[00:00:54] *** adultera1edjedi has joined #opensolaris
[00:05:07] *** vmlemon has quit IRC
[00:06:23] *** sahafeez has quit IRC
[00:06:49] *** infidel has quit IRC
[00:06:58] *** Gman has joined #opensolaris
[00:11:47] <theRealballchalk> is there a rar utility in solaris?
[00:12:22] <jbk> warezing? :)
[00:12:34] <theRealballchalk> noo
[00:12:41] <theRealballchalk> iphone hacking
[00:12:49] <theRealballchalk> too old for warez
[00:12:51] <theRealballchalk> lol
[00:13:08] <jbk> i thought that was the only thing that still used rar :)
[00:13:19] <theRealballchalk> naww
[00:15:54] * delewis loves the warez that has a single rar file that contains 50,000 rar files, which contain zip files.
[00:15:55] *** sahafeez has joined #opensolaris
[00:16:23] *** adulteratedjedi has quit IRC
[00:16:29] <FireflyST> delewis: you get my pm?
[00:16:50] *** phips|mb has quit IRC
[00:24:17] *** BatonT has quit IRC
[00:24:22] *** StFS has joined #opensolaris
[00:24:31] *** adulteratedjedi has joined #opensolaris
[00:26:04] *** ichigo has quit IRC
[00:26:21] <StFS> hello. how do I change the keyboard layout in solaris (nexenta)? I don't have a GUI installed so I'm talking about doing this in the console environment
[00:27:51] <alanc> in real Solaris, kbd -s, not sure about nexenta
[00:29:09] <StFS> alanc: is that permanent then?
[00:29:28] <alanc> think so
[00:30:09] <StFS> hmm... did that... doesn't work :(
[00:31:59] <StFS> ok... another one then... how can I turn off the keyboard beep? again in console only
[00:32:27] <jbk> speaking of keyboards, anyone know of any links on making some of the extra keys (such as volume/mute/etc.) work with X?
[00:32:35] <StFS> I think my collegues are about to kill me
[00:33:05] <StFS> click click backspace BEEP tab BEEP click tab BEEP
[00:34:59] <alanc> jbk: the hard part is usually getting the kernel keyboard driver to pass through the events
[00:35:14] <alanc> casper had a hack in frkit to get them for the Ferrari laptop keys
[00:35:22] <jbk> ahh
[00:37:46] *** adultera1edjedi has quit IRC
[00:38:21] *** duri has quit IRC
[00:38:52] *** perlmonk has quit IRC
[00:41:03] *** duri has joined #opensolaris
[00:41:16] *** sommerfeld has quit IRC
[00:41:20] *** pauliukas has joined #opensolaris
[00:41:27] <pauliukas> I'm installing the latest version of Express...
[00:41:31] *** adultera1edjedi has joined #opensolaris
[00:41:41] <pauliukas> And at the end of the setup I'm getting cannot access errors.
[00:41:43] <pauliukas> Is this normal?
[00:41:44] *** NikolaVeber has quit IRC
[00:52:01] *** jwk404 has quit IRC
[00:58:17] *** adulteratedjedi has quit IRC
[00:58:20] <jbk> ok.. silly question
[00:58:38] *** sponix has joined #opensolaris
[00:58:41] <jbk> was there an ARC case that removed the /etc/hostname.<ifname> functionality from nevada?
[00:59:34] <alanc> don't remember any, or anyone removing it
[00:59:41] <jbk> cause it's not working :)
[00:59:46] *** pfa3rh has quit IRC
[00:59:58] <moazamraja> odd error...
[01:00:03] <alanc> is nwam active?
[01:00:10] <jbk> it appears to _always_ use dhcp (probably nwam related) even if /etc/dhcp.<ifname> doesn't exist
[01:00:14] <alanc> it should work if you're not using nwam
[01:00:16] <jbk> and even if /etc/hostname.<ifname> exists
[01:00:35] <jbk> i would think in such a case, it should defer to the existing configuration
[01:00:36] <moazamraja> Hardware address '03:00:[blah]' trying to be our address 010.013.020.081!'
[01:00:42] *** Pietro_S has quit IRC
[01:00:44] <jbk> least surprise, etc.
[01:00:47] <moazamraja> wtf is up with those zeros before each IP segment
[01:00:53] <CIA-24> dm120769: backout 4693666/6460239: causes 6603251
[01:01:13] <moazamraja> oddly enough, the hardware address is from an interface on the same machine :/
[01:01:24] <moazamraja> but the zeros in the IP, that's new to me
[01:01:53] <jbk> someone probably did printf("%03d.%03d.%03d.%03d", ...)
[01:02:47] <jbk> i should probably dig into the nwam behavior
[01:03:55] <jbk> it's running, but i'm hoping that this is perhaps just interim behavior and won't make it into a solaris release (the behavior i'm seeing)
[01:06:09] *** adulteratedjedi has joined #opensolaris
[01:07:19] *** dunc_ has joined #opensolaris
[01:09:01] *** leal_ has joined #opensolaris
[01:09:14] <leal_> hello all...
[01:09:23] <leal_> Somebody here with AVS skills?
[01:09:26] *** dunc has quit IRC
[01:11:39] <pschow> AVS == Address Verification System?
[01:12:07] <leal_> :)
[01:13:09] <leal_> http://www.opensolaris.org/os/project/avs/
[01:14:05] <SYS64738> where can I find info on about making an smf for clamav ?
[01:18:05] *** adultera1edjedi has quit IRC
[01:19:16] <e^ipi> dump any old manifest &  modify it to your own ends?
[01:19:34] <libkeiser> oh. that's what used to be called SNDR, right?
[01:20:05] <moazamraja> are there known problems with jumpstarting x2100s?
[01:22:14] <palowoda> SYS64738: Or you could look at say the example of blastwaves manifest for clamav example: http://rafb.net/p/GW3kzR24.html
[01:22:46] <SYS64738> thanks
[01:32:00] *** charlesnw has left #opensolaris
[01:33:00] *** karrotx has quit IRC
[01:38:10] *** forelle has joined #opensolaris
[01:38:11] *** forelle has left #opensolaris
[01:38:57] *** forelle has joined #opensolaris
[01:38:58] *** forelle has left #opensolaris
[01:43:02] *** lacaAFK is now known as laca
[01:43:17] *** paul has quit IRC
[01:51:51] *** unixware has joined #opensolaris
[01:52:47] <unixware> hello earths dweller
[01:53:10] *** kadath_ has joined #opensolaris
[01:57:38] *** kadath has quit IRC
[01:57:39] *** Chihan has quit IRC
[01:57:51] *** sahafeez has quit IRC
[01:58:51] *** uebayasi has joined #opensolaris
[02:00:16] <CIA-24> fw157321: 6594328 Solaris "faulted" CPU status is not reported by PICL
[02:00:17] <CIA-24> fw157321: 6592150 PICL SNMP and PRI plug-ins don't handle picld being sent a SIGHUP
[02:03:01] *** dunc__ has joined #opensolaris
[02:05:08] *** dunc_ has quit IRC
[02:05:13] *** dme has quit IRC
[02:06:43] *** libkeise1 has joined #opensolaris
[02:13:27] *** yarihm has quit IRC
[02:13:52] *** paul has joined #opensolaris
[02:14:32] *** adulteratedjedi has quit IRC
[02:15:51] <StFS> ok... what was the command again to see all the disks that can be found?
[02:16:18] <StFS> or, I should say, that are connected (some of them are not formatted)
[02:16:22] <unixware> #df -h
[02:17:21] <unixware> for those are not conected #df -h -unconected
[02:17:39] *** sahafeez has joined #opensolaris
[02:17:45] <jamesd> StFS,  format will list them reguardless of state,  iostat -En  will as well i beleve.
[02:18:12] <jamesd> run   devfsadm  to have the OS scan for new drives.
[02:19:41] *** pauliukas has quit IRC
[02:23:18] *** libkeiser has quit IRC
[02:26:45] <unixware> StFS is the firts person i see on irc from Island :)
[02:28:17] <StFS> :)
[02:33:24] <StFS> urrgh... I'm still having some problems here... I can see in /dev/dsk that I seem to have two controllers (c0... and c1...) but it doesn't make much sense... the second one should have three disks connected but it only shows one...
[02:34:34] <unixware> maybe do u have a raid card? that makes 3 hd on one?
[02:34:52] <StFS> unixware: nope, not a raid card for sure
[02:35:00] <jbk> have you tried running devfsadm?
[02:35:05] <unixware> ohh 0.o
[02:35:27] <StFS> jbk: yes... it runs... and then stops :-/
[02:36:25] <StFS> is there any way for me to see what pci device my second IDE controller is?
[02:37:19] <StFS> cause I was following a howto on how to get a Promise IDE controller card (PCI) to work
[02:37:28] *** Fullmoon has joined #opensolaris
[02:37:45] <StFS> and I want to verify that the second controller that I'm seeing is actually that card
[02:39:49] <unixware> syslog or in dmesg maybe there is the information i think
[02:42:29] <jbk> what you could do (a bit manual) is run (as root) /usr/X11/bin/scanpci
[02:42:46] <jbk> but then, you have to try to patch it to the output of prtconf
[02:45:19] *** sahafeez has quit IRC
[02:46:09] <StFS> jbk: well.. I can see the card in scanpci and I can actually see that same card in prtconf but I can't see anywhere whether that card is being used as that second IDE controller
[02:46:36] <flyingparchment> StFS: cfgadm -alv
[02:46:43] <flyingparchment> (i'm not sure if that works with IDE controllers though)
[02:47:34] <StFS> flyingparchment: thanks, but it doesn't seem to be displaying IDE controllers
[02:51:00] <StFS> ok... well I took a better look at the output of prtconf -pv and it looks like there are two IDE controllers, one is on the mainboard itself and the other one is under the PCI stuff so I'm pretty sure that it's picking it up and that's the c1... I'm seeing
[02:51:25] <StFS> ... however... I don't get why it's only picking up one of the disks that are connected to it :-(
[02:52:35] <StFS> anyways... that's a problem for tomorrow...
[02:52:39] <RElling> b72?
[02:52:45] <StFS> thanks for all your help guys
[02:53:58] *** nachox has joined #opensolaris
[02:54:16] <nachox> PerterB, resistance only lasted so much :P
[02:55:27] <palowoda> The Promise controllers have discussed for years now.
[02:56:18] *** yongsun has joined #opensolaris
[02:57:43] *** alanc is now known as alanc-away
[02:58:03] *** hohum_ has quit IRC
[02:59:41] <StFS> palowoda: hmm... not sure what you mean... but anyways... I found this howto and followed that: http://wiki.complexfission.com/twiki/bin/view/Main/OpenSolarisOS
[03:00:08] <StFS> (at the bottom there's a section about adding a promise IDE controller
[03:00:47] *** sfire||mouse has joined #opensolaris
[03:01:52] *** theRealballchalk has quit IRC
[03:01:59] <StFS> anyways... I think I just figured it out... and it honestly is too embarrassing to reveal...
[03:02:00] *** theRealballchalk has joined #opensolaris
[03:02:27] <StFS> ... but since you guys deserve a good laugh after all the help you gave me I'll tell you.... jumper settings!
[03:02:41] <StFS> all the drives were set to masters :$
[03:02:56] <StFS> I see them all now...
[03:02:57] <unixware> ;S
[03:03:04] <unixware> :P
[03:03:06] <StFS> so thanks again and good night!
[03:03:15] <unixware> good nite
[03:03:24] <StFS> ZFS for the win.... tomorrow!
[03:05:02] <StFS> oh... just one question before I leave... is it normal that dmesg says "MultiwordDMA mode 2 selected" and then "last message repeated 3 times" and then there's a line that says "UltraDMA mode 5 selected"? Does that mean that the drive eventually went into UltraDMA mode 5?
[03:06:45] <StFS> is there any tool that I can use to measure throughput? like hdparm on linux?
[03:07:22] <palowoda> dd, iozone, bonnie, bonnie++
[03:07:39] <palowoda> Just curious what model Promise card you have?
[03:08:03] <StFS> Ultra133 TX2
[03:08:22] <StFS> palowoda: exactly the same one as this guy has: http://wiki.complexfission.com/twiki/bin/view/Main/OpenSolarisOS
[03:08:51] <palowoda> Oh so you did update the driver_aliases file?
[03:09:40] <palowoda> Or added the class to the master file.
[03:10:13] <StFS> files I edited were /etc/system and /boot/solaris/devicedb/master
[03:10:19] <StFS> exactly as it said in his howto
[03:11:01] <palowoda> Ok, wonder why this never got into an RFE that info is like over a year old.
[03:11:08] <StFS> palowoda: do you know bonnie? how exactly am I supposed to use it?
[03:11:11] <StFS> RFE?
[03:11:12] <diomac> whats a good ts client for solaris?
[03:11:41] <palowoda> StFS: Just get the iozone from blastwave.org.
[03:12:14] <StFS> palowoda: well.. I'm trying out nexenta and it has bonnie in the apt repository :/
[03:12:42] <palowoda> Then try bonnie -s 2000  (for a 2G test)
[03:13:07] <palowoda> Try something smaller first bonnie -s 200
[03:13:08] <StFS> but how do I tell it which disk to use???
[03:13:23] <palowoda> What ever directory your in.
[03:13:26] *** Sup3rkiddo has quit IRC
[03:13:28] *** diomac has quit IRC
[03:13:48] *** simford has joined #opensolaris
[03:14:00] *** unixware has quit IRC
[03:14:20] <FireflyST> palowoda: where did you download compiz from?
[03:14:28] <palowoda> hang on.
[03:14:46] <palowoda> http://blogs.sun.com/erwann/entry/3d_desktop_compiz_fusion_0
[03:15:41] <StFS> funny... I can't run bonnie as root :-/
[03:16:02] *** johnlev has quit IRC
[03:16:09] <StFS> damn... I have to get up in 6 hours :@ better get my butt to bed
[03:16:21] <StFS> thanks again everyone!
[03:16:28] <StFS> good nite (again)
[03:16:36] <palowoda> later.
[03:16:44] <StFS> palowoda: count on that ;)
[03:17:09] <FireflyST> Son of a bitch
[03:17:17] <FireflyST> That's where I got mine
[03:17:27] <FireflyST> why the hell didn't it work for me? that sucks
[03:17:38] <palowoda> What kind of hardware do you have?
[03:18:03] <FireflyST> Intel x3100 graphics, c2d 1.5ghz, 2gb memory
[03:18:14] <FireflyST> the hardware check said it was all ok
[03:19:02] <palowoda> Yeah but all the systems I've seen it run on with no problems had Nvidia graphics in them.
[03:19:42] <FireflyST> well that sucks
[03:20:03] <palowoda> Yeah I know Intel makes some really crappy graphics chips.
[03:20:30] <FireflyST> It may just be a support issue
[03:21:00] <palowoda> I heard Intel is busy trying to improve it too.
[03:21:44] <palowoda> Not much you can do with that chipset though.
[03:23:35] <FireflyST> well, I'll buy the upgrade card later if I can
[03:23:53] <FireflyST> Dell had a 128mb nvidia card available
[03:24:16] <FireflyST> it was +$100 to the price though
[03:24:53] <edp> i'm somewhat disappointed with the compiz performance of the Geforce Go 7900
[03:25:17] <e^ipi> does the free trial download of SGD actually ever expire?
[03:25:20] <palowoda> I have a Nvidia 7800GTX in my laptop.  Kind of overkill though.
[03:25:33] <e^ipi> or is it the honour system?
[03:26:56] *** IvanR__ has quit IRC
[03:27:28] <edp> palowoda, are the graphics smooth for you under opensolaris?  I saw a very noticeable performance decrease with compiz on opensolaris as opposed to linux
[03:28:58] <palowoda> When I run it I noticed about 8 percent cpu usage but no real slowdown.  Nothing different from the linux boxes I've been on.  This is on dual core amd cpus
[03:30:16] <ottom> e^ipi: I'm pretty sure it stops working after the 30 days is up.
[03:30:43] <palowoda> For 350.00 a pop it should stop working.
[03:30:46] <FireflyST> Can I install Sun Studio 12 on my OpenSolaris machine?
[03:30:59] <flyingparchment> FireflyST: depends which opensolaris distribution you're using
[03:31:06] <palowoda> Heck it comes with Studio 12.
[03:31:09] <FireflyST> Solaris Express
[03:31:12] <flyingparchment> FireflyST: it will work fine on solaris express, but be aware you cannot use studio 12 to build onnv
[03:31:23] <FireflyST> yeah
[03:31:25] <FireflyST> I know
[03:36:10] *** apersson has quit IRC
[03:36:36] <jbk> hmm.. the ipp listener support in b72 appears to be broken
[03:37:17] <flyingparchment> can vxvm/vxfs make filesystems smaller?
[03:37:25] <jbk> flyingparchment: yes
[03:39:58] <e^ipi> so i'd need a hex editor to keep using it
[03:40:00] <e^ipi> fair enough
[03:48:14] *** yongsun_ has joined #opensolaris
[03:49:20] *** leal_ has quit IRC
[03:50:31] <richlowe> yeesh.
[03:50:34] <richlowe> steleman: ping?
[03:59:31] *** hile_ has quit IRC
[04:04:12] *** jafari has joined #opensolaris
[04:10:02] *** alobbs has quit IRC
[04:10:23] *** alobbs has joined #opensolaris
[04:17:28] <FireflyST> So what would I need the sun compilers to build?
[04:17:48] <flyingparchment> you need studio 11 with a set of patches, http://www.opensolaris.org/os/community/tools/sun_studio_tools/sun_studio_11_tools/
[04:18:32] <e^ipi> where's the build fail w/ studio12?
[04:18:38] <flyingparchment> oh, misread the question, nm.
[04:18:39] <richlowe> various places.
[04:18:48] <richlowe> flyingparchment: I think you have have answered what was meant.
[04:18:50] *** ByteEnable has joined #opensolaris
[04:18:52] <richlowe> flyingparchment: reading it literally it doesn't make sense.
[04:19:03] <flyingparchment> richlowe: i think he means, what software will only build with studio, not gcc
[04:19:03] *** HD9361 has joined #opensolaris
[04:19:08] <ByteEnable> OMG, 5kb/s download of DVD, is this for real?
[04:19:15] <richlowe> e^ipi: there doesn't seem to be an associated keyword, like there was before, but there's a number of bugs you can find searching b.o.o
[04:19:22] <flyingparchment> ByteEnable: wow!  SDLC finally upgraded to 56k
[04:19:49] <ByteEnable> Is there a torrent ?
[04:20:02] <FireflyST> that would be an abhorrent torrent
[04:24:39] *** HD9361 has left #opensolaris
[04:25:04] *** andy_ has joined #opensolaris
[04:26:17] *** halton has joined #opensolaris
[04:26:25] *** Andy_Pease has quit IRC
[04:28:18] *** IvanR_ has joined #opensolaris
[04:31:05] *** ByteEnable has left #opensolaris
[04:34:43] *** Chihan has joined #OpenSolaris
[04:39:43] *** Fullmoon has quit IRC
[04:40:48] *** gnut has joined #opensolaris
[04:40:56] *** gnut has quit IRC
[04:41:25] *** Triskelios has joined #opensolaris
[04:41:30] *** jinx099 has joined #opensolaris
[04:48:45] <libkeise1> has anyone else noticed that src.o.o symbol definition searches against onnv-gate specifically are failing when they shouldn't?
[04:49:22] *** libkeise1 is now known as libkeiser
[04:50:19] <jbk> yes, i've seen that recently
[04:50:31] <Triskelios> ditto
[04:51:43] *** sfire||mouse has quit IRC
[04:56:20] <steleman> richlowe: pong!
[04:56:37] <steleman> sorry for the delay i was busy with a chicken cacciatore
[04:58:12] *** ashner has joined #opensolaris
[04:58:21] *** mypapit has joined #opensolaris
[04:58:35] <mypapit> hey sibikos, aren't you supposed to be working?
[05:01:47] <jmcp> aren't we all?
[05:02:29] <jbk> well you're not broken, are you? :)
[05:02:38] <nachox> no, i'm supposed to be sleeping
[05:03:03] <jbk> well i see sch submitted the new packaging proposal
[05:03:20] <nachox> seriously? where? :)
[05:03:27] <jbk> install-discuss
[05:03:49] <jmcp> about time, too
[05:04:22] <richlowe> Yes.
[05:05:34] <nachox> nod
[05:06:15] <Gman> he did?
[05:06:16] <Gman> woo.
[05:07:36] <nachox> http://mail.opensolaris.org/pipermail/install-discuss/2007-September/005207.html
[05:08:09] <Gman> hawhaw, then dminer throws the constitution at him
[05:08:10] <Gman> nice.
[05:08:14] <Gman> go stop energy!
[05:08:21] <jbk> heh
[05:08:51] <jbk> i have a cousin with the same name, always throws me for a bit when i see posts
[05:08:52] <Gman> oh well, what's another 72 hours
[05:09:27] <jbk> grr this ipp support is going to annoy me
[05:09:28] *** SirFunk has quit IRC
[05:10:43] *** jafari has quit IRC
[05:11:24] <richlowe> Gman: I think that was just reminding Install what they actually needed to do.
[05:11:37] <richlowe> Gman: and ptribble has some very reasonable questions.
[05:12:24] <Gman> not saying they weren't
[05:12:27] <nachox> only 4 and 5
[05:13:11] <richlowe> #1 and #2 are very reasonable, unless you already know what they're doing.
[05:13:19] <richlowe> and that's a rather small set, right now.
[05:13:20] <nachox> the rest of them were already answered
[05:14:48] <nachox> as for 2, they are developing only a packaging system they only seek assistance to integrate the resulting framework in the other projects
[05:15:30] <richlowe> I'm reading #2 as a very different question.
[05:15:53] <richlowe> I'm reading it as "Is this an entirely new system, or are you enhancing the interface to the existing system (and parts of it) to deal with your new goals"
[05:16:02] <richlowe> I already know the answer, but install-discuss doesn't.
[05:16:05] <richlowe> so it's a reasonable question.
[05:16:16] <nachox> and as for the first one, why would develope a prototype if they want to work in something else?
[05:16:41] <richlowe> nachox: because many people develop prototypes to validate their ideas and design.
[05:16:46] <richlowe> that's... well, it's what a prototype *is*
[05:17:38] <nachox> richlowe, the answer to your question is in sch's blog, in the comments of the last entries i believe
[05:17:51] <richlowe> sigh.
[05:18:09] <nachox> hey, that's why they included the links :P
[05:21:29] *** slowhog has joined #opensolaris
[05:21:34] *** jhawk has joined #opensolaris
[05:24:37] *** sfire||mouse has joined #opensolaris
[05:24:47] <jbk> hmm
[05:24:59] <jbk> something doesn't like a hyphen in the print queue name
[05:25:39] <richlowe> printing to what?
[05:25:49] <richlowe> I'm pretty sure whatever I'm printing to has a hyphen in there without issue.
[05:25:59] *** jhawk has quit IRC
[05:26:00] <jbk> with the ipp stuff
[05:26:11] <richlowe> Oh, I'm using plain ol' lp.
[05:26:21] <flyingparchment> ipp, how newfangled
[05:26:22] * dlg yawn
[05:26:30] <jbk> well i'm trying to get my laptop to print to the printer attached to me desktop
[05:26:53] <richlowe> jbk: lpd.
[05:27:06] <richlowe> I gave up on ipp, printers aren't worth the fight. :)
[05:27:11] <flyingparchment> buy a printer with lpd support built-in ;-)
[05:27:16] <jbk> well it's usually running XP
[05:27:28] <jbk> since i seem to have horrible luck w/ wireless cards
[05:27:45] <dlg> which card?
[05:27:53] <jbk> well it has a broadcom built in
[05:28:03] <jbk> i tried to get one that I tought used an atheros chip
[05:28:07] <jbk> turns out it was marvell instead
[05:28:16] <jbk> and i tried a atheros mini-pci card instead
[05:28:18] <dlg> jbk: interesting news on the broadcom: http://marc.info/?l=dragonfly-commits&m=118923223901695&w=2
[05:28:18] *** SirFunk has joined #opensolaris
[05:28:19] * sfire||mouse thinks a printer at work uses ipp, since the mac picks it up broadcasted.  or something that cups understands
[05:28:28] <jbk> that thing would overheat my laptop and cause serious instability
[05:29:09] <jbk> dlg: very interesting
[05:29:16] <jbk> question is, will anyone port it..
[05:29:20] <flyingparchment> for some reason, considering MPIO on linux just seems so wrong to me.
[05:29:21] <dlg> of course they will
[05:29:26] *** karrotx has joined #opensolaris
[05:29:29] <jbk> even though a group reverse engineered the specs on linux
[05:29:34] * flyingparchment suspects the likelyhood of linux failing is higher than that of a path failing :)
[05:29:45] <jbk> and from what they claim, did the whole chinese wall/clean room bit
[05:29:51] <jbk> sun still refused to even consider it
[05:30:06] *** DoYouKnow has joined #opensolaris
[05:30:21] <dlg> which group?
[05:30:42] <jbk> bcm-specs.sipsolutions.net
[05:31:02] <jbk> which appears to be the same group that the dragonfly bsd people used for the specs
[05:31:18] <DoYouKnow> bcm = broadcom?! someone going to write a broadcom driver for opensolaris?! :)
[05:31:19] <dlg> i dont see why a driver for solaris couldnt be based on the dfly work
[05:31:42] <dlg> i can see why it might not be integrated though
[05:31:49] <dlg> but thats more political than technical
[05:31:55] <Tempt> bah
[05:31:56] <Tempt> Politics.
[05:31:57] <Tempt> bah
[05:32:02] <dlg> bah
[05:32:03] <dlg> bah
[05:32:05] * dlg agree
[05:32:28] <Tempt> Don't get me start on GPL license arguments, either. I just won't take the bait today.
[05:32:42] <dlg> my reasons arent to do with the gpl
[05:32:44] <flyingparchment> when you don't pay Sun's insane prices, how much are FC-AL disks compared to SCSI?
[05:32:55] <Tempt> about 5% premium?
[05:33:08] <flyingparchment> hmm, that's not so bad
[05:33:08] <Tempt> unless you're buying on ebay, and then they're about half the price.
[05:33:12] <jbk> it wasn't a license issue
[05:33:20] <e^ipi> it could be chucked on to muryama's pile
[05:33:33] <jbk> i think they were too scared of not being able to prove how the 'cleanliness' of the specs
[05:33:41] <dlg> jbk: exactly
[05:33:56] *** ashner has quit IRC
[05:34:02] *** ashner has joined #opensolaris
[05:34:26] *** sfire||mouse has quit IRC
[05:38:24] <flyingparchment> HP's website is annoying.  there is no "someone should contact me with a quote" link.
[05:41:57] *** victori_ has joined #opensolaris
[05:49:11] *** jinx099 has quit IRC
[05:50:39] *** comay has quit IRC
[05:52:19] <FireflyST> so how do wireless access points work that require a password on a web page? do they just not allow any traffic via closed ports until you sign in?
[05:57:40] *** nachox has quit IRC
[06:01:03] *** slowhog has left #opensolaris
[06:01:14] <dlg> FireflyST: probably
[06:01:21] *** yongsun_ has left #opensolaris
[06:01:58] <victori_> bfu looks nice
[06:02:01] <victori_> or neat
[06:06:58] *** eboutilier has quit IRC
[06:15:06] <Triskelios> FireflyST: look up "walled garden"
[06:15:12] *** Triskelios has quit IRC
[06:15:34] *** jHoNDoE has joined #opensolaris
[06:22:37] <jbk> ahh cool.. got a workaround
[06:24:39] *** dlynes has quit IRC
[06:27:39] *** sibikos has quit IRC
[06:28:24] *** laca has quit IRC
[06:37:55] *** Gman has quit IRC
[06:43:58] *** DoYouKnow has quit IRC
[06:56:22] *** derchris has quit IRC
[06:56:29] *** derchris has joined #opensolaris
[07:00:36] *** theRealballchalk has quit IRC
[07:00:58] *** bthornton has joined #opensolaris
[07:02:11] <bthornton> Netbeans won't execute on my fresh Solaris Express installation--when I run it in a shell, it replies: "Cannot find java. Please use the --jdkhome switch.".  However, 'java' is in the $PATH.  Any ideas?
[07:02:28] <flyingparchment> bthornton: set JAVA_HOME to /usr/java
[07:03:36] <victori_> any decent package manager for open solaris?
[07:03:38] <bthornton> just did it but same result...
[07:04:48] <bthornton> although I just ran it with the --jdkhome switch that it suggested--I'd done it before as "--jdkhome=/usr/java" but it didn't work
[07:04:49] <Tempt> perhaps we should put blastwave.org in the topic? for packages?
[07:05:01] <flyingparchment> it's already in the onjoin
[07:05:02] <bthornton> apparently I needed to omit the equals sign
[07:05:03] <bthornton> 'cause it works now
[07:05:12] <Tempt> flyingparchment: apparently nobody reads that.
[07:05:19] <flyingparchment> you think people read the topic? :)
[07:05:40] <Tempt> actually, scroll-scroll .. it isn't in the onjoin.
[07:05:46] *** flyingparchment has left #opensolaris
[07:05:46] *** flyingparchment has joined #opensolaris
[07:05:53] <flyingparchment> hmm thought it was
[07:06:08] <Tempt> Nope.
[07:06:24] <Tempt> Perhaps we could replace the pastebin link in the topic with blastwave
[07:06:36] <Tempt> (given everyone just uses their favourite pastebin anyway)
[07:08:06] *** jHoNDoE has left #opensolaris
[07:08:27] <bthornton> remotely related question: Is there such a thing as a "UNIX keyboard"?  I'm on a Sun workstation right now and some of the keys are clearly in different positions than on the PCs I use--and I was curious if this is just a Sun thing or if this particular layout is typical of Unix boxen?
[07:08:35] <WickedWicky> Hi all!
[07:09:30] <Tempt> Sun actually calls it the UNIX layout, yes.
[07:09:37] <Tempt> the control key is in the correct place.
[07:10:07] <bthornton> so presumably this is the layout I would find on other Unix boxen?  like SGIs and such?
[07:10:27] <bthornton> aahhh it's the Control key that's driving me nuts!  I keep turning the CAPS on when I want to hit Control on this thing! :p
[07:10:42] *** m0le has joined #opensolaris
[07:10:47] *** Andy_Pease has joined #opensolaris
[07:11:02] <WickedWicky> I dont see anything weird regarding the layout on my sun keyboard o.O
[07:11:05] <bthornton> and the backspace is also funkily positioned.  ah well, I'll get used to it
[07:11:08] <WickedWicky> jus that I have extra keys on the left
[07:11:15] <Tempt> xmodmap ...
[07:11:35] <bthornton> yeah it's just weird to me 'cause I've got a Pee-Cee keyboard right to the left of this one and the differences are obvious
[07:11:38] <bthornton> I guess this is just convention.
[07:11:53] <Tempt> xmodmap
[07:11:59] <Tempt> change the layout to suit your needs
[07:12:02] <WickedWicky> oh look it
[07:12:03] <bthornton> kll
[07:12:07] <bthornton> *koo
[07:12:08] <WickedWicky> Tempt made IRC promotion
[07:12:12] <Tempt> or
[07:12:16] <Tempt> push the PC off the desk
[07:12:21] <Tempt> and enjoy the purity of UNIX.
[07:12:51] <bthornton> oh I can feel the purity of UNIX...
[07:12:54] <bthornton> it's in the air
[07:12:56] *** andy_ has quit IRC
[07:13:15] <bthornton> it accounts for about a 10 degF increase in the ambient temperature in my room actually :p
[07:13:24] <WickedWicky> yippee-ka-yee
[07:13:42] <Tempt> oh, big machine
[07:13:43] <Tempt> all the better.
[07:13:48] *** m0le has quit IRC
[07:13:58] <Tempt> nothing keeps up with WickedWicky's home datacentre though
[07:14:00] <bthornton> yeah I'm going to love this thing in the winter
[07:14:05] <Tempt> he's got bleeding edge technology in there
[07:14:12] <bthornton> oooh
[07:14:13] <WickedWicky> \o/
[07:14:17] <WickedWicky> pwnage!
[07:14:26] <Tempt> His smallest machine cost over $20k when it came out.
[07:14:27] <WickedWicky> ph3ar my sparcstation 20 and ultra 5
[07:14:34] <bthornton> my technology is bleeding edge if I just pretend it's the year 2000
[07:14:45] <Tempt> year 2000 bleeding edge
[07:14:46] <Tempt> hmmm
[07:14:48] <Tempt> E6000?
[07:14:49] <WickedWicky> my sparcstation 20 is so acting 2007
[07:14:55] <Tempt> (in the sun world, that is)
[07:15:00] <bthornton> no
[07:15:02] <bthornton> U80
[07:15:07] <WickedWicky> running Apache 2, PHP5, latest exim
[07:15:12] <Tempt> U80
[07:15:13] <Tempt> Nice machine
[07:15:16] <Tempt> solid.
[07:15:32] <Tempt>  big bulgy power supply to deliver the rich chunky amps that those four CPUs need
[07:15:39] <bthornton> yes indeed.
[07:15:53] <bthornton> and the two-foot-long UPA framebuffer
[07:16:00] <Tempt> Heh.
[07:16:04] <Tempt> That's the one.
[07:16:26] <bthornton> which seems to have the 2D performance of a 10-year-old PC with a $10 VGA card..
[07:16:37] <bthornton> I've yet to figure that one out
[07:16:46] *** victori__ has quit IRC
[07:17:21] <Tempt> Anyway, coffee time.
[07:17:23] *** victori__ has joined #opensolaris
[07:17:41] <brendang> Tempt: is it MOSUG day today?
[07:17:42] <bthornton> mmm good idea
[07:17:49] <bthornton> I'll warm it on my U80!
[07:19:20] *** bthornton has quit IRC
[07:22:06] * jmcp wonders whether "bleeding edge" for WickedWicky actually means "makes you bleed if you bump it"
[07:23:51] <jbk> :)
[07:25:22] <jbk> bah.. there's a nevada-specific hcl right?
[07:25:41] *** Gman has joined #opensolaris
[07:25:57] <jmcp> I didn't think there was
[07:27:18] *** dme has joined #opensolaris
[07:28:37] * jmcp heads off for more coffee
[07:29:41] *** oninoshiko has joined #opensolaris
[07:29:52] * g4lt-sb100 spikes jmcp's coffee
[07:32:45] *** WickedWicky has quit IRC
[07:33:35] *** patrick__ has joined #opensolaris
[07:33:41] <patrick__> oh noes!
[07:33:46] <patrick__> screen just hung on me!
[07:33:55] *** patrick__ is now known as WickedWicky
[07:40:23] *** diomac has joined #opensolaris
[07:40:46] <diomac> how do you get colour in the default terms in b72?
[07:40:59] <jmcp> g4lt-sb100: awww, thanks, just what I needed
[07:41:06] <jmcp> diomac: colour ls is devilspawn
[07:41:18] <diomac> i hate not having it
[07:41:45] <jmcp> diomac: pull down the gnu version via blastwave.org
[07:41:48] *** sdash has joined #opensolaris
[07:41:56] *** sdash has left #opensolaris
[07:42:13] <richlowe> jmcp: /usr/gnu/bin/ls
[07:42:16] <jmcp> that too
[07:42:17] <richlowe> it's *delivered* devil spawn.
[07:42:23] <jmcp> if you're on a sufficiently recent built
[07:42:24] <jmcp> build
[07:42:38] <jmcp> richlowe: just because it's delivered doesn't make it less demonic
[07:42:50] <richlowe> jmcp: I *certainly* didn't intend to imply otherwise.
[07:42:55] <jmcp> :-)
[07:42:57] <jmcp> I know
[07:43:40] <diomac> cheers
[07:43:53] <richlowe> jmcp: that said, Brendan totally needs to add syntax highlighting to ged.
[07:43:54] <WickedWicky> heading to the bus station
[07:43:57] <WickedWicky> see ya all later
[07:44:02] *** estibi has quit IRC
[07:44:02] <oninoshiko> bye
[07:44:13] <jmcp> richlowe: yeah, he does
[07:44:44] *** werlock has joined #opensolaris
[07:52:12] <LeftWing> Haha, ed with syntax highlighting.
[07:56:07] <Tempt> Hmm
[07:56:16] <Tempt> Can I integrate my colour vmstat at the same time?
[07:56:55] <richlowe> color is over used, surely vmstat needs a sound track.
[07:56:57] <Tempt> brendang: MOSUG. That sounds awful.
[07:57:02] <Tempt> brendang: But yes, today.
[07:57:03] <richlowe> the benny hill music, growing fast as load rises.
[07:57:29] *** werlock has quit IRC
[07:58:17] <g4lt-sb100> yakkety sax?
[07:58:23] <Tempt> richlowe: Hmm, that gives me an idea ...
[07:58:31] <richlowe> g4lt-sb100: Yes, but it seems better known as the benny hill theme :\
[08:04:48] <diomac> ah man what a time for blastwave to go down, grrr
[08:05:27] *** Gekkko has joined #opensolaris
[08:10:20] <WickedWicky> it's very up and running here diomac
[08:10:45] <diomac> can't get any packages atm
[08:10:54] <diomac> might try a mirror
[08:12:09] *** dunc__ has quit IRC
[08:14:22] <brendang> richlowe: I've been waiting a long time for ksh93 to exist so that I can use its sound capabilities in a number of tools.
[08:15:01] <Tempt> brendang: Writing to the audio device not good enough for you?
[08:15:14] <brendang> Tempt: try it
[08:15:18] <Gman> soundls, evil.
[08:15:31] <Tempt> brendang: Seems to work for me.
[08:15:43] <brendang> Tempt: try creating waveforms programatically
[08:16:00] <brendang> Tempt: I don't mean cat > /dev/audio; I mean create new sounds :)
[08:16:23] <Tempt> brendang: aaah
[08:16:26] <Tempt> you want to be fancy.
[08:17:28] <brendang> Tempt: python came with a library that almost did it ok. was scratchy.
[08:17:50] <brendang> Tempt: roland has a ksh93 script that does it much better.
[08:18:08] <g4lt-sb100> brendang, what, #include mlib.h too complex? ;P
[08:18:09] <Tempt> brendang: Then all you need is some interesting synthesis
[08:20:08] <brendang> g4lt-sb100: generating the .au output was a bit of work. although last time I was trying was in Perl.
[08:20:36] <g4lt-sb100> well mlib has a whole passl of FFTs, so one could pretty much generate any waveform needed
[08:21:36] <Tempt> How dull.
[08:21:49] <Tempt> Try building waveforms using sections from a mandelbrot
[08:22:05] <Tempt> you can perform seamless transforms by slowly twiddling parameters and views.
[08:22:42] <g4lt-sb100> right, so you can do SFTs as opposed to FFTs
[08:24:10] <brendang> g4lt-sb100: yes. then follow au(4) to wrap it as 8-bit u-law or what have you, and you might be able to play a tone. hmm. I'd rather there was a library that did that for me, so I could issue "play F#,0.25" or some such, which is what python/ksh93 provide.
[08:24:53] <g4lt-sb100> brendang, okay, wirte it in TRS-80C extended basic then
[08:25:16] <brendang> g4lt-sb100: anyhow, have a go at playing a single note to /dev/audio, and see how you feel about the effort. :)
[08:25:40] <g4lt-sb100> so ksh93 is fully compatible with a wannabe-game console from 1981.  W00H00!!!!
[08:26:15] <Tempt> Wouldn't it be easy just to have a few appropriate .aus lying around and dump them into the audio device?
[08:26:26] <g4lt-sb100> in fact, the play "f",0.25 is the exact syntax from TRS-8C BASIC
[08:26:28] <Tempt> Then you wouldn't be tied to a shell that isn't in stable Solaris yet.
[08:26:42] <brendang> Tempt: to a point. I was writing some tools that pushed that to the limit.
[08:27:39] <g4lt-sb100> brendang, what, porting dungeos of daggorath?
[08:28:06] <brendang> no
[08:28:07] *** oninoshiko has quit IRC
[08:29:06] <g4lt-sb100> do you also want the peek and pokee functions from BASIC as well?
[08:30:12] <brendang> g4lt-sb100: not really, peek/poke as an interface is asking for trouble.
[08:31:52] <g4lt-sb100> what about the line ()-() and circle () functions?
[08:32:35] <brendang> g4lt-sb100: are you trying to be helpful, or trying to be an arsehole?
[08:32:56] <g4lt-sb100> if you have to ask, my work here is complete
[08:33:44] <brendang> uh-huh - work wasting my time.
[08:34:22] <g4lt-sb100> no, pointing out in a sarcastic manner that you're asking for things that are only really useful in a flat memory model
[08:34:38] <flyingparchment> what does producing audio have to do with memory model?
[08:35:01] <g4lt-sb100> the play function from basic only workeed because you could rasonably hit the csound chip at a certain place and it would make a certain sound
[08:35:37] <flyingparchment> new computers have this fancy thing where you can create a waveform at a certain frequency and actually play it
[08:35:45] <flyingparchment> even with virtual memory
[08:36:53] <brendang> g4lt-sb100: so what. like I said earlier, python provides a library for playing tones. now ksh93 does.
[08:37:09] *** yarihm has joined #OpenSolaris
[08:37:10] <WickedWicky> it's a freaking shell
[08:37:13] <WickedWicky> who's using this?
[08:41:26] <e^ipi> brendang: you're a dvorak person, non?
[08:42:04] <brendang> e^ipi: yes
[08:42:34] <e^ipi> any way to use it as the layout for OBP on a sparc?
[08:44:11] <richlowe> not that I know of.
[08:44:13] <brendang> e^ipi: not that I've seen documented. I recall seeing how the OBP setup keymaps somewhere (maybe in forth), and thinking that it was doable - but quite hard.
[08:44:24] <e^ipi> drats
[08:44:36] <richlowe> hellish, ain't it?
[08:44:43] <palowoda> Shesh how much time are you going to spend typing in OBP?
[08:45:02] <brendang> e^ipi: not a problem if you are using SSPs or terminal consoles.
[08:45:22] <richlowe> palowoda: it's still a pain  the ass when something goes wrong and you can't touch-type qwerty anymore.
[08:47:42] <palowoda> I guess the demand hasn't been great enough for the feature.
[08:49:00] *** diomac has quit IRC
[08:50:27] <Tempt> boyd: ping
[08:50:37] <boyd_> pong
[08:50:40] *** boyd_ is now known as boyd
[08:50:52] <Tempt> Are you trammin' out down to Sun?
[08:51:42] <e^ipi> for that matter, the sparc & dvorak aren't happening at all at the moment
[08:52:05] <flyingparchment> missing dvorak support is one of those things that annoys me a lot more than stuff like colour ls
[08:52:40] <Tempt> gah, you guys and your fake keyboards
[08:52:45] <e^ipi> gnome has a tool to set the keymap, but like everything else useful in that godforsaken desktop, it doesn't work
[08:52:58] <Tempt> you could try using xmodmap
[08:53:01] <Tempt> and remapping it that way
[08:53:04] <flyingparchment> i run solaris in vmware, so there aint no X
[08:53:11] <palowoda> e^ipi:  The real issue with sparc is it's not popular as a client usage type thing.
[08:53:37] <palowoda> But that is a hardware issue.
[08:57:47] <richlowe> e^ipi: I just build a copy of setxkbmap
[08:57:54] <richlowe> e^ipi: hence my multi-year assault on alanc about that RFE :)
[09:00:41] <palowoda> richlowe: That RFE number == ?
[09:00:43] *** mikefut has joined #opensolaris
[09:00:50] <boyd> Tempt: Getting a lift
[09:00:57] <boyd> Many sun guys next door
[09:01:06] <Tempt> boyd: Aah.
[09:01:10] <Tempt> boyd: See you there then
[09:01:15] <boyd> Ok.. will do.
[09:01:18] * boyd out.
[09:01:26] <richlowe> 4756705
[09:02:38] <palowoda> Your right 5 years is multi year.
[09:02:42] <WickedWicky> have fun at the meeting boyd
[09:02:59] <richlowe> I didn't file the original, I've only been bugging him for 2 and a bit.
[09:03:49] <palowoda> Strange that the fix is known.
[09:06:54] <e^ipi> it is known... build setxkbmap
[09:06:58] *** infidel has joined #opensolaris
[09:07:31] <palowoda> It might need ARC approval.
[09:11:14] *** theRealballchalk has joined #opensolaris
[09:12:12] *** helpmehome has quit IRC
[09:12:29] *** nostoi has joined #opensolaris
[09:16:10] *** diomac has joined #opensolaris
[09:17:53] *** tsoome has quit IRC
[09:18:18] <Tempt> Already, clearin' out for the day
[09:18:20] <Tempt> catchya later.
[09:18:52] <jmcp> ciaociao
[09:19:25] *** cydork has joined #opensolaris
[09:20:14] <WickedWicky> have fun
[09:20:15] <WickedWicky> tchau
[09:20:36] <WickedWicky> I'm gonna have my own meeting today
[09:20:49] <WickedWicky> not osol though, way funnier, capex calculations for next year \o/
[09:21:50] <jmcp> WickedWicky: "give me all your money"
[09:22:16] <WickedWicky> yea
[09:22:25] <WickedWicky> we need 1.000.000 euros
[09:22:28] <WickedWicky> "that's a bit much"
[09:22:32] <WickedWicky> yeah no shit, but we need it
[09:22:36] <quasi> WickedWicky: remember to leave money in there for a few extra e25k's just in case
[09:22:37] <flyingparchment> 1.0 euros?  :)
[09:22:58] <WickedWicky> flyingparchment: I'm european eh, . is for thausands, comma is for decimals
[09:23:04] <flyingparchment> ya ino
[09:23:27] <WickedWicky> 1.00 euro would be doable
[09:23:34] <WickedWicky> we'd be done with the meeting within 5 seconds
[09:24:29] <WickedWicky> what's thec++ compiler in SunSpro?
[09:24:34] <e^ipi> CC
[09:24:41] <WickedWicky> thanks
[09:24:51] *** Cyrille has joined #opensolaris
[09:25:20] <flyingparchment> we do our budgets in reverse, we decide how much we need, and then think about how to get hold of it
[09:25:24] <WickedWicky> well
[09:25:47] <WickedWicky> the thing is, we always plan unforseen projects in our budgets, and spare parts
[09:26:06] <WickedWicky> cause the way it goes is: OH LOOK!! OUR COMPETITOR DOES xxxxxxx WE WANNA DO THAT TOO
[09:26:16] <WickedWicky> fact they forget is that requires hardware
[09:26:37] <quasi> WickedWicky: sounds as if you work for a telco/isp ;)
[09:26:48] <WickedWicky> har di har
[09:26:50] <flyingparchment> i wonder if i can find a pub that opens before 10am
[09:26:51] <WickedWicky> :P
[09:26:58] <WickedWicky> what? sure you can
[09:27:05] <WickedWicky> unless you're in timbuctu
[09:27:06] <quasi> WickedWicky: been there, done that, didn't like the tshirt
[09:27:19] <WickedWicky> me neither and I am browsing shops for new t-shirts
[09:27:32] <WickedWicky> preferable with a cool print
[09:28:52] * WickedWicky is wearing one now saying: NO! I DO NOT LIVE IN A WINDMILL"
[09:29:13] <quasi> WickedWicky: our market people were brilliant at pulling: so when will this feature be ready? we're going to be advertising it next week and then refusing to change the one week deadline because they'd booked advertising 6 months ago and couldn't change it at such a short notice
[09:29:18] <flyingparchment> WickedWicky: do you wear clogs?
[09:29:32] <WickedWicky> if clogs are woodenshoes, then of course
[09:29:34] <quasi> flyingparchment: all dutch people do
[09:29:59] <WickedWicky> quasi: oh hey! so I am not alone with this frustration then :p
[09:31:10] <quasi> WickedWicky: nope, general telco/isp operating mode: "time to market" ~= one week after the competitors come up with a new product
[09:31:42] <WickedWicky> not to mention we have to beg for every euro-cent and they get zilions for marketting campaigns that are doomed to fail
[09:31:51] <WickedWicky> oh hello cutey
[09:31:55] * WickedWicky is not in favor for blonde
[09:32:03] <WickedWicky> but the one I see in the train now is
[09:32:12] <WickedWicky> clear
[09:32:13] *** yarihm has quit IRC
[09:32:29] *** cmihai has joined #OpenSolaris
[09:32:56] <quasi> WickedWicky: so when are you posting pics? ;)
[09:32:57] * WickedWicky is listening to Savatage - If I go away
[09:33:04] <WickedWicky> of the blonde?
[09:33:22] <WickedWicky> I dont think she'd appreciate it if I took pics
[09:34:15] <quasi> who knows - maybe she would ;)
[09:34:37] <WickedWicky> i can always say in court she was flirting with me
[09:38:09] <cmihai> Hooray for Windows update, it's just installed the same update 67 times so far.
[09:38:16] <quasi> if you have a mobile with a cam, you could just pretend to accidentally hit the button :)
[09:38:37] <quasi> cmihai: switch to solaris
[09:39:03] <cmihai> Eh, I've got Solaris, this is a development machine for .net crap.
[09:39:24] <cmihai> Just makes me wonder how can they fuck updates so badly.
[09:39:48] *** tsoome has joined #opensolaris
[09:39:49] <e^ipi> it's microsoft...
[09:40:01] <e^ipi> they take pride in fucking things up in new and innovative ways
[09:40:28] <cmihai> No shit. Funny part is they (try) fix a remote root hole in Visual Studio with this patch hehe
[09:40:47] *** bengtf has quit IRC
[09:47:27] <jamesbrink> cmihai can you explain the zone setting for locked memory? idont understand what that does
[09:51:12] <cmihai> Yeah, sure mate.
[09:51:25] <cmihai> Just give me a second to solve this damn update issue.
[09:51:34] <cmihai> Developers are bugging me ;P
[09:51:39] <jamesbrink> im thinking it has somthing to do with what can be sent to swap but im lost, thanks again =)
[09:51:48] <cmihai> Not sent.
[09:52:25] <cmihai> Locking prevents memory from being paged out by the system.
[09:52:31] <cmihai> Paging is sending stuff to swap.
[09:52:42] <jamesbrink> ahh
[09:53:25] <cmihai> That is (or should be) real memory.
[09:53:28] <cmihai> Not swap.
[09:53:58] <cmihai> You could lock say 64 RAM to make sure the box always has 64 MB of real memory and doesn't get trashed
[09:54:13] <cmihai> Just don't set it too high, or you'll mess things up.
[09:54:36] <cmihai> If you could lock it all, it may serve as additional security too...
[09:54:56] <cmihai> The issue with swap is that it stays there even after reboot... and I've seen systems compromised from data in the swap.
[09:55:19] <cmihai> Kind of like MacOS's File Vault encryption. Strong, but kept the key in plain text in un-encrypted swap. Not fun.
[09:55:42] <cmihai> So most crypto stuff actually try to lock the memory they use (or disable swap)
[09:56:51] <jamesbrink> lol ok cool
[09:56:54] <jamesbrink> thanks alot man
[09:57:57] *** dennisPoland has joined #opensolaris
[09:58:12] *** ballchalk has joined #opensolaris
[10:00:10] *** ballchalk has quit IRC
[10:00:12] *** dmarker has quit IRC
[10:00:17] *** ballchalk has joined #opensolaris
[10:00:23] *** dmarker has joined #opensolaris
[10:02:38] <tsoome> oh nice, damn why they dont include 64bit nss_winbindd in snv_71/snv_72 like they have no clue about 64bit opterons.... "good" integration work indeed:(
[10:05:17] *** mav_eric has joined #opensolaris
[10:05:18] *** peteh has joined #opensolaris
[10:05:24] <mav_eric> good morning
[10:05:54] <trochej> Mornin'
[10:06:14] *** ballchalk has quit IRC
[10:08:38] <jamesbrink> what was the name of the tool that is similar tot op
[10:08:40] <jamesbrink> to top
[10:08:45] <pablomh> prstat
[10:08:49] <jamesbrink> thanks
[10:09:02] *** ballchalk has joined #opensolaris
[10:09:24] <mav_eric> can someone a newbie to opensolaris help with getting my printer to work within a virtual machine?
[10:12:00] <infidel> anyone know where i can get a cisco vpn client for solaris x86_64?
[10:12:45] *** victori__ has quit IRC
[10:13:09] *** victori has joined #opensolaris
[10:15:55] <cmihai> infidel, yeah, vpnc
[10:16:38] <mav_eric> infidel: I got mine from the university. I don't know if cisco offers one since their download area is only accessible via login
[10:16:49] <cmihai> infidel, h nice, damn why they dont include 64bit nss_winbindd in snv_71/snv_72 like they have no clue about 64bit opterons.... "good" integration work indeed:(
[10:16:53] <cmihai> erm
[10:16:56] <cmihai> infidel, http://www.unix-ag.uni-kl.de/~massar/vpnc/
[10:17:08] <infidel> thanks
[10:17:15] <cmihai> Use this. It's open source, works with Cisco VPN concentrators, and works on any OS. It's also better imho... lot less crap.
[10:17:30] <infidel> ok cool
[10:18:31] <cmihai> mav_eric, isn't that SPARC only?
[10:19:23] <cmihai> I've got vpnclient-solaris-4.6.00.0045-k9.tar.Z, and it's all they have for Solaris at the download center. And it's only for SPARC.
[10:19:28] <mav_eric> cmihai: could be. I only have x86 OpenSolaris since a few hours
[10:19:35] <mav_eric> the machines at the university are sparcs yes
[10:20:18] <cmihai> Ah, OK. Thought they might have actually built an x86 version I didn't know about.
[10:23:08] <infidel> cmihai, did you import a pre-existing cisco profile into vpnc?
[10:23:54] <cmihai> Nah, I configured the concentrator, so I know all the settings :P
[10:24:03] <infidel> ahh
[10:24:11] <cmihai> You just need to know the pre-shared secret, your username and password.
[10:24:27] <cmihai> Read the manpage or whatever, it's a very simple text-based config file.
[10:24:52] <infidel> cmihai, i guess that where the cisco password cracker fits in huh?
[10:24:59] <cmihai> The what now?
[10:25:08] <cmihai> Are you wacked?
[10:25:23] <cmihai> You should know all that from your sysadmin. It's the same settings you're using to connect to the VPN with the Cisco client.
[10:25:24] <infidel> i have the pcf files
[10:26:08] *** helpmehome has joined #opensolaris
[10:26:38] <cmihai> Ehem, are you legally allowed to connect to that VPN?
[10:26:46] <cmihai> If so, just ask the damn sysadmin.
[10:27:08] *** ballchalk has quit IRC
[10:30:28] <infidel> it's not that easy, we don't have a "sys admin' we can just directly. first we have to call the helpdesk, then we have to submit a trouble ticket there are 7 pcf files i have from different concentrators accross the US. the have to find and call every admin to see if they know the password. we have over 100,000 employees. i work in japan and when it's day there it's late night here.
[10:30:47] *** Vorbis_w has quit IRC
[10:31:43] <dlg> mmmjapan
[10:31:55] <cmihai> Mmmgrmrmr Japan
[10:32:34] * quasi wouldn't want to be the sysadmin that 100000 people could call directly
[10:32:58] <infidel> anyway i have a win and linux client, but i installed opensolaris to learn it and wanted a working cisco vpn client
[10:33:46] <infidel> quasi, me either
[10:34:16] <quasi> does cisco have a version of their vpn client for solaris?
[10:34:19] <cmihai> Well, it's up to you get the pre-shared key stuff from the .pcf files...
[10:34:23] <cmihai> quasi, only SPARC.
[10:34:31] <mav_eric> quasi: yes. sparc only
[10:34:35] <infidel> quasi, for sparc
[10:34:46] <cmihai> And the Linux one is crap too. You can't lxrun it either (kernel modules and blobs and crap)
[10:34:53] <quasi> hmmm, there's my excuse to get sparc at home ;)
[10:35:07] <infidel> lol
[10:38:58] <Jigsaw> hey all, when I truss something ... how and where can I find out what the number after the open syscall means ... for example? open("/path/to/file", O_RDONLY) = 59 ...
[10:39:37] <Cyrille> errno
[10:39:52] <Cyrille> in the case of an open it's just the inode number
[10:40:15] <Jigsaw> so 59 is the inode number?
[10:40:29] *** Vorbis_w has joined #opensolaris
[10:40:42] <Cyrille> yes, so if you then see calls to read(59, ...) or write(59, ...) or close(59) they will be related to that /path/to/file.
[10:41:13] <Jigsaw> ah cool.
[10:41:17] <Cyrille> the return code depends on the function, man <function>, section return values is your friend.
[10:41:51] <Cyrille> (because read or write will return the number of bytes read or written, some others will just return 0 or -1 for success or error, etc.)
[10:42:42] <Jigsaw> ok thanks a lot
[10:45:44] *** Erwann has joined #opensolaris
[10:45:58] *** yarihm has joined #OpenSolaris
[10:48:26] *** NikolaVeber has joined #opensolaris
[10:55:45] <palowoda> quasi:  If you are successfull in getting a sparc at home via the cisco vpn problem you could start an industry revolution.  Not that anyone would understand how you did it but understanding it wouldn't be needed. :)
[10:55:58] <palowoda> s/Not/Note
[10:58:57] <palowoda> By the way CISCO is rather high on Linux internal but not so high on open source in general.
[11:00:21] <palowoda> It was horrible when they bought Linksys and had to close the wireless router technology with a proprietary architecture.
[11:01:09] <quasi> palowoda: "getting" as in buying something off ebay or finding something redundant in one of our serverrooms ;)
[11:01:13] <cmihai> Ever tried their VPN client on Linux:
[11:01:26] <quasi> cmihai: I have
[11:01:32] <cmihai> Damn thing uses horrible blobs they call kernel modules and only works on some distros without a lot of work.
[11:01:39] <palowoda> The Ebay factor.  No future.
[11:01:40] <cmihai> It also killed another module here... hehe
[11:01:58] <quasi> cmihai: worked fairly well on my gentoo install
[11:02:07] *** migi has joined #opensolaris
[11:09:28] *** dosiu has quit IRC
[11:11:40] <jamesbrink> cmihai, i setup a zone with the settings you suggested to me the other day, and my and a friend are just kinda playing around with the zones, and i notriced he can still bring my sytem to a crawl, i setup cpu-shares in the global zone and in his zone, if my system is idle while he is pouding his zone, would it be normal for the gloabl zone to respond very slowly
[11:11:43] *** Snake007uk has joined #opensolaris
[11:11:50] <jamesbrink> for example just doing a pwd or ls is really really slow
[11:11:55] <cmihai> Duh.
[11:12:17] <cmihai> He can still saturate the CPU. Did you set up locking to only on thread?
[11:12:18] <cmihai> As in, core.
[11:12:34] <jamesbrink> no
[11:12:39] <cmihai> isainfo -vvv
[11:12:40] <cmihai> how many?
[11:13:32] <cmihai> http://unixsadm.blogspot.com/2007/08/solaris-container-with-1-cpu-resource.html - try making a resource pool and give it CPUNumber-1 CPUs and share those along your containers or something.
[11:13:33] <palowoda> "-vvv" where do you get such ideas?
[11:13:35] *** mirror176 has joined #opensolaris
[11:13:54] <cmihai> stuck key
[11:14:16] <cmihai> Point is keeep a dedicated CPU for the global zone
[11:14:36] *** yongsun has quit IRC
[11:15:22] *** simford has quit IRC
[11:15:50] <palowoda> Boy you do have a suck keys.  Somebody told me you can wash your keyboard in a dishwasher.  As long as it goes through the dry cycle.  Never tried it though.
[11:16:01] *** Gropi has joined #opensolaris
[11:16:37] <jamesbrink> cmihai isainfo prints out 32-bit i386 aplications, sse,fxsr,amd_3d_now,amd_3dnowx,amd_mmx,mmx,cmov,sp,cx8,tsc,fpu
[11:16:47] <cmihai> Only once?
[11:16:54] <jamesbrink> yeah
[11:16:55] <cmihai> How many CPU cores do you have
[11:16:58] <cmihai> 1?
[11:16:59] <jamesbrink> just one
[11:17:02] <jamesbrink> yes
[11:17:13] <richlowe> isainfo would only print out once per-ISA, no materr how many...
[11:17:18] <richlowe> matter, too.
[11:17:21] <jamesbrink> this isnt an important machine, i was just testing this crap out
[11:17:34] *** Cyrille has quit IRC
[11:17:58] *** Snake007uk has quit IRC
[11:18:01] <cmihai> Bah, I meant /usr/sbin/psrinfo -v
[11:18:10] *** Snake007uk has joined #opensolaris
[11:18:54] <jamesbrink> just lists the one proc
[11:18:58] <jamesbrink> 1150MHZ wooohoo
[11:19:11] <mirror176> a keyboard can be washed as long as what you use to clean it doesnt dissolve the plastic or paints used. then just make sure its dry with no left over residue
[11:19:12] <palowoda> Heh a real winner
[11:19:13] <cmihai> Uhh
[11:19:24] <cmihai> mirror176, I know, I give mine showers
[11:19:32] <cmihai> Hell, I've given motherboards showers :-)
[11:19:42] <palowoda> I use glass cleaner on mine.
[11:20:02] <cmihai> I use those on LCDs
[11:20:05] <mirror176> it was fun watching my dad put a friends laptop motherboard under a running faucet
[11:20:08] <cmihai> Just the rag
[11:20:11] <jamesbrink> when mine get to dirty i buy a new one
[11:20:16] <cmihai> lol
[11:20:35] <palowoda> Wireless keyboards can get expensive.
[11:20:49] <jamesbrink> meh whats 30 bucks every 3-6 months lol
[11:21:17] <palowoda> 60.00 for a Logitech S500 wireless keyboard and mouse.
[11:21:28] <palowoda> S510
[11:22:06] <jamesbrink> when i used to play alot of video games i had to buy cheap $10 keyboards because i would get mad and smash them
[11:22:11] <jamesbrink> so i would buy like 3 at a time lol
[11:22:22] <palowoda> jamesbrink: You just need a faster system.
[11:22:52] <palowoda> You shouldn't be playing with salvage stuff from the junkyard.
[11:23:28] <jamesbrink> lol
[11:23:59] <palowoda> Leave that to the guys who think Ebay is your savour.
[11:25:00] <jamesbrink> so when you use the FSS, the shares only matter if the diff zones are not idle correct
[11:25:11] <jamesbrink> if they are idle the one system will get all of the cpu time
[11:25:27] <jamesbrink> i would like to have a quad core to play with
[11:26:39] <palowoda> How much money do you have?
[11:26:57] <palowoda> I mean we are talking hardware you know.
[11:27:06] <jamesbrink> me im broke all the time =) but i do have a core 2 dou here, i might load it up on here soon and try it out again
[11:27:20] <jamesbrink> the machine we were playing with is my workstation at work
[11:27:25] <jamesbrink> its a peice of shit lol
[11:27:38] <quasi> jamesbrink: FSS only kicks in when you've got a shortage of cpu
[11:28:52] <palowoda> jamesbrink: What kind of cpu you running opensolaris on anyways?
[11:29:27] <jamesbrink> amd 1ghz 768mb of ram
[11:29:34] <palowoda> amd what?
[11:29:39] <jamesbrink> athalon
[11:30:16] <palowoda> Athlon can in at 1ghz?  I don't remember that. When.
[11:30:51] <jamesbrink> well its 1.somthing
[11:30:58] <jamesbrink> i dont remember the exact specs
[11:31:09] <jamesbrink> it was a peice of shit i found laying around with a motherboard
[11:31:12] <palowoda> Well it's an old K6 architecure isn't it?
[11:31:15] <jamesbrink> my company is cheap lol
[11:31:22] <jamesbrink> no its not
[11:31:29] <jamesbrink> oh idk
[11:31:40] <palowoda> Athlons are 64bit.
[11:32:06] <palowoda> Your isa was 32bit.
[11:32:34] <jamesbrink> i thought the older athlons were 32bit
[11:32:43] <jamesbrink> unless im thinking of the wrong name
[11:33:13] <palowoda> Oh retro naming that is right. Helps companies increase thier performance. :)
[11:33:15] *** BatonT has joined #opensolaris
[11:33:31] <LeftWing> Not all Athlons are 64bit, surely.
[11:33:34] *** ichigo has joined #opensolaris
[11:34:06] <delewis> palowoda: bullshit.
[11:34:09] <delewis> Opterons are 64-bit.
[11:34:14] <delewis> all Athlons are most certainly not.
[11:34:26] <LeftWing> I have an Athlon Thunderbird under my desk which is very, very 32bit.
[11:34:32] <cmihai> Sigh.
[11:34:37] <palowoda> delewis: You understand reto naming?
[11:34:52] <cmihai> It's 32 bit. Really.
[11:34:59] <delewis> palowoda: I understand I have a 1.4GHz Thunerbird sitting in a peecee beside me that is most certainly 32-bit.
[11:35:10] <delewis> there were slot-based Athlons that were way under 1GHz.
[11:35:20] <palowoda> Ok when did that Athlon name come out?
[11:35:23] <cmihai> Only the new generation Athlon 64 is... 64 :-).
[11:35:38] <LeftWing> 1999, according to Wikipedia.
[11:35:49] <cmihai> O... long time ago, I remember something along the lines of 500-600Mhz stuff
[11:35:58] *** Cyrille has joined #opensolaris
[11:36:15] <jamesbrink> i have had 3 athlons
[11:36:19] <jamesbrink> they were all 32bit
[11:36:36] <jamesbrink> the only 64bit adm i had was a turion (sp?) and it was a mobile proc
[11:37:33] <cmihai> That's the new Athlon 64 based stuff, yes. And btw palowoda it's not a k6 architecture, it started with a K7 core...
[11:38:46] <jamesbrink> i bought my first intel this year
[11:38:53] <jamesbrink> set me back 500 bucks for a core 2 dou
[11:39:00] <palowoda> Ok so james has a 32bit k7 core yes?
[11:39:02] <jamesbrink> now its come down like at least 200 in price
[11:39:04] <jamesbrink> probobly more
[11:39:41] <cmihai> jamesbrink, probably some newer core, k7 were only used in the beginning.
[11:40:20] <Vanuatoo_> how do I resolve sendmail error message about qualifing short name without shutting down the service?
[11:40:21] <cmihai> There's lots more cores like Thunderbird, Palomino, Barton etc. I think he has a Thunderbird core.
[11:40:40] <cmihai> Vanuatoo_, get yourself a REAL FQDN, and use that.
[11:40:45] <palowoda> Now is the k7 core 32bit (what 256/512L2) worth running with the newer versions of opensolaris.  Or is it just frustration?
[11:41:01] <jamesbrink> yeah i used to have a barton... not that i know wtf the diff is between them all
[11:41:02] <cmihai> Frustration.
[11:41:21] <palowoda> I'd say a PINA.
[11:41:41] <cmihai> I'd say a minimum of 2GB RAM and a 2-way 64bit 2Ghz x86-64 CPU is a minimum to enjoy all the cool features like ZFS, Containers and all that.
[11:41:56] <palowoda> I would agree.
[11:42:19] <palowoda> Friggen Intel P4 are crap for solaris.
[11:42:43] <jamesbrink> how about the dou
[11:42:58] <jamesbrink> or is it duo
[11:42:59] <cmihai> Look into dual core Opterons or Core 2
[11:42:59] <jamesbrink> lol
[11:43:06] <cmihai> No, Core Duo is not 64 bit
[11:43:07] <cmihai> Core 2 is.
[11:43:20] <jamesbrink> yeah i currently have a core 2 duo
[11:43:22] <delewis> isn't Core 2 Duo extremely buggy?
[11:43:27] <jamesbrink> 4gb of ram
[11:43:39] <cmihai> delewis, meh, so is the new Opteron.
[11:43:41] <palowoda> Why is Core 2 Duo buggy?
[11:43:52] <quasi> delewis: not all that much more than the usual intel procs
[11:43:53] <cmihai> I've seen the eratta, I've seen what the community "leaders" had to say...
[11:44:00] <cmihai> And I don't give a crap.
[11:44:12] <cmihai> Sure, you want data integrity, run your stuff on RISC
[11:44:21] <palowoda> Sun employees buy a lot of apple laptops to use with opensolaris.
[11:44:31] <palowoda> That is Core 2 Duo yes?
[11:44:36] <cmihai> Yes
[11:44:37] <quasi> palowoda: yes
[11:44:51] <palowoda> Not that I would by an Apple laptop.
[11:44:56] <jamesbrink> i got the shaft lol i have a ppc mac mini, guess what i use it for
[11:44:57] <cmihai> The new stuff uses the Intel Santa Rosa platform, very well supported on Solaris.
[11:44:57] <jamesbrink> irc
[11:44:59] <jamesbrink> thats it
[11:45:06] <cmihai> Get an Acer 5920g
[11:45:35] <palowoda> Santra Rosa stuff looks reasonably priced.
[11:45:41] <cmihai> Core 2 Duo 2Ghz, 4MB cache, NVIDIA 8600 GT / 256 RAM + 1GB Turbocache, 250GB hdd SATA, 1GB TurboMemory, subwoofer stuff..
[11:45:50] <cmihai> This laptop kicks ass.
[11:45:57] <cmihai> Plus it was half the price of a worse Apple.
[11:46:00] *** timsf has joined #opensolaris
[11:46:04] <timsf> hi all
[11:46:09] <palowoda> What is this TurboMemory stuff?
[11:46:19] <cmihai> Onboard memory disk for cacheing.
[11:46:26] <palowoda> Ah.
[11:46:30] <cmihai> Usually for disk and stuff.
[11:46:38] <quasi> TurboMemory? does it have a turbo button? ;)
[11:46:45] * cmihai throws a 386 at quasi
[11:46:56] <palowoda> So in combination with a 7200rpm drive is should kick ass.
[11:46:59] <cmihai> Anyway, I'm very happy with mine, it kicks ass.
[11:47:06] <cmihai> palowoda, 250GB, built in.
[11:47:09] <cmihai> SATA-II too.
[11:47:28] <cmihai> It's got firewire, so you can add fast external storage too.
[11:47:37] *** Vanuatoo_ has quit IRC
[11:47:50] *** Vanuatoo has joined #opensolaris
[11:47:53] <palowoda> You got a 8600GT video too?
[11:47:56] <cmihai> And with 4GB of DDR2-800 (upgraded) it _really_ kicks ass.
[11:48:03] <cmihai> palowoda, yeah, it's faster then my desktop :-)
[11:48:14] <palowoda> I believe it.
[11:48:27] <Vanuatoo> what about warning syslog.d line 24, loghost could not be resolved
[11:48:34] <cmihai> And I have a 7900GT in the desktop hehe
[11:49:00] <palowoda> I'm using 7800GT's in my desktop.  Got them for 89.00 new.
[11:49:12] <jamesbrink> does solaris support hardware acceleration with nvidia?
[11:49:16] <cmihai> Yep
[11:49:17] <quasi> Vanuatoo: define loghost in your /etc/hosts file
[11:49:20] <palowoda> of coarse.
[11:49:22] <cmihai> That's why I got an NVIDIA laptop :-)
[11:49:23] <jamesbrink> cool
[11:49:26] <kohju> I have a 8600GTS in the desktop... :)
[11:49:29] <cmihai> palowoda, it's built into SXCE now.
[11:49:44] <jamesbrink> i got an 8800GTX and i dont do crap with it lol
[11:49:45] <cmihai> You need to download from NVIDIA for Solaris 10... but nobody uses that on a workstation :P
[11:49:50] <jamesbrink> i just use putty and ie 7
[11:49:53] <palowoda> And Sun won't send the CDDL police out to get you because you use Nvidia. :)
[11:50:14] <kohju> Today, I bought Thinkpad T61(with nVidia)....
[11:50:41] <cmihai> 8400?
[11:50:53] <cmihai> Heh, this has 8600GT :P
[11:50:57] *** uebayasi has quit IRC
[11:50:58] <jamesbrink> chihai nobody uses what on a workstation?
[11:51:28] <cmihai> Solaris 10, it's more of a production / server thing. Unless you're doing stuff like software development or CAD that requires stable bla bla, just use SXCE.
[11:51:55] <delewis> I use Solaris 10 on my workstation. :-)
[11:52:32] <cmihai> Your SPARC workstation? :P
[11:52:43] <palowoda> I'm going to pickup one of these and cmihai gave me more of a go ahead: http://www.discountlaptops.com/index.php?section=specs&model_id=1451&category_id=2&category_theme=c1
[11:52:46] <cmihai> kohju, I kind of hate Thinkpads.. small disks.
[11:52:48] <delewis> yes :-P
[11:52:51] <cmihai> 80-100GB... nasty stuff.
[11:53:03] <delewis> and I would use Solaris 10 on my laptop if it weren't for the fact I'd have to run 3/05.
[11:53:11] <trochej> There is 200 GB disc in mine
[11:53:38] <cmihai> Oh, that's nice.
[11:53:49] <trochej> 7200 rpm
[11:53:53] <cmihai> Really nice :-)
[11:54:03] <cmihai> Slap in 4GB RAM and you're set.
[11:54:06] <Vanuatoo> quasi, /etc/hosts file is a symlink to /etc/inet/hosts and that file is readonly
[11:54:18] <Vanuatoo> should I change the attributes and edit, or use some tool?
[11:54:21] <trochej> cmihai: I'm waiting for them to come, 2GB now :)
[11:54:28] <cmihai> Neat.
[11:54:33] <palowoda> That DDR2 800mhz stuff really makes a difference.
[11:55:00] <cmihai> It does
[11:55:09] <quasi> Vanuatoo: I just use :wq! in vi ;)
[11:55:18] <cmihai> May I suggest a bag?
[11:55:21] <cmihai> DICOTA Executives AluSlight ;-)
[11:55:55] <palowoda> Londen Fog is also good.
[11:56:11] <cmihai> http://www.dicota.com/start.php?office=62&page=page_productDetails&ds=345 :-)
[11:56:45] <palowoda> Oh that is like a Haliburton.
[11:57:04] <palowoda> I have two of them. :)
[11:57:19] <quasi> cmihai: have you added wheels or do you have a cart to move it around?
[11:57:21] <trochej> Right now I'm looking at making that: http://blogs.sun.com/icedawn/entry/a_physical_mashup_opensolaris_laptop
[11:57:26] <trochej> :)
[11:57:29] <cmihai> quasi, cart mate? It
[11:57:30] <dennis-> hmm, solaris works as an iscsi initiator?
[11:57:31] <cmihai> It's very small.
[11:57:37] <cmihai> dennis sure.
[11:57:42] <dennis-> k, thanks
[11:57:59] <kohju> I build 2 harddisks to thinkpad :)
[11:58:04] *** Cyrille has quit IRC
[11:58:23] <trochej> kohju: Pictures?
[11:58:27] <trochej> :)
[11:59:12] *** cmihai has quit IRC
[12:00:02] <kohju> one of them is installed to WinVista.  another is for SXCE71.
[12:00:23] <kohju> no, picture :D , maybe, maybe..
[12:00:50] <trochej> kohju: What, you deny us our geek pr0n? :)
[12:01:42] <trochej> You really need Vista?
[12:02:17] <PerterB> Everybody needs Vista... Microsoft said so, so it must be true.
[12:02:44] <palowoda> Yet MS backtracked and extended XP.
[12:03:30] *** Cyrille has joined #opensolaris
[12:04:19] <PerterB> it'll be like solaris 8, they'll never get rid of it
[12:05:01] <delewis> Microsoft has the ability to can XP. Sun does not have that ability with Solaris 8, because of customer demand.
[12:05:10] <delewis> (specifically, in the government and healthcare sector)
[12:05:57] <palowoda> goverment and healthcare can stay right where they want.  I don't care.
[12:06:02] <PerterB> at least with this new Etude thingie, there's scope to consolidate those sol8 apps onto newer servers
[12:06:22] <PerterB> palowoda: they will, they don't care about you either :)
[12:06:42] <palowoda> I don't want them to spend anymore tax dollers on anything.
[12:07:04] <palowoda> dollars.
[12:07:13] <mav_eric> bye
[12:07:16] *** mav_eric has left #opensolaris
[12:07:16] <delewis> it would be nice to see the government adopt Etude, so they can run something modern on their laptops.
[12:08:10] <palowoda> Absolutely not.  The US goverment is irresponsible in their spending as it is.
[12:08:40] <delewis> what does that have to do with adopting Etude?
[12:08:47] <palowoda> Support cost.
[12:09:04] <delewis> they're spending a lot more money for Solaris 8 support, I suspect.
[12:09:07] <palowoda> They will make up any excuse to spend money.
[12:09:08] <jamesbrink> wtf is etude
[12:09:17] <delewis> jamesbrink: Solaris 8 in a zone, basically.
[12:09:56] <jamesbrink> oh
[12:10:19] <palowoda> A good way to degrade your Solaris 8 performance too.
[12:10:20] <delewis> ideal for customers that still need to run Solaris 8 apps, but run them on a modern platform.
[12:10:37] <delewis> and gives you a chance to consolidate your Solaris 8 applications.
[12:11:19] <delewis> palowoda: not if it's newer hardware, and if not, the performance degradation in a zone is only minimal for workloads.
[12:11:29] <PerterB> yeah, my previous boss is gagging to get hold of it for exactly those two reasons (they have all sorts of odd sol8 shit they can't get rid of, on shonky hardware)
[12:11:42] <palowoda> Solaris 8 isn't a performer to begin with.
[12:12:08] <PerterB> since when?
[12:12:39] <palowoda> Solaris 8 isn't winning any new benchmarks now is it?
[12:12:55] <palowoda> Even Sun doesn't care how Solaris 8 performs.
[12:13:00] <logic_> account off 0
[12:13:01] <delewis> palowoda: most of the stuff that Solaris 8 doesn't perform well in is kernel-level, and will thus benefit from running as a zone (as it'll be using a newer Solaris release atop) -- at least if I understand Etude correctly.
[12:13:11] <delewis> you'll get FireEngine, because you're running Solaris 8 in a zone.
[12:13:43] <logic_> sorry, my irc went foobar
[12:13:55] <delewis> palowoda: most customers don't care how well Solaris 8 performs, as it performs "good enough" in most cases.
[12:14:14] <delewis> the healthcare sector wouldn't be entrenched in it if that weren't the case, don't you think?
[12:14:22] <PerterB> and "significantly better than 2.6", which was the key at the time
[12:15:24] <palowoda> delewis: I wish I could agree with your dream but it's not happening.
[12:15:37] <delewis> what isn't happening?
[12:15:44] <palowoda> Etude
[12:16:05] <PerterB> And what makes you think that?
[12:16:12] <delewis> why? it's very attractive to any Solaris 8 shop.
[12:16:15] <palowoda> How much does it cost?
[12:16:29] <richlowe> there's way more interesting (technical) questions.
[12:16:39] <delewis> palowoda: it's free...
[12:16:46] <palowoda> Solaris 8 is free?
[12:16:51] <delewis> has been.
[12:16:53] <delewis> for up to 8 procs.
[12:17:01] <delewis> what planet have you been on?
[12:17:14] <richlowe> and, beyond that, if you're wanting to migrate an *existing solaris 8 install*...
[12:17:30] <PerterB> richlowe: go on...
[12:17:47] <palowoda> Yes what is the price?
[12:18:00] <PerterB> (the example I saw was basically transplanting a flar from a live system into a zone)
[12:18:28] <palowoda> The price was?
[12:19:19] <palowoda> Free, don't pay attention to that man behind the curtain.
[12:19:36] <PerterB> I haven't actually seen the price, I didn't think it would be free but you never know... the issue is how much are these sites currently spending on support for sol8
[12:19:53] <delewis> and how much Sun is spending on Solaris 8.
[12:20:16] <delewis> that's definitely a non-zero value at the moment, and unless something like Etude takes hold, will remain non-zero.
[12:21:02] <palowoda> Wake me up when Etude makes a market impact.
[12:21:54] <delewis> you almost sound like kaiwai, now.
[12:21:59] *** dosiu has joined #opensolaris
[12:23:04] *** Gman is now known as GmanAFK
[12:28:01] <PerterB> Now if only I could get shot of these bloody unpatched Solaris 7 boxes the same way...
[12:28:22] <delewis> Solaris 7? unpatched? what?
[12:28:39] <PerterB> don't ask...
[12:29:01] *** mypapit has quit IRC
[12:29:14] <delewis> plese tell me you've informed whoever is responsible for this that those Solaris 7 systems can be rooted with a telnet client.
[12:29:23] <PerterB> of course
[12:30:57] * PerterB looks under the rock... last patch cluster seems to have been installed in Dec 2000, with one single patch in Nov 2002 *sigh*
[12:31:37] <delewis> what hardware?
[12:32:00] <PerterB> SUNW,Ultra-5_10
[12:32:05] <delewis> oh, dear.
[12:32:17] <PerterB> long overdue for being taken out back and shot
[12:32:34] <PerterB> however, mine is not to reason why
[12:32:35] <delewis> that's an understatement. :-)
[12:32:46] <delewis> I was hoping you weren't going to say Ultra 5 or Ultra 10, but I knew it was coming.
[12:33:23] <PerterB> I think one of 'em is an E450 too... it's all great stuff
[12:33:34] <delewis> real winners there. :-)
[12:33:42] <PerterB> oh yes
[12:34:20] <delewis> now if you only had an E10k, you'd have "Sun's worse systems collection"
[12:34:55] <PerterB> awww, be fair, the 450 was a great little workhorse in its day... it's just that that day was 5 years ago
[12:36:00] <delewis> I always thought that a box of its size should have more than 4 procs.
[12:36:17] <PerterB> true, but you couldn't half load it up with disks
[12:38:42] <PerterB> it's about time I got a new job anyway, so I'm treating them as "not my problem"
[12:39:04] <Vanuatoo> Is next SXDE going to be released this month?
[12:39:10] <delewis> least you aren't supporting any Pizzaboxen :-)
[12:40:09] * delewis imagines a lone SPARCstation 2 that's been running so long, it's been lost in the last building renovation.
[12:40:16] <PerterB> just the one at home which I can't quite switch off yet
[12:40:25] <PerterB> which is indeed a SS2 :)
[12:42:14] <kohju> trochej: only for adobe software and Office2007.
[12:43:44] <kohju> trochej: sorry, I was away from keyboad.
[12:46:01] <trochej> kohju: NP, I am often too away
[12:46:12] <trochej> kohju: Sad reality, isn't it?
[12:46:21] *** triplah_ has joined #opensolaris
[12:48:47] *** nostoi has quit IRC
[12:50:33] <kohju> trochej: Yes. So, I bought win4solaris for Adobe's and MS's ones :)
[12:52:37] <trochej> kohju: How it works?
[12:52:55] <madhatter> Just a quick question, as I am stuck in the middle of the zfs documentation: is it possible to add mirroring to a pool later when starting with one or two non-redundant disks?
[12:53:37] <delewis> yes, zpool attach.
[12:54:55] <madhatter> delewis: Thank you
[12:55:18] <kohju> trochej: too slow.... :'(  It need fortitude for me.
[12:56:48] <kohju> trochej: maybe, I think that it is based on qemu...,isn't it. I don't know it used kqemu.
[13:01:16] <jamesbrink> has anyone here built E17 on solaris?
[13:01:17] *** yarihm has quit IRC
[13:02:09] <quasi> jamesbrink: benr is probably the expert
[13:02:31] <jamesbrink> i was just curious how long it takes lol
[13:02:44] <triplah_> the build is quick
[13:02:59] <triplah_> e17 is quite small comparatively
[13:03:28] <triplah_> benr seems to have been busy lately though. i think his docs are out of date :(
[13:03:57] <quasi> http://www.cuddletech.com/e17/ - does keep a warning about out of date
[13:03:58] <jamesbrink> there is alot of little packages to build
[13:04:25] <jamesbrink> yeah thats what im reading now
[13:05:06] <jamesbrink> but i cannot get that stupid FAM package
[13:05:07] *** Vorbis_w has quit IRC
[13:05:21] *** huw-l has joined #opensolaris
[13:06:44] <jamesbrink> this is just a slow ass pc i guess
[13:07:17] *** halton has left #opensolaris
[13:07:58] *** deather_ has joined #opensolaris
[13:10:45] *** deather has quit IRC
[13:10:46] *** deather_ is now known as deather
[13:17:46] *** cmihai has joined #OpenSolaris
[13:26:33] *** dennisPoland has left #opensolaris
[13:27:08] *** FelipeBare has joined #opensolaris
[13:33:39] *** _dennis has joined #opensolaris
[13:37:29] *** Baton has joined #opensolaris
[13:41:55] <asyd> "QNX Software Systems is going open source -- sort of."
[13:41:56] <asyd> nice
[13:43:48] <trochej> asyd: link?
[13:44:30] *** BatonT has quit IRC
[13:45:19] <quasi> trochej: http://www.qnx.com/news/pr_2471_1.html (guessing)
[13:46:18] *** yarihm has joined #OpenSolaris
[13:53:32] <trochej> thnx
[14:14:10] *** master_o1_master has joined #opensolaris
[14:18:16] *** triplah_ has quit IRC
[14:20:58] <trochej> http://www.youtube.com/watch?v=kr42ARHwFPU
[14:26:04] *** master_of_master has quit IRC
[14:32:48] *** Risky has joined #opensolaris
[14:35:19] *** ashner has quit IRC
[14:37:04] *** Gekkko has quit IRC
[14:40:01] <jbk> morning
[14:41:09] <trochej> mornin'
[14:44:27] <trochej> Coffee?
[14:44:54] <madhatter> trochej: Please ;)
[14:49:17] * Berny too please!
[14:49:33] <Berny> double espresso if possible :-)
[14:49:42] <trochej> :)
[14:51:29] * jbk is strange -- doesn't drink coffee
[14:51:37] <cmihai> YOU SUCK
[14:51:39] <cmihai> I mean, hi :-)
[14:52:12] <cmihai> Keys are like, right next to each other. Especially after a quadruple espresso
[14:52:24] *** dunc_ has joined #opensolaris
[14:53:02] *** TBCOOL has quit IRC
[14:53:53] <jbk> :)
[14:54:28] <victori_> ok so any decent package management projects for solaris? or am I stuck with a limited option like pkgsrc?
[14:54:44] <victori_> and bfu a good idea to keep a system up to date?
[14:55:32] *** Gekkko[PDA] has joined #opensolaris
[14:56:04] <Gekkko[PDA]> Twitter? Are you fucking serious?
[14:57:14] <trochej> cmihai: Key are like one key, after such espresso :)
[14:57:27] <cmihai> :-D
[14:57:46] <cmihai> victori_, www.blastwave.org, pkg-get, like apt-get.
[14:57:56] <victori_> apt sounds nice
[14:57:57] <cmihai> Wrapper around native SVR-4 packages
[14:58:09] <cmihai> (like apt is around dpkg / rpm) -> that's a good thing.
[14:59:05] <victori_> and do you recommend keeping updated with bfu?
[14:59:08] *** diomac has quit IRC
[14:59:11] <cmihai> no
[14:59:24] <cmihai> You'll end up with a system that is out of sync.
[14:59:29] <victori_> so what is the preferred way of keeping a server updated
[14:59:32] <cmihai> It's a much better idea to LiveUpgrade.
[14:59:33] <richlowe> cmihai: and?
[14:59:46] <cmihai> And you have something to fall back to in case shit happens.
[15:00:03] <richlowe> Oh, I totally managed to misread the context.
[15:00:06] <richlowe> so, yeah, my bad.
[15:00:07] <cmihai> victori_, I don't recommend SXCE on a server :-)
[15:00:14] <cmihai> You know, a real server.
[15:00:23] <victori_> sxce = solaris express?
[15:00:26] <tomww> LiveUpgrade plus patches in case of the Sun Solaris 10
[15:00:29] <cmihai> The main production Oracle db or application cluster or whatever.
[15:00:42] <cmihai> victori_, no, SXDE = Solaris Express
[15:00:53] <cmihai> SXCE - Solaris Express Comunity Edtion. The only thing BFU works on.
[15:00:54] <tomww> or LiveUpgrade plus regular Upgrade of OpenSolaris/SXDE to stay current
[15:01:02] <victori_> I see
[15:01:10] <cmihai> As in "OpenSolaris"
[15:01:43] <victori_> I am looking into learning opensolaris , and hoping for better tomcat performance than what I am getting out of my freebsd rack
[15:01:53] <trochej> But you may just install Debian or Windows 2003 server and sleep well
[15:02:02] <trochej> Got a rack?
[15:02:03] <trochej> Nice?
[15:02:09] *** eboutilier has joined #opensolaris
[15:02:26] <cmihai> I suggest you look at Project Glassfish.
[15:02:34] <cmihai> It's the Java reference platform Application Server.
[15:02:36] <victori_> eh decent keeps on ticking but not too enthusiastic with freebsd's pthread / java performance
[15:03:04] <victori_> cmihai: don't need a full blown j2ee container when my projects revolve around spring/hibernate
[15:03:23] <victori_> unless grizzly can be run without the j2ee bloat
[15:03:24] <cmihai> I find glassfish easier and lighter... and faster too.
[15:03:37] <victori_> how much ram does it take for you?
[15:03:47] <madhatter> As we are talking about live upgrade here: does this work with sxce?
[15:03:53] <cmihai> Yes.
[15:04:02] <cmihai> LiveUpgrade works with anything from Solaris 8 to 11
[15:04:18] <cmihai> https://glassfish.dev.java.net/ - give it a spin mate. Benchmark for your own platform.
[15:04:31] <madhatter> cmihai: I am just trying to figure out how to resize the slizes to get a second slice for live upgrade available.
[15:04:32] *** ocr has quit IRC
[15:04:41] <cmihai> dump / restore :P
[15:05:14] <cmihai> victori_, you can get http://cooltools.sunsource.net/coolstack/ if you just want tomcat. Good for the benchmark even if you decide on Glassfish after.
[15:05:36] <victori_> ya I did notice that
[15:05:51] <victori_> but won't that be redundant if I have pkg-get?
[15:06:00] <cmihai> Don't use pkg-get for this
[15:06:01] <cmihai> Read this:
[15:06:02] <cmihai> victori_, read this too: http://unixsadm.blogspot.com/2007/08/installing-coolstack-on-solaris-and.html -> Installing CoolStack, creating the SMF manifests and
[15:06:15] <cmihai> Putting the tomcat && friends in a container (zone)
[15:06:34] <cmihai> Also, CoolStack is optimized for Solaris and bla bla, it's better mkey? :P.
[15:06:45] <victori_> I am skeptical on that
[15:06:55] <cmihai> Why?
[15:07:19] <cmihai> You can have CoolStack and pkg-get CSW stuff at the same time.
[15:07:26] <cmihai> They won't conflict. Then test both.
[15:07:27] <victori_> well ill try both
[15:07:38] *** hile_ has joined #opensolaris
[15:07:45] <victori_> thanks for all the information
[15:08:31] <flyingparchment> does vxvm work with sun cluster?
[15:08:41] <jmcp> flyingparchment: yes
[15:08:54] <jmcp> flyingparchment: it should be there in the well-written doco from docs.sun.com
[15:09:09] <flyingparchment> i'm sure the docs are there, just needed a quick answer and i'm lazy ;-)
[15:09:41] <flyingparchment> since it looks like all enterprise storage solutions for linux suck, i may actually be able to deploy solaris here.  that would be nice.
[15:09:42] <jmcp> gah
[15:10:19] <cmihai> victori_, one reason to use CSK though mate: it doesn't depend on 1000 things like blastwave does. Easier to deal with, and I'd consider it more "stable". Just make sure you integrate everything with SMF and put them in containers, makes your life easier after.
[15:10:48] <victori_> out of curiosity why bother dealing with containers?
[15:10:51] <cmihai> flyingparchment, yeah, and the stuff that's there is very distro dependent.
[15:10:53] *** nachox has joined #opensolaris
[15:11:09] <victori_> it isn't like I use jails for each hosted deamon on freesd
[15:11:10] <flyingparchment> cmihai: yeah, we don't use RHEL, so veritas on solaris would actually be easier than linux anyway
[15:11:19] <flyingparchment> (not to mention cheaper)
[15:11:24] <cmihai> victori_, Security, performance (you can actually give CPU shares, memory locks / min / max and so on), share CPU's between containers, or lock them to specific CPUs and such.
[15:11:38] <jmcp> victori: (a) because you can, (b) because they allow a certain sort of administrative abstraction, and (c) because they're very, very secure
[15:11:39] <cmihai> Easier to manage too.
[15:11:42] <victori_> oh neat but I doubt ill use that
[15:11:48] <cmihai> Your loss.
[15:11:49] <cmihai> It's _easy_
[15:12:02] <cmihai> Takes like 5 minutes, uses virtually no additional resources...
[15:12:40] <Gekkko[PDA]> tasty.
[15:13:01] <cmihai> It also allows you to isolate containers with applications so they won't interfere with eachother in case you have multiple clients. You can go as far as to bind them to specific CPU's and fixed memory size (locked, cannot be paged) and so on.
[15:13:24] <cmihai> I've had issues when an application went out of control and blocked the whole machine.
[15:13:45] <nachox> cool! sun acquired lustre :)
[15:13:54] <cmihai> This way, it's nice and isolated, there's no way it can damage the other containers or the root zone.
[15:14:19] <jbk> heh i'm even planning on setting up containers on my home pc :)
[15:14:22] <cmihai> nachox, cool... I wonder how this will reflect in ZFS.
[15:14:36] <Gekkko[PDA]> what's lustre?
[15:14:38] <victori_> and you would say brandz has better linux compat than what freebsd-6.2 has?
[15:14:41] <cmihai> victori_, PS: you get actual performance boosts using Containers on top of ZFS... add compression and you're set
[15:14:42] <jmcp> Gekkko[PDA]: cluster file system stuff
[15:14:45] <cmihai> Gekkko[PDA], filesystem
[15:14:52] <Gekkko[PDA]> i see.
[15:14:53] <cmihai> victori_, yeah, it's better.
[15:14:58] <flyingparchment> nachox: seriously?
[15:15:03] <cmihai> victori_, but not by much... it's like a Linux Compat jail.
[15:15:17] <nachox> flyingparchment: http://www.sun.com/aboutsun/pr/2007-09/sunflash.20070912.2.xml
[15:15:24] <victori_> I am just expecting it to be enough to run oracle-xe
[15:15:46] <victori_> freebsd's compat barely runs it , some stuff does not work correctly like the tns listener
[15:15:59] <flyingparchment> victori_: if this isn't for production, why not just use real oracle on solaris?
[15:16:00] <cmihai> Oracle makes Solaris stuff mate
[15:16:05] <victori_> well not xe
[15:16:08] <cmihai> Just grab Oracle 10g Solaris
[15:16:12] <cmihai> You don't need XE
[15:16:16] <cmihai> Grab Enterprise
[15:16:20] <cmihai> It's free for development phases.
[15:16:29] <cmihai> You only need to pay once you put it in production....
[15:16:43] <victori_> we have a small oracle7 production db and I need to have it exporting data to a back up db (oracle-xe in his case)
[15:16:57] <cmihai> Just setup RAC mate :P
[15:17:04] <nachox> i really hope sun opensources that as cddl
[15:17:08] <cmihai> Or migrate the whole thing to Oracle 10 inside a container....
[15:17:27] <victori_> can't, the programmer who made our point-of-sale app hit the bucket and stuck with oracle7 for compat reasons
[15:17:34] <cmihai> I've had to migrate some Oracle 7 stuff from OpenVMS machines too, it's not big deal.
[15:17:52] <cmihai> So run Oracle 7 on Solaris.
[15:17:58] <victori_> no source to our point of sale app, so stuck with oracle7
[15:18:13] <victori_> its binded to oracle7 libs etc
[15:18:29] <Berny> jmcp: got a minute? can you explain how biosdev works? from what i gathered so far it asks the bios what disks are there, get some data from the bios, compares them and reads the first block of each disk and tries to match them?
[15:18:55] <jmcp> yeah, that's pretty much it
[15:19:43] *** pde has quit IRC
[15:19:52] <Berny> hmm, so basically i could confuse it when i have two disks of the same model and dump one disk to the other? (like mirroring the beasts)
[15:19:54] *** yarihm has quit IRC
[15:20:07] <jmcp> yeah
[15:20:16] *** Sup3rkiddo has joined #opensolaris
[15:20:18] *** pde has joined #opensolaris
[15:20:22] <Berny> right that i verified :-)
[15:20:39] <jmcp> and if the bios only adheres to the very, very basic spec (vintage 1995 iirc) which doesn't specify how to handle most things which we assume as default and sensible these days
[15:20:57] <cmihai> victori_, at least, with Containers you can bind your zone to one CPU, give it 1GB of RAM (maximum Oracle XE can use) so it won't waste additional resources (those can be dedicated). But if you actually manage to lxrun Oracle XE, that would be something fun. Should run if it doesn't depend on Linux kernel modules though.
[15:21:08] <cmihai> Just install a CentOS zone.
[15:21:09] <Berny> i wonder now a fixed bios would help here... would it provide more data about the disks?
[15:21:20] <cmihai> Hell, it's a very bad idea, but it's your funeral :-).
[15:21:48] <jmcp> Berny: it might, but don't count on it
[15:21:54] <trochej> And use the new and tasty lx26 zone
[15:22:27] <StFS> hello... is there some "user friendly" alternative to the solaris fdisk? something similar to fdisk on linux?
[15:23:02] <Berny> jmcp: got your point... so in fact there seems to be little hope in my case other then either replacing biosdev with a script or give the disks a different layout?
[15:23:02] <jmcp> nope
[15:23:06] <jmcp> sftnope
[15:23:09] <jmcp> StFS: nope
[15:23:13] <Gekkko[PDA]> lol
[15:23:20] <jmcp> Berny: biosdev does the best it can
[15:23:21] <Gekkko[PDA]> jmcp: wtf not?
[15:23:34] <jmcp> Berny: I'd be using a hacked up script
[15:23:40] <jmcp> in fact, I *do* use a hacked-up script
[15:23:43] <jmcp> Gekkko[PDA]: wtfnot what?
[15:24:00] <Gekkko[PDA]> why not nice fdisk
[15:24:10] <Berny> jmcp: biosdev looks clean in what it does... so script would be the way to go if support can't fix it in the bios...
[15:24:13] <jmcp> Gekkko[PDA]: because the command line version we've got has always been sufficient
[15:24:18] <jmcp> Berny: yeah
[15:24:19] <Gekkko[PDA]> why cruel, nasty, user rapist fdisk.
[15:24:20] *** victori_ has quit IRC
[15:24:35] <flyingparchment> you only need one command in fdisk anyway, 'create solaris partition on entire disk'
[15:24:42] <Berny> Gekkko[PDA]: run "nice 19 fdisk" makes it much nicer :-P
[15:24:46] <cmihai> lol
[15:25:58] *** fbo has joined #opensolaris
[15:26:49] <trochej> :)
[15:26:58] <trochej> flyingparchment: True
[15:26:59] *** huw-l has quit IRC
[15:27:40] <Berny> hmm, does grub put some stuff in the first block on the disks? or could i just leave one disk intact and zero out the first block on the second disk?
[15:28:00] *** Anders has joined #opensolaris
[15:28:50] <jmcp> Berny: preserve the first block
[15:30:47] <Berny> hmm... idea was to zero out that block, let biosdev do its work and the lus stuff can put the bootloader there...
[15:32:48] *** karrotx has quit IRC
[15:32:52] *** karrotx has joined #opensolaris
[15:33:02] * jmcp runs away screaming from the lus nightmare
[15:33:40] * Berny giggles
[15:34:02] <Berny> so i found a way to scare you then :-)
[15:34:18] *** pfa3rh has joined #opensolaris
[15:35:38] *** ocr has joined #opensolaris
[15:35:56] *** laca has joined #opensolaris
[15:36:00] <Berny> btw i wonder why biosdev -d tells me match_edd bdf 0 5 0 for both disks... if i understand things right it should be 0 5 1 for my second disk...
[15:36:18] *** RaD|Tz has joined #opensolaris
[15:39:44] *** tsoome has quit IRC
[15:40:55] <trochej> Berny: I have to tell you, you giggle like you mean it :)
[15:41:06] * Berny does
[15:41:29] <Berny> why would i giggle if i didn't mean it? :-)
[15:41:57] *** Gekkko[PDA] has quit IRC
[15:43:06] *** _judas_ has joined #opensolaris
[15:43:59] <dunc_> as a decoy?
[15:44:02] *** hile_ has quit IRC
[15:44:11] * _judas_ stabs dunc_ in the face
[15:44:42] * dunc_ blinks
[15:45:00] <dunc_> bit sore that
[15:45:23] <Berny> na... decoys are for wimps... real sysads say it straight ;-)
[15:49:29] *** jafari has joined #opensolaris
[15:50:18] <trochej> http://blogs.eweek.com/careers/content001/recruiting/hey_genius_wooing_young_hip_microserfs.html?kc=EWKNLCSM091107STR3
[15:50:21] <trochej> ROTFL
[15:50:46] * _judas_ bites trochej's prostate
[15:51:14] <Berny> ouch
[15:51:18] <trochej> _judas_: I don't have one. I donned it for homeless children.
[15:51:30] *** FelipeBare has quit IRC
[15:51:55] <dunc_> well somebody fell out of the wrong side of their bed this morning
[15:52:00] *** FelipeBare has joined #opensolaris
[15:52:24] * dunc_ trips _judas_ down some really hard metal steps
[15:52:35] <_judas_> oh
[15:52:39] <_judas_> :)
[15:52:42] <dunc_> :)
[15:54:24] <trochej> _judas_: What you bitten was my metal ball transplant. Say goodbye to your teeth
[15:55:23] <Berny> .oO(are these magnetic?)
[15:55:41] <_judas_> :F
[15:56:19] <trochej> Berny: Of course they are, other way I wouldn't be able to get my wife to bed in any way
[15:56:26] <trochej> I feed her with magness
[15:56:31] <_judas_> http://www.geocities.com/ujb81/images/beforeafter.jpg
[15:57:18] <trochej> Uhm
[15:57:26] <_judas_> i cleaned it :)
[15:57:27] <trochej> _judas_: How much was for surgery?
[15:57:38] <trochej> _judas_: We wouldn't be able to see it, ourselves
[15:57:39] <_judas_> not much :)
[15:59:05] <trochej> But I make a nice _kling_ noice, when I walk
[15:59:20] *** ocr has quit IRC
[15:59:25] <trochej> The only problem is airport security gate
[15:59:30] <_judas_> hehe
[15:59:57] <_judas_> ok, night guys
[15:59:58] *** _judas_ has left #opensolaris
[15:59:59] *** ocr has joined #opensolaris
[16:00:13] <trochej> Night, so long, see ya, sayonara, dowidzenia, dobranoc, bye bye
[16:01:30] <trochej> ok
[16:01:32] <trochej> 4 PM here
[16:01:35] <trochej> Time to go homne
[16:01:36] <trochej> [d]
[16:01:59] <Berny> 2hours till train home...
[16:03:14] *** hile_ has joined #opensolaris
[16:03:42] <jbk> the day's just starting here :(
[16:03:58] <Berny> my day started 10hrs ago...
[16:07:56] *** cmihai has quit IRC
[16:09:13] <logic_> hi, i am having a problem with dladm, i am trying to create a aggregate, but it fails :(
[16:09:18] <logic_> i keep getting : dladm: create operation failed: invalid link
[16:09:35] <logic_> the command i tried is : % dladm create-aggr -d vfe0 1
[16:09:47] <Berny> logic_: are your nics using gldv3 drivers?
[16:10:02] <logic_> yes, i recompiled with gldv3 support
[16:10:15] <Berny> hu whats vfe?
[16:10:22] <logic_> vfe0            type: non-vlan  mtu: 1500       device: vfe0
[16:10:36] <Tempt> been busy around here tonight
[16:10:39] <logic_> the non-vlan is correct right?
[16:10:51] <Berny> cheerio Tempt
[16:11:09] <Tempt> Heya
[16:11:16] <Tempt> Melbourne OpenSolaris thingy was pretty good.
[16:11:26] <logic_> Berny: vfe is from: vfe, former rh, driver for VIA Rhine family fast ethernet chipset
[16:11:29] <Tpenta> cool
[16:11:34] <Tpenta> I have to organise another sydney one :)
[16:11:46] <Tempt> Apparently I'm giving the preso on ZFS at the next one
[16:12:13] <Berny> logic_: sorry never worked with that driver before...
[16:12:24] <logic_> Berny: ok thanx
[16:12:26] <Berny> ho Tempt becoming a star... :-)
[16:12:31] <Tempt> Ha.
[16:12:36] <Berny> btw how many pints on the new jobs? ;-)
[16:12:59] <jbk> the local opensolaris thing here kinda fizzled
[16:13:08] <Tempt> More like said "giving presentations is easy, I'll do one if nobody else is motivated" and got assigned ZFS
[16:13:11] <jbk> the guy organizing it never told anyone where it'd be :)
[16:13:30] <Tempt> Berny: There will be a mandatory heavy drinking session on Friday.
[16:14:06] <Tempt> Boyd's LU presentation was pretty good tonight, there wasn't even much heckling opportunity.
[16:14:22] <Berny> Tempt: just that it's to far away...
[16:14:37] *** Vorbis_w has joined #opensolaris
[16:14:43] <Tempt> Btw, anyone who wants to run with my idea of using Niagara-2's crypto hardware to offload ZFS/raidz calculations is welcome to it.
[16:14:51] <Berny> maybe is should go and get a loan for the flight ;-)
[16:15:14] <Tempt> Berny: Could put you up on the couch for the duration, saves accomodation costs?
[16:15:53] <Berny> hehe cheers...
[16:16:09] <logic_> Berny: if you do a creation of the aggr, do you first have to unplumb the cards?
[16:16:20] <logic_> Berny: they must be inactive?
[16:16:28] <Berny> aussie is supposed to be warm... maybe i go and stay with the metermaids on the beach ;-)
[16:16:48] <Berny> logic_: i vaguely remember so...
[16:17:08] <Berny> or was it plumbed but no ip assigned... but i think unplumbed
[16:17:19] <Tempt> logic_: Yes.
[16:17:22] <Tempt> logic_: Unplumb 'em
[16:17:38] <Tempt> Berny: metermaids etc is a QLD thing.
[16:17:56] <logic_> ok thanx, then i will try it when i get home, not want to risk to loose the connection ;)
[16:18:05] <Berny> order some to your place Tempt ;-)
[16:18:22] *** dunc_ has quit IRC
[16:18:28] <Berny> logic_: you loose connection until your aggr is running :-)
[16:22:12] <flyingparchment> item on quote request: Thank you for choosing Dell.  Qty: 1
[16:22:44] *** sfire||mouse has joined #opensolaris
[16:22:51] <Tempt> flyingparchment: "I'll have a large server of thanks, thanks."
[16:23:50] <flyingparchment> wtf, this EMC array has 5 146GB FC-AL disks for the array OS?!
[16:23:51] <nachox> ?
[16:24:42] <Berny> isn't that a little overkill?
[16:24:52] <jbk> perhaps they switched to windows :)
[16:24:56] <jbk> for the OS
[16:25:12] *** sfire||mouse has quit IRC
[16:25:32] *** sfire||mouse has joined #opensolaris
[16:27:35] <Berny> woohoo... a week of wrestling over... case escalated...
[16:29:38] <flyingparchment> Berny: it seemed a little odd to me..
[16:30:21] <Tempt> Five spindles is probably the size of a parity group.
[16:30:24] <Tempt> So it'd make sense.
[16:30:27] *** sioraiocht has left #opensolaris
[16:30:46] <flyingparchment> Tempt: i'm used to firmware being stored in flash, not fast, expensive disks :)
[16:31:01] *** sioraiocht has joined #opensolaris
[16:31:04] <Berny> but 146GB? wouldn't 32mb flash be enough?
[16:31:07] <WickedWicky> why doesnt Sol10u4 recognize the internal smart array adapter on a DL380 :s
[16:31:50] <WickedWicky> hey tempt, how was the osol meeting?
[16:31:51] <Berny> what controller is in there?
[16:31:58] <WickedWicky> lemme reboot
[16:32:00] <Berny> hey wicky btw
[16:32:05] <WickedWicky> I'm on the box via ILO
[16:32:07] <flyingparchment> Berny: compaq smart array i think
[16:32:12] <WickedWicky> yea
[16:32:20] <flyingparchment> i thought solaris had a driver for it
[16:32:37] *** karrotx has quit IRC
[16:33:08] <Berny> osol/sol didn'tz like the adaptec raid controller in my box... which happened to be rebranded by dell with a nifty firmware missing half the options (including "work with solaris")
[16:33:21] <WickedWicky> Smart Array 6i
[16:33:22] <flyingparchment> dell rebranded controllers are shitty
[16:33:27] <elektronkind> you might be interested in the mfi driver
[16:33:28] <Berny> ack
[16:33:35] <jbk> s/rebranded controllers//
[16:33:39] <Berny> replaced it with an lsi hba
[16:33:48] <WickedWicky> no, I am interested in out of the box controllers that are on all the DL380s we have to work
[16:33:53] <WickedWicky> :P
[16:34:05] *** hohum has joined #OpenSolaris
[16:34:28] <WickedWicky> anyway, I cant install it now I am afraid
[16:34:39] <WickedWicky> unless there is some driver available that I can put on a floppy
[16:34:45] <WickedWicky> lemme google
[16:34:48] <WickedWicky> google always is my friend
[16:35:02] <WickedWicky> oh btw, about my question yesterday , the MSA1500 + Solaris combo
[16:35:05] <WickedWicky> technicaly it works
[16:35:09] <WickedWicky> HP just doesnt give support on it
[16:35:30] <quasi> WickedWicky: hp has plenty of sol 8 drivers for the Smart Array
[16:35:33] <Berny> if it works don't tell hp :-)
[16:35:50] <flyingparchment> http://h18023.www1.hp.com/support/files/server/us/locate/38_1122.html
[16:36:10] <flyingparchment> i guess the S8 DPU won't work, but the driver might
[16:36:23] <WickedWicky> let's try that
[16:36:24] <WickedWicky> thank you
[16:36:56] <WickedWicky> ah look
[16:37:07] <WickedWicky> http://h18000.www1.hp.com/support/files/server/us/locOsCat/1123.html
[16:37:12] *** forelle has joined #opensolaris
[16:37:14] <WickedWicky> HP has a solaris 10 x86 section
[16:37:27] <quasi> yeah
[16:37:33] <flyingparchment> oh, must be well hidden, it wasn't in the list i was looking at
[16:37:33] <quasi> http://h18004.www1.hp.com/support/files/server/us/download/24869.html
[16:37:57] <quasi> gah, paste
[16:38:58] <Tempt> EMC man
[16:39:05] <Tempt> the price you're paying, you expect overkill
[16:39:05] <WickedWicky> so, I never did this. Do I download the the Driver Update diskette and that's all I need?
[16:39:10] <flyingparchment> yes
[16:39:13] <WickedWicky> awesome
[16:39:16] <flyingparchment> you get prompted to insert it during the install
[16:39:19] <Tempt> Oh, and Melb osol tonight
[16:39:24] <Tempt> turn-out was around 50
[16:39:29] <Tempt> 100% male attendance
[16:40:00] *** TBCOOL has joined #opensolaris
[16:40:35] <WickedWicky> brb, quic smolke
[16:40:56] <jbk> Tempt: and that was a surprise? :)
[16:41:10] <Tempt> I thought there might be *one*
[16:41:34] <Berny> just drag some random girls along
[16:42:04] <Tempt> I think Australia's UNIX community is 100% male
[16:42:27] <Tempt> Kinda depressing in a way.
[16:42:32] <flyingparchment> i know for a fact that australia is composed only of men who like bbqs and UNIX, and women who spend all day on the beach.
[16:42:41] <flyingparchment> (i know this because i saw australian soaps on the tv.)
[16:42:54] <Tempt> Which soap has UNIX people in?
[16:43:01] <Tempt> (Enquiring minds demand to know!)
[16:43:03] <flyingparchment> i implied the unix people from the bbqing
[16:43:16] <flyingparchment> you know unix people have long hair and like that kind of thing
[16:43:41] <richlowe> flyingparchment: you're forgetting the forthright older women with the heavily dyed hair.
[16:43:55] <Berny> hehe
[16:43:58] <richlowe> if you're going to base this on australian soaps, you have to get the whole thing down.
[16:44:11] <Tempt> I've had a few women as DBAs, and a couple of developers
[16:44:15] <Tempt> but no UNIX at all.
[16:44:18] <Berny> Tempt: btw the female unix community here in germany i wouldn't invite for bbq anyway...
[16:45:59] <Tempt> Ouch.
[16:46:10] <renihs> heh :)
[16:46:20] <quasi> hmm, sun is supposed to be buying Lustre
[16:46:26] <TBCOOL> hrm, that was fairly interesting. wish I had known about 6598201 before I was trying to upgrade to b72..
[16:47:32] <Tempt> quasi: To kill it?
[16:48:14] <Tempt> or port to solaris?
[16:48:25] <quasi> Tempt: more likely to kill netapp :)
[16:48:26] <Tempt> Either way, probably ends up being related to something fishy.
[16:48:46] <quasi> yeah, el reg seems to think so as well - http://www.theregister.co.uk/2007/09/12/sun_buys_lustre/
[16:49:03] <Tempt> Hmm
[16:49:07] <Tempt> El Reg is usually on the ball
[16:49:17] <flyingparchment> sun made a press release about it
[16:49:20] <flyingparchment> so i'd say it's official :)
[16:49:31] <Tempt> Hopefully they re-release under CDDL
[16:49:32] <Tempt> Solaris only
[16:49:37] <Tempt> (Okay, Solaris/OpenSolaris)
[16:49:42] <Tempt> and terminate any linux support.
[16:49:49] <Tempt> And we can laugh *again* at the penguins
[16:49:58] *** rasputnik has joined #opensolaris
[16:50:33] <flyingparchment> hey, i have to work in a linux shop, be nice :(
[16:50:38] <rasputnik> anyone feeling helpful? I've got a corrupted zpool here, not sure if it's recoverable
[16:50:53] <quasi> Tempt: http://wiki.lustre.org/index.php?title=Lustre_OSS/MDS_with_ZFS_DMU - there is a bit of solaris there already
[16:51:32] <rasputnik> I need to figure out what the output of 'zpool status -xv' is trying to tell me
[16:52:24] <Tempt> rasputnik: Pastebin it
[16:52:53] <quasi> Tempt: and the official bit - http://www.sun.com/aboutsun/pr/2007-09/sunflash.20070912.2.xml
[16:53:27] <rasputnik> here it is : http://opensolaris.pastebin.ca/693964
[16:53:54] <rasputnik> the 'range' bit looks worryingly large :(
[16:57:46] <Tempt> why does everyone call their zpool tank?
[16:57:56] <flyingparchment> Tempt: it's the example in the zfs docs
[16:58:01] <Berny> yep
[16:58:03] <Tempt> ffs
[16:58:03] <Berny> copy&paste
[16:58:08] <Tempt> ORIGINALITY PEOPLE
[16:58:14] <Tempt> and stop *CARGO CULTING YOUR CONFIGS*
[16:58:17] <flyingparchment> i called mine 'ift'
[16:58:20] <flyingparchment> because it's on an infortrend array
[16:58:21] <Berny> i call mine data :-P
[16:58:22] <flyingparchment> :)
[16:58:26] <LeftWing> I called mine modicum.
[16:58:41] <Tempt> err
[16:58:45] <Tempt> sata750 and "global"
[16:58:54] <rasputnik> Tempt: I was gonna rename it, but then I have to rewrite like 4 pages of docs :)
[16:59:18] <Berny> lol
[16:59:23] * rasputnik runs 'find /zones/foobar -inum 8572'
[16:59:28] <Berny> put <yournamehere> in the doc
[16:59:49] <flyingparchment> maybe if i get to deploy zfs here, i'll call the pool 'linsux'
[17:00:03] <richlowe> call it 'ufs'
[17:00:04] <rasputnik> hah, it's /usr/bin/bash ! I got a bazillion of those :)
[17:00:05] <LeftWing> I'll be calling my next zpool "netapp" I suspect.
[17:00:10] <jbk> antitux :)
[17:00:11] <richlowe> if you need two, call the other s5fs
[17:00:12] <Tempt> only because global matches the previous mountpoint
[17:01:48] <rasputnik> Tempt: cheers, I'm a happy man now. Do i just 'zpool clear' the error?
[17:01:59] <Tempt> zpool clear should be your friend.
[17:02:28] <Tempt> I'd probably give it a zpool scrub afterwards for good luck
[17:02:29] <Tempt> anyway, 1am
[17:02:33] <Tempt> whisky then bedtime
[17:02:36] <Tempt> See you all tomorrow.
[17:02:49] <rasputnik> cheers again
[17:03:03] <Berny> nighty Tempt
[17:04:01] <flyingparchment> 'zpool status' is annoying slow under moderate-high load
[17:04:45] *** sfire||mouse_ has joined #opensolaris
[17:05:50] *** mikefut has quit IRC
[17:06:13] *** mikefut has joined #opensolaris
[17:06:41] *** yage has joined #opensolaris
[17:07:37] <flyingparchment> i have to say, i'm leery of zfs since i deployed it in production and found an unfixed panic within 12 hours
[17:08:38] *** leal has quit IRC
[17:08:41] *** jwk404 has joined #opensolaris
[17:08:50] <Berny> how did you do that?
[17:08:51] <yage> Hi, i'm new to solaris, where can i find a good command reference? For example "ifconfig -a" doesn't return all network interfaces (enabled and disabled) but only enabled ones, how do i enable the other interfaces?
[17:09:20] <flyingparchment> Berny: created an fs and put mysql on it.  a few hours later it paniced :)
[17:09:29] <Berny> yage: ifconfig plumb -a
[17:09:32] <jbk> yage: dladm show-link
[17:09:57] <Berny> yuck mysql
[17:10:02] *** Vanuatoo_ has joined #opensolaris
[17:10:30] *** Vanuatoo has quit IRC
[17:10:32] <jbk> perhaps it was revolted at the thought of having mysql using it :)
[17:10:33] <yage> dladm, great :) So, is there any reference splittend in arguments, for example "network commands... disk commands..." next i read the man page, but i need first to know their names...
[17:11:06] <Berny> yage: did you check the "beginners" section on opensolaris.org?
[17:11:07] <yage> (i come from linux and bsd)
[17:11:31] *** sfire||mouse has quit IRC
[17:11:42] <yage> Berny: yes, but there are so many links there...
[17:11:52] <Berny> the documentation section has a new to solaris part
[17:12:34] <yage> great :)
[17:13:31] * Berny heads off to train station
[17:17:13] *** cypromis has joined #opensolaris
[17:18:06] <rasputnik> yage: man ifconfig ? The manpages on solaris are much much better than the linux ones
[17:18:42] *** JoergB has joined #opensolaris
[17:18:58] *** johnlev has joined #opensolaris
[17:20:50] *** ballchalk has joined #opensolaris
[17:21:30] <yage> rasputnik: the problem is that only with dladm i can display the interface i need to bring up and the ifconfig manpage has no reference to it
[17:22:23] *** ballchalk has quit IRC
[17:22:30] <rasputnik> no fair enough. manpage would only help if you knew what 'plumb' meant
[17:26:58] <yage> ahhh, ok, understood now
[17:27:07] <yage> was "ifconfig -a plumb" not "ifconfig plumb -a"
[17:27:15] <yage> got it
[17:28:47] *** _dennis has quit IRC
[17:29:27] <rasputnik> yay manpages \o/ !
[17:30:13] *** Vanuatoo_ has quit IRC
[17:30:45] *** sfire||mouse_ has quit IRC
[17:33:59] *** Vanuatoo_ has joined #opensolaris
[17:40:25] *** Jondice_ver_2 has joined #opensolaris
[17:47:19] *** Jondice has quit IRC
[17:49:37] *** slowhog has joined #opensolaris
[17:53:28] *** forelle has quit IRC
[17:55:55] *** yarihm has joined #OpenSolaris
[17:57:13] *** rasputnik has quit IRC
[18:03:12] *** yage has quit IRC
[18:05:24] *** estibi has joined #opensolaris
[18:05:57] *** Erwann has quit IRC
[18:06:15] <flyingparchment> hmm, free veritas storage foundation
[18:06:25] <flyingparchment> wonder how that compares to QFS for OLTP workload
[18:07:20] *** yippi has joined #opensolaris
[18:08:28] *** hile_ has quit IRC
[18:09:30] <nachox> i wonder how many people is actually  using QFS
[18:10:01] <flyingparchment> i am!
[18:10:11] <mustang> two.
[18:10:12] <mustang> me and him.
[18:10:13] <mustang> two.
[18:10:17] <mustang> :)
[18:10:40] <flyingparchment> only QFS though, not the SAM part
[18:10:51] <mustang> ok. I've got some of that too.
[18:11:51] <flyingparchment> filesystems seem to hate me, as well as panicing zfs, i made QFS give a repeating error in syslog that wouldn't go away until i rebooted
[18:11:59] <flyingparchment> wonder what i'll manage to break in vxfs
[18:12:05] <nachox> what's the use of QFS without sam? just a shared filesystem?
[18:12:21] <flyingparchment> nachox: it's less lame than UFS for larger filesystems (>1TB)
[18:12:22] <jbk> flyingparchment: just fill it up 100% :)
[18:12:25] <flyingparchment> and shared if you need that
[18:12:49] *** vmlemon has joined #opensolaris
[18:12:53] *** estibi has quit IRC
[18:12:58] <flyingparchment> jbk: i think that's what broke qfs :)
[18:13:15] <jbk> there was a bug a few years back with vxfs that cause it to catastrophically corrupt the filesystem when it hit 100%
[18:13:24] <jbk> where a mkfs + restore was your only option
[18:13:38] <jbk> (so much for 'time-proven' solutions :P)
[18:13:44] <jbk> and
[18:13:48] <jbk> for extra fun
[18:13:55] <jbk> if using clustered vxfs (multi-writer)
[18:14:29] <jbk> try gzipping or unzipping a large file while simultaneously accessing other files in the same directory (even from the same server)
[18:14:52] *** Fish has joined #opensolaris
[18:14:53] <jbk> any processes that do will freeze until the gzip/unzip is complete
[18:15:01] <jbk> and cannot be killed
[18:15:09] <nachox> jbk: that is obviously a feature :P
[18:15:14] *** hile_ has joined #opensolaris
[18:15:17] <jbk> (they get stuck waiting on a cv in the kernel)
[18:15:24] <Fish> hello
[18:17:14] <nachox> cv?
[18:17:28] <jbk> condition variable
[18:18:03] <jbk> thankfully veritas was already engaged w/ a VOS engineer
[18:18:27] <elektronkind> cv_wait strikes again
[18:18:28] <jbk> trying to explain that one to a level 1/2 tech would not have been fun :)
[18:20:36] <mustang> ick
[18:21:34] <nachox> "hello, the filesystem you guys made is stuck in a condition variable" "err, what?" "you know, a condition variable, those cool syncronization stuff" "..." "is your manager round?" "*loud bang*"
[18:22:45] <flyingparchment> so i finally got a correct quote for solaris support
[18:22:54] <flyingparchment> now the guy here who can actually buy it is on holiday for 2 weeks
[18:22:56] 
[18:23:05] 
[18:23:30] <nachox> flyingparchment: what are you buying?
[18:23:45] <flyingparchment> nachox: solaris standard support
[18:24:03] <elektronkind> flyingparchment: so wikipedia can't buy anything for two weeks until this guy comes back?
[18:24:23] <flyingparchment> elektronkind: this is Wikimedia Deutschland e.V., separate people
[18:24:46] <axisys> I get `this version of Solaris 10 is not compatible' when booting using sol 10 u3 x86 cd (provided by SUN) and point to network image of sol 10 u4 x86.. is this normal?
[18:26:27] <flyingparchment> (the e.V. is the only part of wikimedia that runs solaris ;-)
[18:28:03] *** charlesnw has joined #opensolaris
[18:28:13] <axisys> i create a lofi device of the iso and then monted it locally and then used star to create net image like this ->  star -copy -xdot -p -C /mnt/image . /net/image
[18:28:35] <axisys> could that be why the image showed not compatible?
[18:28:46] <flyingparchment> why are you installing U4 using a U3 CD?
[18:29:21] *** ichigo has quit IRC
[18:31:21] <axisys> flyingparchment: because I dont have U4 cd yet.. waiting for sun to ship them
[18:31:57] <elektronkind> you can download the isos ...
[18:32:16] <axisys> elektronkind: that is what i did.. but systems dont like burned dvds
[18:33:03] <axisys> elektronkind: i could not jump the x86 box becaseu there is another dhcp server in same network conflicting with mine
[18:33:38] <axisys> elektronkind: too many issues w/ jumping from net.. that is why picked a sun provided dvd
[18:34:15] <axisys> installing from dvd gives u an option to point to net image and that is what I was trying
[18:35:28] *** Cyrille has quit IRC
[18:36:47] *** yarihm has quit IRC
[18:36:57] *** Pietro_S has joined #opensolaris
[18:37:47] *** Kush- has left #opensolaris
[18:39:10] *** johnlev has quit IRC
[18:42:32] *** Kush- has joined #opensolaris
[18:44:32] <FireflyST> what the hell
[18:44:33] *** trs81 has quit IRC
[18:44:52] <FireflyST> accuweather.com thinks I'm using WAP from my OpenSolaris machine on Firefox
[18:45:09] <g4lt-sb100> you're not? ;P
[18:49:24] *** FelipeBare has quit IRC
[18:51:46] *** peteh has quit IRC
[18:56:27] *** derchris has quit IRC
[18:56:34] *** derchris has joined #opensolaris
[19:01:20] *** mikefut has quit IRC
[19:01:43] *** tsoome has joined #opensolaris
[19:02:34] *** trs81 has joined #opensolaris
[19:03:49] <CIA-24> rb144127: 6591930 domain ETM omits check for dup xid on ALERT msgs, 6591931 domain ETM falsely reports etm_xport_open_fail stat, 6594506 domain ETM and LDC likely deadlock when xmit queue full
[19:04:14] *** hohum has quit IRC
[19:04:43] *** timsf has quit IRC
[19:04:45] *** ichigo has joined #opensolaris
[19:09:50] <madhatter> Does Live Upgrade work with two slices on the same disk, too?
[19:11:07] *** yippi has quit IRC
[19:12:59] <madhatter> And is there another tool but fdisk to reorganize disks? This, again, seems to work very different from dos, linux or freebsds fdisk :(
[19:14:14] *** eboutilier has quit IRC
[19:14:40] <nachox> format
[19:15:18] <holcomb> yes to the liveupgrade thing
[19:15:50] <madhatter> holcomb: Do the slices need the be identical in size or is there anything else I should take care of?
[19:15:55] <madhatter> nachox: Thanks
[19:16:17] <holcomb> i don't think so
[19:16:23] <holcomb> never tried it though :)
[19:16:37] <holcomb> all it does is cpio over the files, so it could probably even be smaller
[19:16:46] <madhatter> holcomb: Oh, I see :) But you are sure about the slices
[19:17:04] <holcomb> yes.  i do that on a regular basis
[19:17:08] *** Triskelios has joined #opensolaris
[19:17:34] <madhatter> holcomb: What? So you did not try it with different sizes?
[19:17:44] <madhatter> holcomb: Seems I misunderstood you before
[19:17:59] <holcomb> i have in the past liveupgraded from one slice to another on the same disk
[19:18:02] <holcomb> they were the same size
[19:18:15] <madhatter> Ah okay
[19:18:33] <madhatter> To have same size makes more sense I guess
[19:18:48] <madhatter> As far as I read the howto you stick with the patched copy, right?
[19:19:06] *** Snake007uk has quit IRC
[19:22:40] <gdamore> for anyone that want's a simple kernel task, it would be nice to eradicate all instances if mi_strtol() from the kernel... and replace them with ddi_strtol().
[19:23:09] *** Mdx4 has joined #opensolaris
[19:23:30] *** logic_ has quit IRC
[19:24:36] <jbk> now sun is oem'ing windows
[19:24:50] <jbk> anyone see any guys on horseback riding around the bay area? :)
[19:25:39] *** cmihai has joined #OpenSolaris
[19:27:37] *** perlmonk has joined #opensolaris
[19:27:59] *** bubbva has joined #opensolaris
[19:31:18] *** estibi has joined #opensolaris
[19:35:04] *** Daniel_J has joined #opensolaris
[19:35:59] <jbk> or really, really dark clouds overhead? :)
[19:38:00] *** victori_ has joined #opensolaris
[19:38:15] <victori_> oh I just noticed there is portage for solaris, is it any good?
[19:38:25] <cmihai> no
[19:38:40] <cmihai> Portage is horrible enough as it is...
[19:38:51] <victori_> ok blastwave it is
[19:38:53] *** sahafeez has joined #opensolaris
[19:38:58] <victori_> cmihai: I disagree
[19:39:00] <victori_> but okay
[19:39:28] *** libkeiser has quit IRC
[19:40:06] *** yippi has joined #opensolaris
[19:40:09] <Triskelios> victori_: if you want to build stuff from source, spec-files-extra has packages
[19:40:35] <victori_> well I would want to avoid building if possible but it does add flexibility
[19:40:50] *** sahafeez has quit IRC
[19:40:59] *** sahafeez has joined #opensolaris
[19:41:21] *** cmihai has quit IRC
[19:41:34] *** cmihai has joined #OpenSolaris
[19:42:09] *** johnlev has joined #opensolaris
[19:44:13] <victori_> ok stupid question since solaris uses ufs, do I need to enable some type of softupdates system?
[19:44:25] <cmihai> No BSD boy
[19:44:29] <victori_> ;-)
[19:44:34] <cmihai> Solaris UFS uses logging
[19:44:38] <cmihai> It's like journaling.
[19:44:42] <cmihai> Unlike SoftUpdates
[19:45:04] <cmihai> Besides, their UFS is more of a FFS :P
[19:45:47] <victori_> thanks for answering all my questions, I appreciate it
[19:46:04] *** dunc has joined #opensolaris
[19:46:37] * Triskelios really needs to work out an arrangement for building SFE binary packages
[19:50:07] *** cydork has quit IRC
[19:54:35] <axisys> what is boot -a equivalent on sol x86?
[19:56:56] *** LoBoX has joined #opensolaris
[19:57:12] <jbk> adding -a on the end of the kernel line in grub
[19:57:52] *** bondolo has quit IRC
[20:00:11] *** bondolo has joined #opensolaris
[20:00:19] <CIA-24> rs200217: 6498936 TCP RST ignores IP_NEXTHOP for unexpected TCP data packets
[20:01:33] <victori_> pity no postgres part of the coolstack
[20:01:51] <wesolows> why should it be?  It's in OpenSolaris proper.
[20:01:51] *** comay has joined #opensolaris
[20:02:17] <wesolows> The CoolStack people are taking the absolute wrong approach, and the SFW C-Team told them so last year.
[20:02:35] <axisys> jbk: thnx
[20:02:38] <victori_> so what would you advise?
[20:02:40] <richlowe> but, but, SFW is such a boring name!
[20:02:43] <victori_> someone mentioned to just use the coolstack
[20:02:50] <cmihai> Yes
[20:02:52] <wesolows> They should be working on providing the right set of compilation options for best performance in the regular OpenSolaris consolidations, rather than building their own duplicate versions.
[20:02:55] <cmihai> me
[20:03:10] <cmihai> Postgres is part of Solaris though
[20:03:18] <cmihai> But not integrated in the php && co of CoolStack
[20:03:42] <wesolows> Nevertheless, if you just need Postgres and don't care about a hypothetical 0.0002034% performance boost on T2000, just use your favourite OpenSolaris distribution.
[20:04:02] <wesolows> cmihai: Which is a perfect example of why their approach is wrong.
[20:04:20] <cmihai> No idea why they don't put that stuff in Solaris
[20:04:25] <richlowe> also, if you do need the minor performance boost, dodge it anyway.  Stupidity shouldn't be rewarded.
[20:04:28] <cmihai> I have to spend hours writing SMF manifests
[20:04:41] <victori_> I am guessing this coolstack is installed in a zone
[20:04:51] <victori_> would be nice to just to see how much of a difference it makes if any
[20:05:04] <wesolows> Zones make no difference for performance.
[20:05:16] <victori_> not zones but the coolstack
[20:05:23] <victori_> it won't conflict with the sunw stuff
[20:05:29] <delewis> "no difference" meaning only '0-5% performance degradation for most workloads'
[20:05:32] *** RElling has quit IRC
[20:05:42] <cmihai> No, CoolStack installs in the global zone
[20:05:47] <wesolows> delewis: It should be extremely close to 0 for nearly if not everything.
[20:05:54] <wesolows> If not 0.
[20:05:57] <cmihai> It's just a good idea to use it in a zone. Like everything else :P
[20:06:03] <delewis> wesolows: Solaris Internal Ed/2 says otherwise.
[20:06:06] <victori_> cmihai: I concur
[20:06:28] <wesolows> delewis: Ask comay or dprice if you want a maximally authoritative answer.
[20:06:47] <brendang> delewis: which page?
[20:07:15] <delewis> I don't have a copy with me, but it's in the zones chapter towards the end, IIRC.
[20:07:24] <comay> for almost all workloads, we've seen virtually no degradation (< 1%s)
[20:07:33] * brendang looks
[20:07:48] <delewis> there's a chart that compares various workloads (Java, network-intensive, etc.) in the global zone and local zones.
[20:08:09] <cmihai> Besides, the new isolation stuff in u4 rocks.
[20:08:32] <cmihai> Memory, CPU shares, locked memory... and the older stuff like binding a zone to specific CPUs and such..
[20:08:34] *** migi has quit IRC
[20:08:40] <comay> i can imagine some pathoglogical cases (opening up 10,000 files in a short period of time) where you might see some degradation
[20:08:49] <delewis> cmihai: that stuff was there before Update 4.
[20:08:50] *** yippi has quit IRC
[20:08:57] <cmihai> delewis, not all of it.
[20:09:05] <delewis> most of that was present in 3/05.
[20:09:08] <brendang> still looking... certainly if you turn on FSS and rcapd you can get measurable overheads...
[20:09:35] <cmihai> Well, the tools got enhanced mostly.
[20:09:40] <wesolows> and IP instances may have some cost; I don't know enough about that to comment.  But none of that is Zones, per se.
[20:09:46] <comay> rcapd itself can cause some overhead but that isn't related to zones
[20:09:50] <cmihai> Like rcapadm on the fly stuff
[20:10:26] <comay> cmihai: yes, the tools did get enhanced but one of the more important changes is the accouting of shared memory segments is accurate now
[20:10:28] <cmihai> delewis, locked memory was there?
[20:10:41] <delewis> I did say 'most'
[20:10:42] <FireflyST> hmm
[20:10:45] *** yippi has joined #opensolaris
[20:10:53] <cmihai> comay, yeah, I think it overcounted them before
[20:10:54] <delewis> you could certainly create a CPU set and bind it to a zone, exclusively.
[20:11:10] <cmihai> delewis, I know, hence the "older stuff"
[20:11:11] <comay> cmihai: indeed.  it made things like rcapd useless with database workloads
[20:11:21] <cmihai> " and the older stuff like binding a zone to specific CPUs and such.."
[20:11:29] *** FelipeBare has joined #opensolaris
[20:11:45] <brendang> the page is 410
[20:12:31] <cmihai> This isolation stuff works quite well too. Too bad people aren't usually informed...
[20:12:43] <cmihai> And you see things like forkbombs in a container take down the whole machine
[20:13:43] * comay doesn't have his internals book (gave it to a customer!).  is the table/graph available online anywhere?
[20:13:45] <brendang> the table shows the performance loss between -0.3% and 4%, for different workloads (yes, "Networking" benchmarked faster).
[20:14:22] <cmihai> Were they using ZFS? :-)
[20:14:25] <brendang> the section unfortunately doesn't explain how the zones were setup -- if they were using FSS, rcapd, IPQoS, etc.
[20:14:27] <cmihai> Zones are so much faster on ZFS
[20:14:28] *** mikefut has joined #opensolaris
[20:15:07] <richlowe> that's a fairly nonsensical statement.
[20:15:25] *** hile_ has quit IRC
[20:15:29] <brendang> "The 4% degradatation in the time-sharing work-load is primarily due to the overhead associated with accessing commands and libraries through lofs."
[20:15:34] <brendang> sigh
[20:15:45] *** FireflyST has quit IRC
[20:16:46] <comay> brendang: right, that's the reason a benchmark that just does 100,000 open() calls will show a degradation
[20:17:04] <brendang> I should have spotted that in edit, sorry. it doesn't sound right - "commands and libraries" should be cached, by passing much of the FS overhead. unless the work-load was strange.
[20:17:14] <nachox> do you guys know if N2L was opensourced as part of opensolaris?
[20:17:25] <cmihai> grok it
[20:17:37] <comay> sun has an internal benchmark which simulates something like 100 users pounding on a multi-user system and that might show that degradation
[20:17:55] <cmihai> nachox, http://src.opensolaris.org/source/
[20:18:11] <brendang> comay: yep - I'd be amazed if someone can show a >1% performance loss for a real-world workload.
[20:18:12] <delewis> that'd be a nice benchmark to make available as part of the testing suite.
[20:18:17] <cmihai> comay, I just install Oracle and hammer the hell out of it :-)
[20:18:23] <nachox> excellent, it is :)
[20:18:26] <delewis> any possibility that could happen?
[20:18:27] <cmihai> Just grab some of the Quest stuff like foglight
[20:18:46] <cmihai> nachox, grab the onnv with mercurial
[20:19:23] <jbk> heh.. the last place at work felt it was necessary to recertify (a year+ long process) all the commercial apps they use under zones (even though all but 2 of them run as regular users)
[20:19:28] <jbk> it was frustrating
[20:19:33] <cmihai> hg clone ssh://anon at hg dot opensolaris.org/hg/onnv/onnv-gate
[20:19:56] <comay> one performance issue that was fixed in nevada & s10 concerned auditing but that doesn't seem common
[20:20:07] <nachox> cmihai: i just checked, it was opensourced
[20:20:32] <brendang> comay: actually - much of the zones chapter came from older sources (pre FCS); so not only may the work-loads be odd, but it may have been on very early bits.
[20:20:52] <cmihai> nachox, and.. I assume you also want it? Or.. just checking?>
[20:21:59] <delewis> brendang: interesting, so I guess I should stop warning users they may see a very slight performance degradation for some workloads, eh?
[20:22:42] <brendang> delewis: "slight performance degredation" sounds fine, so long as they understand that slight means near zero. :)
[20:22:54] <nachox> cmihai: N2L is part of the nis tools in solaris, a customer is interested in a nis->ldap migration and cares about source availability, i'm just giving him the options, i dont need it right now
[20:23:02] <delewis> brendang: I usually jus say "at most 5%"
[20:23:05] <delewis> just*
[20:23:18] <brendang> delewis: as with everything, they should benchmark it themselves. and we should help customers do that (filebench, etc).
[20:23:18] <delewis> and that was based on that chart in SI ED/2
[20:23:33] <delewis> brendang: right
[20:24:04] <brendang> delewis: "at most 5%" would probably be true - we just don't know how contrived the workload and setup was to get it to be so high.
[20:24:32] <wesolows> I don't think we should warn customers of performance degradations that are unlikely to be measurable.  It just discourages use of Zones.
[20:24:35] *** RaD|Tz has quit IRC
[20:25:15] <wesolows> It makes Zones sound like VMware or Xen or any of the other virtualisation players, who also use terms like "slight degradation" except in their case they really mean 10-95%
[20:25:39] * comay agrees with wesolows
[20:25:58] <madhatter> Is there some kind of autodetection of disks in 'format'? I added a 60G disk, should be c1d0, but format now wants to know the number of cylinders and stuff
[20:25:59] *** hile_ has joined #opensolaris
[20:26:01] <comay> it's important to point out that especially for i/o workloads
[20:26:15] <cmihai> It's like saying: don't put apache in a chroot, it's much slower
[20:26:36] <nachox> 95%? :)
[20:26:43] <cmihai> comay, IOzone!
[20:26:47] <wesolows> nachox: Have you ever actually used VMware?
[20:26:54] <wesolows> IME 95% was a good day
[20:26:56] <jbk> sun still needs to do a better job of explaining how zones are not vmware -- a lot of mgmt at companies don't get it
[20:27:22] <wesolows> jbk: Agreed.  And why Zones is a fundamentally better answer for nearly everyone.
[20:27:25] <nachox> wesolows: yes, i have, it was bad but not that much
[20:27:54] <brendang> wesolows: yep. I think the SI ED/2 perf chart came from the days when zones were new, and 0-4% sounded good. these days we'd be more careful, and if it was 4% we'd take a longer look at how sensible the work-load was.
[20:28:36] <wesolows> brendang: You're assuming they didn't just ask the Zones team "what's the worst degradation you could ever imagine accepting?" and then use that figure without actually testing anything.
[20:29:00] <comay> jbk: believe me, i know.  i've done many presentations trying to explain this.
[20:29:38] <jbk> me and a few other people were trying to do that internally at our old job as well
[20:29:44] <jbk> but finally gave up
[20:29:44] <wesolows> comay: Well, on the plus side of that "confusion", maybe we should spin off ZonesWare into the capital markets for a cool $10b. :-)
[20:29:59] <comay> LOL
[20:30:02] *** sommerfeld has joined #opensolaris
[20:30:10] *** ChanServ sets mode: +o sommerfeld
[20:30:11] <cmihai> comay, I've tried IOzone and zones on ZFS perfromed better then native Oracle on Veritas
[20:30:25] <cmihai> Or UFS (duh)
[20:30:40] <jbk> but i guess they get what they deserve
[20:31:16] <elektronkind> jeez
[20:32:37] <jbk> they'd rather fire a bunch of people than actually do anything smart that'd save more money than they 'saved' from the layoffs
[20:32:44] <elektronkind> internally with zones, the biggest change is that several subsystems now keep track of what zone context something is executing in. not a major thing at all.
[20:37:26] *** Jondice_ver_2 has quit IRC
[20:38:03] *** victori has quit IRC
[20:38:06] *** johnlev has quit IRC
[20:38:22] *** victori has joined #opensolaris
[20:38:34] <elektronkind> jbk "Saving money" in the corp world never, or rarely, means actually saving money. Perhaps over some longer term, yes, but it usually involves shifting costs from one ledger to another. You know, the whole "recurring vs. capital costs" shell game :/
[20:39:25] *** forelle has joined #opensolaris
[20:39:37] <cmihai> When ever I hear "Save money" or save now means spending money so you'll know not to buy anything the next time...
[20:40:22] <wesolows> Usually it means shifting costs into the future.
[20:40:43] <wesolows> Almost never does it mean shifting costs from the future into the present, layoffs being one of the few exceptions.
[20:40:56] <jbk> well their 'san' for example was designed to maximize expense :)
[20:41:07] <wesolows> All SANs are. :-)
[20:41:14] <jbk> even for a san
[20:41:17] <wesolows> SAN == boondoggle
[20:41:28] <jbk> all storage for every 'tier' was the most expensive in its class
[20:41:36] <jbk> and not necessairly the best performing
[20:41:44] <wesolows> See http://online.wsj.com/article/SB118955478194424452.html?mod=hpp_us_pageone for example
[20:41:53] <jbk> and they also then would only do glorified port aggregration
[20:42:14] <jbk> because EMC said it was 'bad' to mask luns from two different arrays to the same hba port
[20:42:27] <jbk> even if you have like 5% utilization
[20:43:02] <wesolows> you don't buy from EMC because you want good data centre equipemnt
[20:43:03] <cmihai> I love it.
[20:43:07] <jbk> so we're run into situations one some smaller boxes where we'd have to do entire migrations
[20:43:15] <jbk> because they needed another 50gb of san
[20:43:24] <cmihai> Your calling a big budget a problem? :-)
[20:43:26] <jbk> and the array they were using didn't have the space
[20:43:42] <jbk> and no room for extra hbas in the server
[20:43:47] *** Jondice has joined #opensolaris
[20:44:26] <jbk> cmihai: well it's more of paying firing thousands of people when for things like storage (and that's just one of _many_ examples) you're paying 10x/gb what you should really be paying
[20:44:43] <cmihai> ouch
[20:44:43] <wesolows> A big budget actually is a problem; it encourages unethical behaviour, and it encourages waste.
[20:44:55] <wesolows> The budget should be just big enough to do the job and no bigger.
[20:45:01] <cmihai> It encourages creativity!
[20:45:04] <cmihai> Domino servers :D
[20:45:22] <cmihai> Hell, I've seen 200 domino PCs..
[20:45:39] <cmihai> http://youtube.com/watch?v=5_tXcRYOYZ0 :D
[20:49:18] *** forelle has quit IRC
[20:49:35] *** forelle has joined #opensolaris
[20:49:38] <jbk> but like i said, they get what they deserve -- their customers are leaving them by the hundreds of thousands each quarter, they're far from a leader in their industry and fall farther behind the leaders every quarter
[20:53:36] *** FireflyST has joined #OpenSOlaris
[20:55:20] <dunc> dominos piz^H^Hcs
[20:56:16] <madhatter> Please help!
[20:56:44] <madhatter> I tried to create a partition label using 'format' but I don't get what that layout should tell me
[20:57:40] <madhatter> I have 9 partitions it seems. WIth unassigned spaces mostly. And even though I got 37g assigned to partition 0 I am not able to mount it because it's not the correct fstype
[20:57:55] <wesolows> that has nothing to do with format
[20:58:00] <wesolows> format just writes the VTOC
[20:58:10] <wesolows> you still have to create a filesystem if you want to mount it
[20:58:13] <madhatter> But I can only assign the fstype with fdisk. WHich shows me one partition from the type solaris
[20:58:18] <wesolows> see zpool(1M) or mkfs(1M)
[20:58:27] <wesolows> no, that's the partition type
[20:58:39] <wesolows> fstype refers to the type of filesystem that's actually been written to the allocated space in the slice
[20:58:42] <madhatter> wesolows: Ah, okay
[20:58:48] <wesolows> what are you actually trying to accomplish?
[20:58:59] <nachox> madhatter: you used freebsd?
[20:59:00] <madhatter> So i only have one partition?
[20:59:01] *** FireflyST has quit IRC
[20:59:11] <madhatter> nachox: yes
[20:59:22] <wesolows> If you want a filesystem on the whole disk, just do zpool create poolname cXtXdX
[20:59:26] <wesolows> zpool(1M)
[20:59:27] *** johnlev has joined #opensolaris
[20:59:39] <nachox> madhatter: well, freebsd also uses slices iirc
[20:59:41] *** jafari has quit IRC
[20:59:42] <wesolows> correct
[20:59:44] *** yippi has quit IRC
[21:00:04] <wesolows> but again, that's why I asked what you're trying to end up with, because slices are generally useless in a ZFS world
[21:00:17] <madhatter> nachox: True, but it's been a while since I added harddisks to my bsd box
[21:01:01] <madhatter> wesolows: I'd like to create a slice on a new disk, almost same size as my / for live upgrade
[21:01:09] <wesolows> aha
[21:01:24] <wesolows> well, then, presumably you're using UFS for your root filesystem?
[21:01:35] <richlowe> set the slice up with format, let lucreate the filesystem.
[21:01:37] <richlowe> since it's going to anyway.
[21:01:38] <wesolows> In that case, you've already created the slice with format and now you need to do newfs
[21:01:54] <wesolows> or listen to richlowe, who actually uses LU
[21:03:13] <madhatter> wesolows: Okay then, but I am not sure, if the partition layout still is correct. I modified the partition after creating it with fdisk
[21:03:23] <wesolows> Let's get our terminology clear
[21:03:46] <wesolows> Partitions are created with fdisk.  They are recognised by the BIOS.  You should have exactly one SOLARIS2 partition on any one physical disk.
[21:04:00] <wesolows> Slices are created by format within a SOLARIS2 partition.
[21:04:13] <madhatter> wesolows: Uh, switched to Solaris, so I probably should switch back again?
[21:04:24] <wesolows> Filesystems are normally created on either an entire disk (no partition or slice, as with ZFS), or on slices (UFS)
[21:04:36] <wesolows> Yes, you should use SOLARIS2 on all new installations.
[21:04:42] <PerterB> now if format(1M) would only refrain from calling slices partitions... ;)
[21:05:01] <wesolows> PerterB: Or just die and let ZFS take over the universe.
[21:05:10] <PerterB> that works for me too
[21:05:17] <madhatter> wesolows: Okay, I created a partition and then modified the slices somehow to make the 37gb not only appear on slice 2 but on slice 0 also
[21:05:32] <wesolows> madhatter: That's correct.  slice 2 is a special slice.
[21:05:52] <wesolows> madhatter: It contains the "whole disk" which in this case just means the whole partition.  You should not use it to store a filesystem.
[21:06:10] <wesolows> madhatter: Also, you should control the size of slices using format, not fdisk, unless you are sharing the disk with another operating system.
[21:06:20] <wesolows> Generally, the SOLARIS2 partition should consume then entire disk.
[21:06:21] <madhatter> wesolows: But I thought as I wanted the same layout as on the / I should get to slice 0
[21:06:41] <wesolows> If your root filesystem is on slice 2, you're as good as lost anyway.
[21:07:10] <wesolows> Don't we publish docs on docs.sun.com about how to use LU?
[21:07:15] <wesolows> Surely they talk about this.
[21:08:06] <madhatter> wesolows: No, I have my / on slice 0. But on the new created partition slice 0 was 0 in size
[21:09:59] *** FireflyST has joined #OpenSOlaris
[21:10:39] <jbk> this is a new disk
[21:10:45] <jbk> is anything else on it or going to be on it?
[21:11:01] <jbk> the sizes don't have to match, just needs to be big enough to hold the contents
[21:11:08] <madhatter> jbk: Nope
[21:11:29] <jbk> so just go in and make the slice as big as needed
[21:11:40] <jbk> it doesn't even need to be s0->s0 for LU
[21:11:46] <madhatter> jbk: I thought it would be good to have the exact same sizes. As far as I understood I switch at upgrade from one disk to the other
[21:13:06] <jbk> when you do a live upgrade, you will select which slices to upgrade (obviously / is one of them, but if /var or /opt is separate, thise might be included as well)
[21:13:50] <madhatter> What do I do, when format forgets what I changed?
[21:13:53] <jbk> then you select the target location, all it does is copy the contents of the filesystem over (i.e. cp, cpio, tar, ufsdump, etc.), then runs the upgrade on that
[21:14:08] <madhatter> I just made the slice 0 big enough, named it, saved it and quit.
[21:14:35] <jbk> if you run format again, it _should_ show the changes
[21:14:40] <madhatter> Now newfs tells me c1d0s0 does not exist. When going back into format the slice 0 is gone
[21:14:47] <jbk> hmm..
[21:14:50] <jbk> and no errors?
[21:14:55] <madhatter> None at all
[21:16:03] <jbk> pastebin 'prtvtoc /dev/rdsk/c1d0s2' somewhere
[21:17:07] <madhatter> pardon? pastebin?
[21:17:59] <nachox> crap, sun is now a windows oem...
[21:18:12] *** estibi has quit IRC
[21:18:15] <jbk> nachox: you missed my comments earlier :)
[21:18:37] * jbk asked about horsemen riding around the bay area :)
[21:18:46] <nachox> jbk: sorry, it doesnt look like it but i sometimes work
[21:18:54] <jbk> madhatter: paste the output to opensolaris.pastebin.ca
[21:19:58] <jbk> it'll generate a link, paste the link here
[21:20:07] <jbk> (it's to avoid spamming the channel with lots of output)
[21:20:22] <madhatter> jbk: Thanks, I did not know about that
[21:21:05] <madhatter> http://opensolaris.pastebin.ca/694387
[21:22:02] <jbk> you are doing a 'label' after you change the 0 slice right?
[21:22:17] <madhatter> no, only name
[21:22:23] <jbk> oh
[21:22:42] <jbk> go back in and do that..
[21:23:11] <madhatter> jbk: Much better
[21:23:17] <madhatter> jbk: Thanks a lot
[21:23:20] *** logic100 has joined #opensolaris
[21:23:21] <quasi> label is what writes the slices
[21:23:37] <madhatter> quasi: What does 'save' do then?
[21:24:10] <quasi> writes it to disk
[21:24:20] <logic100> hi, i got my aggregation working with two network cards, is it possible to assign an aggr1 to a zone? or do i really need v-lan?
[21:24:21] <jbk> doesn't that write out the definition to format.dat ?
[21:24:55] <madhatter> jbk: It does nothing else? I thought that's just for backup
[21:26:12] <jbk> really no idea :)
[21:26:34] <jbk> on a related note, has anyone looked at the source for format? :)
[21:26:37] *** nachox has quit IRC
[21:30:07] <jbk> i'm guessing that it's really, really, really old code :)
[21:30:46] <cmihai> logic100, yeah
[21:31:19] <PerterB> jbk: given how similar it still looks to the SunOS format, I would imagine so
[21:32:22] <cmihai> logic100, you can just set physical=aggr1
[21:35:21] <logic100> cmihai: ok thanx, i will try that
[21:37:13] <logic100> cmihai: thanx, that worked
[21:38:02] *** leal has joined #opensolaris
[21:38:05] <leal> hello
[21:38:18] <leal> there is no rsync on solaris 10 default install?
[21:38:43] <leal> i mean, there is tar, wget... but no rsync?
[21:39:11] <wesolows> doesn't look like it - might still be on the companion
[21:39:31] <wesolows> yeah, as ok S10 it's still on the companion, not in the WOS - not sure about Nevada
[21:39:38] *** Dar is now known as Dar_HOME
[21:47:06] <e^ipi> I need a better tamper, this plastic piece of crap isn't working
[21:49:33] * hile_ does a double-take, reads that again
[21:49:45] <jbk> doing yardwork?
[21:49:53] <e^ipi> erm... no
[21:49:55] <e^ipi> making espresso
[21:50:50] *** fbo has quit IRC
[21:50:59] *** tek-ops has joined #opensolaris
[21:51:33] <sommerfeld> it's in the WOS in nevada
[21:51:34] <elektronkind> man, if there's anything I'm good at today, it's deref'ing null pointers in the kernel mod I'm working on
[21:51:55] <quasi> e^ipi: should be made of a heavt metal, surely not plastic
[21:52:02] * elektronkind puts "Dereferences null pointers well" on his resume
[21:52:04] <quasi> heavy
[21:52:13] <sommerfeld> heavy, but non-toxic
[21:52:33] <e^ipi> preferably not "heavy metals", since they tend to get clogged in your liver & kill you
[21:52:37] <e^ipi> but metal, yes
[21:52:40] <tek-ops> hey, this may not be work appropriate but it's finally a "I'm a ..." that includes sun
[21:52:41] <tek-ops> http://img1.purerave.com/3/76/3849776_.jpg
[21:52:42] *** Murmuria has joined #opensolaris
[21:52:45] <elektronkind> no pewter tamper then, eh
[21:52:52] <e^ipi> but 'spro machines tend to come with shitty plastic things
[21:52:55] *** libkeiser has joined #opensolaris
[21:52:56] <tek-ops> if that link doesn't work for you guys let me know
[21:53:15] <sommerfeld> elektronkind: well, solaris specifically tells you when you dereference a "null" pointer as opposed to an otherwise-bogus pointer.
[21:53:24] <sommerfeld> which tells you you're not the first one down that path :-)
[21:53:26] <quasi> e^ipi: amateurish setup ;)
[21:53:48] *** GmanAFK is now known as Gman
[21:54:36] *** logic_ has joined #opensolaris
[21:54:37] *** bengtf has joined #opensolaris
[21:54:40] <e^ipi> heh
[21:54:45] *** logic100 has quit IRC
[21:55:04] *** FireflyST has quit IRC
[21:56:53] *** johnlev has quit IRC
[22:00:17] <CIA-24> aj: 6589362 unnecessary overhead caused by device allocation check when booting zones
[22:00:30] *** triplah_ has joined #opensolaris
[22:02:19] *** Daniel_J has quit IRC
[22:04:12] *** Temporo has joined #opensolaris
[22:04:16] <Temporo> Lo.
[22:05:08] <Temporo> I am looking to contact somebody about usage of the Open Solaris Logo.
[22:05:24] <Temporo> Would any of you know who i should contact?
[22:05:25] <elektronkind> that would be the marketting people
[22:05:44] <elektronkind> http://www.opensolaris.org/os/community/advocacy/marketing/
[22:05:44] <Temporo> I can't find a contact for them
[22:05:55] *** johnlev has joined #opensolaris
[22:05:57] <Temporo> Thank you.
[22:06:06] <elektronkind> you can also try the advocacy-discuss mailing list
[22:06:11] <Temporo> Ta.
[22:06:30] * pjd- just off-line extended raidz from 3 disks to 4 disks.
[22:06:57] *** ocr has quit IRC
[22:07:01] *** estibi has joined #opensolaris
[22:07:02] <Temporo> Catch ya later.
[22:07:05] *** Temporo has quit IRC
[22:10:00] *** invect3 has joined #opensolaris
[22:11:43] *** invect3 has quit IRC
[22:15:23] *** leal has quit IRC
[22:17:43] *** FireflyST has joined #OpenSOlaris
[22:21:07] *** elektronkind_ has joined #opensolaris
[22:21:08] *** elektronkind has quit IRC
[22:24:59] *** holcomb has quit IRC
[22:27:09] *** Murmuri1 has joined #opensolaris
[22:29:15] *** forelle has quit IRC
[22:29:38] *** forelle has joined #opensolaris
[22:29:43] <WickedWicky> I love undocumented configure options
[22:29:57] <BadKarma> such as?
[22:30:46] <WickedWicky> by default gcc tries to compile sparcv9 binaries
[22:30:53] <WickedWicky> even when you're on a sparcv8
[22:30:55] <WickedWicky> very usefull
[22:31:09] <wesolows> huh?
[22:31:12] <WickedWicky> yeah
[22:31:13] <wesolows> if that's true, it's a bug
[22:31:17] <wesolows> we should fix that
[22:31:30] <WickedWicky> I think its in the gcc source code
[22:31:52] <wesolows> I don't really care what the gcc source code does, I care what we ship
[22:32:15] <WickedWicky> this is not opensolaris
[22:32:19] <WickedWicky> dont worry :)
[22:32:21] <wesolows> not true
[22:32:27] <wesolows> it builds v8 by default
[22:32:49] <WickedWicky> well
[22:33:03] <WickedWicky> I am on a ss20, solaris 9, trying to compile 3.4.3 with 2.95.3
[22:33:15] <wesolows> sucks to be you
[22:33:18] <WickedWicky> and it tries to make a libgcc_s_sparcv9
[22:33:19] <WickedWicky> I know
[22:33:20] <WickedWicky> anyway
[22:33:51] <WickedWicky> it tries to build a sparcv9 end result, and fails with "skipping incompatible /usr/lib/libc.so
[22:34:05] <WickedWicky> then I use --with-cpu=v8
[22:34:06] <WickedWicky> and it works
[22:34:14] <WickedWicky> (I also have an ultra 5 now btw, runs very cute)
[22:34:28] *** Vorbis_w has quit IRC
[22:34:31] <WickedWicky> seems related to this
[22:34:33] <BadKarma> heh
[22:34:34] <WickedWicky> http://gcc.gnu.org/ml/gcc/2007-07/msg00877.html
[22:34:44] <WickedWicky> read the bottom
[22:34:55] <WickedWicky> It appears that for Solaris 2.8 gcc
[22:34:56] <WickedWicky> will assume --with-cpu=sparcv9 if you don't explicitly use --with-cpu
[22:35:00] <delewis> I really don't want to feed the off-topic conversation, but gcc does produce 64-bit libs in the build process, should you want to build a 64-bit binary with it afterwards. As for the linking errors you are encountering, those are because you don't have the 64-bit libs installed on your SS20.
[22:35:02] <WickedWicky> seems the same for solaris 2.9
[22:36:09] <WickedWicky> delewis: yep, I didnt toggle the option to do 64 bit support, very possible its that. I didnt see the reason to enable it on a 32bit machine
[22:36:18] <WickedWicky> anyway, I just wanted to show my love for undocumented options
[22:36:25] *** holcomb has joined #opensolaris
[22:36:47] <delewis> WickedWicky: there's your problem.
[22:36:51] <wesolows> I didn't think --with-cpu was undocumented, but ok.
[22:37:03] <WickedWicky> btw, kinda on topic, today I handed in a migration plan to migrate from Linux + XFS + LVM + RAID0 to solaris x86 + ZFS
[22:37:04] <delewis> that's the #1 encountered problem when building gcc on sun4m.
[22:37:13] <wesolows> yay!
[22:37:22] <wesolows> also...raid0?!!!
[22:37:26] <WickedWicky> yea
[22:37:28] <WickedWicky> listen
[22:37:32] <WickedWicky> current construction is this
[22:37:39] <WickedWicky> MSA1500 + 8 MSA20s
[22:37:47] <WickedWicky> MSA20s have RAID6
[22:37:47] <hile_> wesolows, linsux LVM only does RAID0
[22:37:59] <wesolows> hile_: and people use this in the real world?  shocking.
[22:38:04] <WickedWicky> now, all MSA20s are seen as a seperate LUN
[22:38:05] <WickedWicky> no
[22:38:10] <WickedWicky> LVM can be on RAID1/RAID5 too
[22:38:12] <wesolows> I see...you have expensive hardware raid and stupid software raid.
[22:38:15] <delewis> LVM2 can.
[22:38:17] <delewis> LVM cannot.
[22:38:21] <delewis> LVM relies on mdtools
[22:38:22] <WickedWicky> ok, this is LVM2
[22:38:28] <wesolows> all right, that makes more sense
[22:38:42] <WickedWicky> anyway, we choose RAID0 cause it was the only way we could make one big storage pool and still have some performance
[22:38:43] *** FelipeBare has quit IRC
[22:38:53] <WickedWicky> so what I wanna do, is use raidz
[22:39:04] <wesolows> not raidz2?
[22:39:05] <WickedWicky> it would cut on the LVM layer, the software raid layer, and XFS file system
[22:39:08] <WickedWicky> no?
[22:39:12] <wesolows> that's probably closest to what HP calls raid6
[22:39:18] <wesolows> double-parity raid z
[22:39:20] <WickedWicky> raid6 == double parity
[22:39:28] <wesolows> and if you have multiple vdevs, it stripes across them
[22:39:44] <wesolows> so it's not so unlike raid6 hw + raid0 lvm
[22:40:11] <wesolows> anyway, yes it's a big win for simplicity and reliability
[22:40:18] <WickedWicky> yea
[22:40:21] <WickedWicky> and
[22:40:31] <WickedWicky> it will be supported software
[22:40:41] *** eboutilier has joined #opensolaris
[22:40:50] <wesolows> Surely your GNU/Linux vendor gives you excellent support for LVM2!
[22:41:06] <WickedWicky> RHEL doesnt do XFS, we only got it in cause a friend of mine is a kernel engineer at redhat and made a patch for me, but with the restriction I'd never file a bug against it at RH
[22:41:14] <WickedWicky> lol
[22:41:29] <WickedWicky> they do for LVM2, not for XFS
[22:41:38] <wesolows> Well, someone could write an XFS module for Solaris, and Sun wouldn't support that, either.
[22:42:03] <WickedWicky> exactly, but it would be nice to have some sort of support in case something goes wrong
[22:42:14] <wesolows> yup
[22:42:21] <WickedWicky> anyway, I found it a nice opportunity to introduce ZFS at work
[22:42:50] *** Fish has quit IRC
[22:42:59] <hile_> has the LU doesn't play happy wtih ZFS bug been fixed yet?
[22:43:53] <holcomb> negative
[22:43:56] <WickedWicky> HP support wrote me an email that technicaly Solaris is supported with the MSA1500
[22:44:05] <WickedWicky> but we wont get any support from HP support
[22:44:23] <WickedWicky> not that that will be a loss, we dont get it on Linux either, hardly anyway
[22:44:32] *** elektronkind_ has quit IRC
[22:44:36] *** elektronkind_ has joined #opensolaris
[22:44:42] *** elektronkind_ has joined #opensolaris
[22:45:29] <WickedWicky> anyway, I am going to get some sleep
[22:45:30] <WickedWicky> take care all
[22:45:54] <WickedWicky> wesolows, delewis thanks for the off-topic explaination, it's highly appreciated :)
[22:46:09] *** Gropi has quit IRC
[22:47:40] *** Vorbis_w has joined #opensolaris
[22:47:47] *** Murmuria has quit IRC
[22:49:01] <tek-ops> I missed the LU + ZFS bug, haven't had much time for my home projects
[22:49:07] <tek-ops> and I work in a LAMP shop
[22:49:21] <tek-ops> is ZFS now concidered production ready for root partitions?
[22:49:32] <tek-ops> or rather the root partition
[22:50:04] <hile_> I woudln't use it, but then that's because it's not yet been released back to Solaris, and that is all I run in production.
[22:51:13] <tek-ops> thats fair
[22:51:28] <tek-ops> and obvious not use it for OpenSolaris if there are LU issues...
[22:51:34] *** FireflyFJ has joined #OpenSolaris
[22:51:34] <tek-ops> what is the bug exactly?
[22:51:49] <tek-ops> feel free to drop a link if it's documents
[22:53:35] *** elektronkind_ has joined #opensolaris
[22:56:49] <victori_> solaris express does not pop you into x11 by default?
[22:57:09] <victori_> this is on vmware so x11 should of worked if it was the default?
[22:57:31] *** Murmuri1 has quit IRC
[22:58:26] *** triplah_ has quit IRC
[23:01:14] <madhatter> Do you need to apply any patches for live upgrade on sxce? I checked #72099 on sunsolve, but there are only Solaris versions < 11 listed
[23:01:35] <wesolows> there are no patches in SX
[23:01:47] <wesolows> if something is busted, use a newer build
[23:01:57] <wesolows> if the newest build is busted, wait or fix the bug yourself
[23:02:14] <madhatter> wesolows: No, what I ment was for doing the live upgrade itself
[23:02:21] <cmihai> madhatter, there are no issues with LU and SXCE. There is one thing about the new DirectBoot, but that's what onnv flag days are for.
[23:02:23] <wesolows> you mean from S10 to SX?
[23:02:58] <cmihai> Should work too, LU is supposed to work across 2 versions... but since it's not a release, your mileage may vary
[23:03:06] <madhatter> wesolows: I only found a howto for live upgrading solaris 10, and there you have to apply some patches before you can start with the live upgrade
[23:03:23] <cmihai> madhatter, http://unixsadm.blogspot.com/2007/08/liveupgrade-sxce.html - no patches...
[23:04:22] *** forelle has quit IRC
[23:05:14] <madhatter> cmihai: Thanks, I will read on in that document then ;)
[23:05:27] <tek-ops> are there currently any sata controllers that can utilize hot-plugging in opensolaris?
[23:05:27] *** alanc-away is now known as alanc
[23:08:04] *** elektronkind__ has joined #opensolaris
[23:08:05] *** elektronkind_ has quit IRC
[23:13:18] *** FireflyFJ has quit IRC
[23:15:12] *** millhouse has joined #opensolaris
[23:15:19] <millhouse> hey
[23:16:48] *** forelle has joined #opensolaris
[23:17:29] <axisys> i have two identical disks on x4100.. i created mirror using SVM .. they are in sync.. took primamry disk out.. second disk gets Bad PBR Sig
[23:17:50] <axisys> do I need to do anything special beyond SVM disk mirror ?
[23:18:16] *** nrubsig has joined #opensolaris
[23:18:17] *** ChanServ sets mode: +o nrubsig
[23:18:18] <sommerfeld> did you create a mirror of the entire disk or of slices of the disk?
[23:18:32] <nrubsig> Anyone from Sun marketing here ?
[23:18:35] <sommerfeld> svm doesn't mirror the boot blocks
[23:18:35] <axisys> silce 0 (/) and slice 1 (swap)
[23:18:37] <millhouse> i'm trying to get apache2 to start...  I've edited /etc/apache2/httpd-std.conf (renamed it to httpd.conf) but when i enable the service it just says "offline"
[23:19:07] <axisys> millhouse: give it few mins.. it is starting
[23:19:47] <axisys> sommerfeld: so how do I copy boot blocks from primary disk to  secondary disk?
[23:20:09] <hile_> axisys, see installboot(1M)
[23:20:23] <jbk> millhouse: does svcs -x give anything more?
[23:20:41] <axisys> hile_: so i need to installboot.. cannot copy bootblk? which is fine
[23:20:55] <jbk> also, check /var/svc/log/network-http:apache2.log
[23:21:02] <hile_> why do something stupid when there's a documented supported way to do it?
[23:21:10] <axisys> hile_: :P
[23:21:44] <millhouse> axisys:  I've let it run for a while and it never starts
[23:21:56] <axisys> jbk: check the log as jbk suggested
[23:21:58] *** cypromis has quit IRC
[23:22:05] <millhouse> manually trying to start apache2 says that it can't bind to port 80, so i changed it but it still can't bind to it
[23:22:24] <millhouse> i ran 'svcs -x' but it gave nothing on apache2.  all it reported on was printing service
[23:22:37] <coffman> nrubsig: bad joke dude :P
[23:23:15] <millhouse> i checked that log and it said that Method "start" exited with status 0
[23:23:28] <axisys> hile_: installboot is obsolete. Use installgrub(1M)
[23:24:27] <nrubsig> coffman: ?!
[23:24:28] <hile_> ah, I'm on SPARC
[23:24:38] <jmcp> morning all
[23:24:42] *** jcsmith has quit IRC
[23:24:46] <jbk> morning jmcp
[23:24:58] <jmcp> gday gday
[23:25:08] <axisys> x4100 was what I mentioned.. i guess not so stupid huh?
[23:25:16] <axisys> hile_: :P
[23:25:41] <axisys> hile_: thnx tho
[23:27:47] *** elektronkind__ is now known as elektronkind
[23:28:04] <elektronkind> huh wow
[23:28:08] <elektronkind> http://www.sun.com/aboutsun/pr/2007-09/sunflash.20070912.2.xml
[23:28:17] <elektronkind> looks like Lustre is coming to Solaris
[23:28:46] <Triskelios> clustered ZFS?
[23:28:57] <elektronkind> not really
[23:29:00] *** Emmedics4 has joined #opensolaris
[23:29:14] <elektronkind> You can think of Lustre as a proto-pNFS
[23:29:15] *** Mdx4 has quit IRC
[23:29:20] <elektronkind> which begs the question... what about pNFS
[23:29:34] <axisys> elektronkind: wow.. nice .. how appropriate
[23:29:45] <jbk> haha the arstechnica graph is amusing :)
[23:29:59] <axisys> we (in my work) were looking for something similar but within solaris.. this will do just fine
[23:31:58] <coffman> nrubsig: "23:18 @nrubsig> Anyone from Sun marketing here ?" - i consider that a bad joke, or did you ever seen one of them in here?
[23:31:59] *** elektronkind has quit IRC
[23:32:02] *** elektronkind has joined #opensolaris
[23:32:33] <wesolows> the only people who like lustre are hpc wonks; everyone else wants pnfs.
[23:32:38] <coffman> elektronkind: lustre is using zfs as there backend if im not wrong
[23:32:49] <elektronkind> hpc wonks also like pNFS :)
[23:33:00] <nrubsig> wesolows: how did you call me ?
[23:33:00] <wesolows> whether lustre will become more interesting to non-hpc people as Sun works on it is unclear
[23:33:06] <elektronkind> coffman: Lustre as it is right now is a Linux-only think
[23:33:07] <wesolows> nrubsig: hpc wonks
[23:33:10] <wesolows> ?
[23:33:11] <coffman> and in there anounce they said they are doing that on solaris, linux and windows
[23:33:12] <nrubsig> grrrrrr
[23:33:21] <elektronkind> s/think/thing
[23:34:11] <elektronkind> also, what about QFS
[23:34:20] <nrubsig> Well, first Sun ruined QFS, then they failed with the ZFS thing and now it's lustre...
[23:34:31] <wesolows> sorry man, you hpc guys are weird.  The stuff you want costs a huge amount of money to design and build, but the market is tiny.  You're a niche market with expensive tastes.
[23:34:36] <jmcp> nrubsig: a "wonk" generally means a really really smart person who is so removed from reality that they have no idea that their field of interest is so small
[23:34:39] <wesolows> nrubsig: Excuse me?  ZFS failed how exactly?
[23:34:58] <nrubsig> wesolows: ZFS is a bit detached from reality.
[23:35:02] <coffman> i might be wrong about windows..
[23:35:11] <jmcp> nrubsig: generally seen in the term "policy wonk" - particularly referring to the economists who get hired by government treasuries :-)
[23:35:23] <coffman> http://www.clusterfs.com/cfscom/news/latest/cluster-file-systems--sun-to-integrate-lustre-with-opensolaris-zfs-file-systems.html
[23:35:24] <jmcp> and how has Sun ruined QFS ?
[23:35:35] <nrubsig> wesolows: like having no user quoats means that it's unuseable for almost every university. nor is ZFS very fast, even if you try hard.
[23:35:36] <wesolows> ZFS is a mainstream technology that scales well from top to bottom and solves probably 98% of all the filesystem needs any SMB, SME, or enterprise customer might ever have.
[23:35:59] <nrubsig> wesolows: if you exclude univesitiies as customers, sure.
[23:36:00] <wesolows> user quotas can easily be implemented by giving each user a filesystem
[23:36:09] <jbk> and once it's integrated with some sort of live upgrade facility
[23:36:12] *** millhouse has quit IRC
[23:36:13] <nrubsig> wesolows: again: it will not work.
[23:36:28] <wesolows> that your university refuses to do that, despite the fact that creating filesystems is virtually free is not a failing of ZFS but of your administration
[23:36:29] <Auralis> which is simply not practical in edu env, with 10000s of users
[23:36:36] <wesolows> perfectly practical in fact
[23:36:51] <jbk> the one thing that zfs could use in that situation is some sort of automount-like support
[23:37:14] <nrubsig> wesolows: 1. it will not work for shared user dirs and 2. 80000 filesystems in ZFS is unmanageable. University Munich, Stuttgard and the vienna folks tried and failed.
[23:37:21] <wesolows> jbk: the automounter is independent of ZFS - do you mean that it sucks so we should write a new one? :-)
[23:37:42] <wesolows> (and yes, the automounter DOES suck)
[23:37:50] <hile_> 8000 thousand users? ha! try 120,000+ active accounts.
[23:37:54] <holcomb> i thought the "lots of filesystems are slow to mount" thing was fixed?
[23:37:59] <jbk> isn't it basically nfs + lofs
[23:38:10] <jbk> holcomb: there's still non-zero overhead per mounted filesystem
[23:38:20] <wesolows> jbk: the kernel component gets into fucked up states and requires rebooting sometimes
[23:38:21] <jbk> multipled by 10,000 or 100,000 i could see where that would add up
[23:38:45] <jbk> when probably maybe 20 of those filesystems are needed at any given time
[23:39:08] <wesolows> if you want to put yourself in the 2% by refusing to use the tools we offer, so be it.  But don't expect some company to invest $2b in solving your niche problem.
[23:39:36] * wesolows should go away now; HPC people irritate the fuck out of him
[23:40:03] <sommerfeld> holcomb: i believe there is still ongoing work on speeding up zfs mount at boot time
[23:40:18] <sommerfeld> it's been sped up some but there's known room for improvement, too
[23:40:57] <coffman> i dont think thats a niche problem, nor does 80000+ (z)fs has any to do with hpc or is uncommon
[23:41:20] <wesolows> coffman: No, it's not.  It's perfectly within the use cases ZFS can handle.
[23:41:51] <coffman> might be okay for zfs, but its not for the tools around it
[23:42:05] <coffman> yes, zfs needs a special "treat"
[23:42:19] <wesolows> Because the UFS tools are so much better?
[23:42:22] <wesolows> give me a break
[23:42:25] <coffman> no
[23:42:43] <coffman> cause most system manage tools are not ready to deal with zfs
[23:43:03] <wesolows> most system management tools, including those Sun delivers, are basically useless, ZFS aside
[23:43:41] <coffman> sure, thats why most edu folks got there own stuff
[23:43:47] <wesolows> exactly
[23:44:11] <elektronkind> we don't use them here
[23:44:17] <elektronkind> well, use any
[23:44:20] <coffman> i dont say that this is pretty.
[23:44:23] *** Atomdrache has quit IRC
[23:44:24] <wesolows> when I worked for the University, I wrote my own tools, too.  And it's for damn sure I'd have rather build on top of ZFS than the traditional filesystems we had at the time
[23:45:00] <victori> just curious if I boot from the solaris express CD, can I continue the install from an NFS export?
[23:45:29] <elektronkind> sure you can
[23:45:46] <victori> perfect
[23:45:47] <jmcp> victori: I do it quite often
[23:46:05] <coffman> its a thing where in the workflow you have to fix certain things
[23:46:09] <victori> weak that our servers don't have dvd roms old p3 dual boxes
[23:46:12] <elektronkind> instead of selecting the CD as the source for packages, select NFS and enter the host:path to where the Product dir is on your NFS server
[23:46:25] <Triskelios> victori: NFS is faster than CDs/DVDs on some systems actually
[23:46:27] <nrubsig> wesolows: too bad that ZFS can't handle shared workspaces.
[23:46:44] <victori> and zfz isn't the default fs for solaris express is it?
[23:46:48] <nrubsig> wesolows: e.g. how can ZFS filesystems protect single users from filling a filesystem up ?
[23:47:06] *** Emmedics4 has quit IRC
[23:47:20] <coffman> but hack, sun is good on scare of edu folks and worse, the students
[23:47:35] <elektronkind> nrubsig: by carefully managing file system quotas and/or reservations
[23:48:01] *** Atomdrache has joined #opensolaris
[23:48:04] <wesolows> there's no way to do that.  If you want "shared" filesystems, you're taking on a "shared" policy problem, too.
[23:48:17] <wesolows> "It's full?  Too bad, talk to your teammates."
[23:48:20] *** mikefut has quit IRC
[23:48:48] <wesolows> Otherwise it's not really a shared anything anyway.  It's really just a collection of independent filesystems with group read permission.
[23:48:55] <jbk> yeah either each user gets its own bit, or if there's a shared area for the group, the group gets a bit to use by anyone in the group -- it is _shared_ after all
[23:49:31] <wesolows> Also, there are so many ways to get around user quotas it's not even funny.
[23:49:35] *** tek-ops has quit IRC
[23:50:31] <jbk> however, i do think the stumbling block is when the number of filesystems on a system gets large
[23:50:33] <elektronkind> we have users who like to install crap in /tmp :/
[23:51:02] <wesolows> jbk: Agreed - if people are hitting scalability problems like we had on jurassic at one time, taking waaaay too long to mount, those need to be fixed.
[23:53:31] *** Atomdrache has quit IRC
[23:53:55] *** bondolo has quit IRC
[23:53:56] <jbk> elektronkind: that's why you cap the size of /tmp :)
[23:54:47] <elektronkind> jbk: yeah, well, one could. Then when it's 100% full kerberos stops working because that's where users' ticket cache is stored :)
[23:54:50] <hile_> elektronkind, their AFS quotas are too low? (so they think?)
[23:55:11] <jbk> wesolows: i also seem to recall that the amount of kernel ram used/fs while small, tended to really add up when the #'s got into the 10s of thousands (or higher)
[23:55:40] <elektronkind> hile_: it's not uncommon for me to find the latest Thunderbird/Firefox install there... by multip
[23:55:41] <jbk> which is why I thought there was talk of possibly doing some sort of mount-on-demand feature (since presumably not all of them are going to be active at once)
[23:56:00] <hile_> LOL
[23:56:05] <hile_> "install it in your fucking home directory"
[23:56:44] <elektronkind> er... by multiple users. Also it's not umcommon to find <full length movie>.mpg  or ISO rips there, too
[23:57:03] <elektronkind> you know, because we don't give them 6GB quotas ;)
[23:57:14] *** Atomdrache has joined #opensolaris
[23:57:57] <jbk> elektronkind: oh.. not messed with kerberos before :)
[23:58:07] <jbk> at my old job, they decided /tmp should be it's own ufs filesystem
[23:58:18] <jbk> which i fought tooth and nail
[23:58:42] <hile_> so get a filer of some sort and say "put all your stupid shit in $directory; realise that this space is NOT BACKED UP"
[23:58:48] <jbk> their reasoning: 'well in the past 5 years, we've seen around 3 apps that are poorly behaved, therefore we must change it for everything else that is'
[23:59:20] <elektronkind> there's odd shit that ends up in /tmp, too. I once found a professor's personal TurboTax files in there for the past 4 tax years
[23:59:27] <hile_> LOL nice!
[23:59:36] <delewis> I guess he'd never used Solaris before, eh?
[23:59:36] <elektronkind> people are dumb
[23:59:47] <hile_> write something  that cleans up /tmp every few hours to teach 'em





top