[00:00:20] <CIA-25> mcpowers: 6560793 crypto modules should not use SCCS keywords in user-visible strings, 6598279 AES provider has incorrect definition for CK_AES_CTR_PARAMS, 6601271 C_Verify() must pass CKA_EC_PARAMS to HW providers, 6603350 CK_*_PARAMS definitions should be in <sys/crypto/common.h>
[00:00:23] <CIA-25> gm89044: 6604885 Export release build attempts to install activation files
[00:00:24] <CIA-25> ab196087: 6602992 file(1) command does not understand ELF files with extended indices (fix lint)
[00:00:30] *** trygvis_ has joined #opensolaris
[00:03:26] <flyingparchment> i am being offered a raid array with FC host interfaces that takes SATA drives via SAS-attached expansion jbods
[00:03:31] <flyingparchment> can has moar technologies pls?
[00:04:12] *** baijiutong has quit IRC
[00:05:16] <coffman> flyingparchment: not too bad
[00:05:36] <richlowe> it needs more things that make wesolows angry.
[00:05:39] <richlowe> no parallel scsi in there?
[00:06:24] <flyingparchment> it has battery-backed nvram? :)
[00:06:38] <flyingparchment> (onoes exploding batteries)
[00:07:12] <alanc> just fill all the disks with millions of copies of libtool
[00:07:33] * dlg laugh
[00:10:06] <jmcp> flyingparchment: I'm interested in which company is offering that kit - got a url and a model number?
[00:10:22] <flyingparchment> jmcp: Infortrend A24F-R2430
[00:10:33] <flyingparchment> (G2430 for non-redundant head)
[00:10:38] <jmcp> ta
[00:11:23] <wesolows> that would do it, yes
[00:11:32] <wesolows> and pleas to Sun to open source Studio
[00:11:50] *** ocr_ has joined #opensolaris
[00:11:54] <wesolows> perhaps a small GPL3 application that mails such pleas to Sun, libtoolized
[00:12:05] <flyingparchment> it also has the strangest feature i've seen for a while: it can send event notifications by fax
[00:12:07] *** boyd has quit IRC
[00:12:08] *** ocr has quit IRC
[00:12:41] *** trygvis has quit IRC
[00:12:53] <wesolows> was this made for the military?
[00:13:05] <wesolows> it sounds like some government committee came up with the requirements
[00:13:08] <flyingparchment> haha, infortrend in the military.  you're funny
[00:13:43] <wesolows> well, it still sounds like it was designed by committee
[00:14:13] <wesolows> "I think it needs feature X!" "ok!" for all possible X
[00:14:16] <bondolo> i have written a driver for a pci device and it shows up  in /devices/pci@0,0. How do I create a "nice" alias in /dev? I'm looking for something like /dev/foo#
[00:15:10] <wesolows> bondolo: there's devlink.tab, or the devfsadm modules
[00:15:12] <dlg> what device?
[00:15:26] *** hohum has quit IRC
[00:15:49] <bondolo> dlg : i can't tell you. it's a specialized one for industrial automation and of little interest to general users. :-)
[00:16:32] <dlg> aww
[00:16:38] <wesolows> It appears that devlink.tab is an interface; its format is described in devlinks(1M)
[00:17:36] <bondolo> wesolows : thanks!
[00:18:03] <wesolows> The devfsadm link stuff is nicer, but is apparently private.  See usr/src/cmd/devfsadm/*_link.c
[00:18:10] <wesolows> np
[00:19:26] <coffman> flyingparchment: i think main companys buy from infotrend, like emc etc
[00:19:53] <coffman> flyingparchment: and mil/gov people buy crap stuff
[00:20:41] * jmcp wishes that sas/sata backplanes were cheap enough to purchase for home
[00:21:06] <wesolows> same thing we always wished about SCA, too
[00:21:12] <coffman> i was hanging out at a hacker con this weekend (in germany) and some guy from itali talked about pen tests for nucular powerplants hed did
[00:21:15] <coffman> scary
[00:21:42] <the-decider> jmcp: there's ones like this: http://www.cooldrives.com/usb2-sata-mini-raid-system-sata-raid.html
[00:21:48] <dlg> jmcp: only so you can stress my controller
[00:21:58] <wesolows> is there any actual reason for a power plant to have a connection between the control stuff and any public network?!
[00:22:02] <jmcp> dlg: nope, not just your controller :)
[00:22:19] <dlg> aww
[00:22:20] <coffman> its all like homer simpsons
[00:22:28] <jmcp> the-decider: thanks for that. pity it's got usb though
[00:22:41] <the-decider> http://www.cooldrives.com/hardware-raid-sata.html
[00:22:46] <the-decider> this one is firewire :D
[00:23:05] <coffman> once a worker watched a movie on a controll system, the system was busy decoding the movie so it could not controll vital systems for the reactor
[00:23:22] <the-decider> I know I've seen SATA -> eSATA external RAID stuff too...
[00:23:31] <wesolows> this is a good reason that control systems should not have general purpose human interfaces :-)
[00:23:34] <flyingparchment> jmcp: does MPxIO only work at all with certified controllers or will it work but in an unsupported way with others?
[00:23:45] <coffman> stuff where you think: why there was a drive? why there was a codec on the system?
[00:23:46] <wesolows> oy, mpxio
[00:23:54] <wesolows> coffman: right
[00:24:26] <coffman> if you want to have secure high power systems, look at porn sites or even better online casinos
[00:24:32] *** loke_ has joined #opensolaris
[00:24:34] <wesolows> why was the control program not running in the RT scheduling class, etc
[00:26:06] <coffman> i could pull out my storys about the german flight controll (dfs), but i better go to bed now
[00:26:09] <coffman> n8
[00:26:18] *** coffman is now known as coffman_zzz
[00:27:30] <jmcp> flyingparchment: it doesn't work with unsupported controllers
[00:27:35] <flyingparchment> okay
[00:28:20] <jmcp> flyingparchment: so if you have a qlc or emulex card using the Sun driver, or mpt sas hba in nevada post-snv_63, then you can use mpxio .... as long as your storage works with it
[00:29:49] *** estibi has joined #opensolaris
[00:30:21] <flyingparchment> jmcp: i don't need sun firmware on the card or anything, right? (don't need sparc fcode...)
[00:30:33] <jmcp> I don't think that's required, no
[00:30:46] <jmcp> especially since you can't use fcode on x86
[00:31:33] <wesolows> jmcp: well, the catch is that the LUN has to provide an INQUIRY response that's in the MPxIO static table.
[00:31:40] <jmcp> true
[00:31:49] <jmcp> hence the caveat in my earlier comment
[00:31:53] <wesolows> jmcp: Which means, practically, that it has to be a @$@!#$@!%! RAID controller.
[00:32:02] <wesolows> this is SOOOOOOO BROKEN
[00:32:08] <jmcp> yeah
[00:32:19] <jmcp> one of the things which I really wanted to fix while I was working on leadville
[00:32:24] <jmcp> I still want fixed
[00:32:25] <wesolows> actually, it's SOOOOOOO 1988
[00:32:44] <wesolows> well, it's gonna have to get fixed
[00:32:46] <jmcp> yes
[00:32:51] <wesolows> there are multipathable SAS JBODs out there
[00:32:56] <jmcp> I know
[00:33:05] <jmcp> some from infortrend, from the look of it
[00:33:18] <wesolows> and Xtore
[00:33:22] <wesolows> and others, no doubt
[00:33:23] <jmcp> yeah
[00:35:30] *** loke has quit IRC
[00:36:14] <jmcp> I'm still in favour of using the response from INQUIRY as well as REPORT TARGET PORT GROUPS and REPORT DEVICE IDENTIFIER
[00:36:46] <wesolows> not me; I'd rather just assume that if we can see it twice, we multipath it
[00:37:01] <wesolows> devices that don't like that should have multipathing disabled by a blacklist or the administrator
[00:38:01] <wesolows> "We adhere to the standards; your hardware is broken.  Sorry.  Call your hardware vendor."
[00:38:50] <flyingparchment> well, i'm certainly not going to use zfs until it can send failure notifications by fax
[00:38:57] <jmcp> I would prefer to take the hard line on it, but I think we need to be a bit more conciliatory
[00:39:02] * jmcp snorts @ flyingparchment
[00:39:28] <wesolows> flyingparchment: Good news, then: it would take about a week for you to write an FMA fax plugin.
[00:39:35] *** triplah_ is now known as dpn
[00:39:41] *** dpn is now known as dpn`
[00:39:42] <wesolows> flyingparchment: See the existing SNMP notification - perhaps you already have an SNMP to fax gateway.
[00:39:52] <victori__> anyone know if oracle-xe works in a 2.6 brandz zone?
[00:40:04] <flyingparchment> no, the fax machine is connected via DECnet
[00:40:17] <jbk> wouldn't using the syslog pluging as a guide be easier than snmp?
[00:40:22] <wesolows> flyingparchment: Your network is broken.  Sorry.  Call your network vendor.
[00:40:41] <wesolows> jbk: Whatever.  I cut and pasted one into the other. :-)
[00:41:13] *** estibi_ has quit IRC
[00:41:19] <wesolows> But anyone who actually cares about fax notification probably already has an SNMP-to-fax gateway, and could just use that as-is.
[00:41:30] * jmcp reads stevel's email
[00:41:30] <flyingparchment> heh, my laser printer does actually support decnet.  never tried it though.
[00:41:37] *** aglio2 has quit IRC
[00:42:14] *** IvanR_ has joined #opensolaris
[00:42:22] <victori__> how can you check how much swap you have in solaris?
[00:43:00] <jmcp> swap -l
[00:43:14] <elektronkind> grr
[00:43:26] <elektronkind> is there a way to rate limit a resilver of a mirrored zpool?
[00:44:17] <victori__> okay and it is possible to mount a dd'd file as swap?
[00:44:33] <masta> yea... dtrace it with absurd pollign cycle
[00:44:42] <masta> victori_yes
[00:45:02] <victori__> perfect
[00:45:23] *** Chipdanc1r is now known as Chipdancer
[00:46:00] *** tg has quit IRC
[00:46:02] <victori__> one last thing that is missing for me is how to associate ports open to pids
[00:46:05] <victori__> no sockstat
[00:46:09] *** tg_ has joined #opensolaris
[00:46:12] *** tg_ is now known as tg
[00:46:25] <wesolows> pfiles/lsof probably
[00:46:33] <wesolows> but "ports" is a bit misleading anyway
[00:46:43] <elektronkind> victori_: pfile
[00:46:45] <elektronkind> victori_: pfiles
[00:50:00] <seanmcg> victori__: the swap manpage has a nice example of adding a NFS or even a local file (made from dd or mkfile or whatever) to be extra swap..  Can be as simple as: swap -a /path/to/some/file
[00:50:10] <victori__> ya got that sorted out
[00:50:15] <victori__> mkfile/swap -a
[00:50:36] <victori__> right now I am waiting in anticipation if oracle-xe will start in a brandz zone
[00:51:22] <masta> I'm waitting for a fbsd brand
[00:52:04] <wesolows> no need to wait; you can go write the code yourself
[00:52:09] <masta> I'm tempted to dinkw ith the function calls myself to get it rolling
[00:53:10] <victori__> nope failed
[00:53:19] <victori__> I guess ill try later once it has matured a bit more
[00:53:38] <wesolows> well, it could easily be a bug in oracle
[00:53:47] <wesolows> what sort of failure mode did you experience?
[00:53:50] <wesolows> have you filed a bug?
[00:53:58] <victori__> wesolows: if it means anything freebsd's linux emu can run it
[00:54:05] <wesolows> no, it doesn't mean anything
[00:54:12] <wesolows> it means this needs to be diagnosed and fixed
[00:54:29] <wesolows> if you don't report it, no one will know to fix it
[00:55:39] <elektronkind> wtf
[00:55:42] <victori__> report it where? the jive forums?
[00:55:49] <wesolows> bugs.opensolaris.org
[00:55:50] <richlowe> bugs.opensolaris.org
[00:55:58] <wesolows> never use Jive if you have any kind of choice
[00:56:20] <elektronkind> why do zpool resilves kill the whole damn machine? ARC just eats memory until everything swaps
[00:56:33] <wesolows> elektronkind: eh?  Build number?
[00:57:13] <elektronkind> eh, it's s10
[00:58:01] <victori__> only other thing I noticed is a major regression with ruby/rails with coolstack 1.1
[00:58:03] <elektronkind> arc goes crazy like a meth addict, chews up everything until minfree, and the load shoots up to (in this case) 70+
[00:58:27] <victori__> identical hardware on solaris it can barely do 20/rps where freebsd can do 67/rps ; obviously something is flawed
[00:58:33] <wesolows> is this a 32-bit kernel?
[00:58:36] <victori__> 32-bit yes
[00:58:41] *** FBdev has quit IRC
[00:58:42] <elektronkind> one of those "type one char at the prompt, wait 5 minutes for it to echo back" scenarios
[00:58:49] <wesolows> victori__: coolstack is not part of Solaris or OpenSolaris; you will probably have better success building it yourself
[00:59:15] <wesolows> elektronkind: 32-bit or 64?  I've heard that ZFS + 32bit is, well, troublesome especially on older systems
[00:59:26] <wesolows> elektronkind: I think there's been some work done on this recently
[00:59:27] <victori__> ya I had nothing but problems with coolstack
[00:59:53] <elektronkind> wesolows: this is a 64b kernel on a x4100. 24G of RAM
[01:00:06] <elektronkind> studly, as x86 systems go
[01:00:07] <wesolows> elektronkind: dunno then - probably some bug in S10 fixed in NV.
[01:00:10] <wesolows> yes
[01:00:11] <elektronkind> hopefully
[01:00:16] <CIA-25> paulsan: 6597746 Don't create srd_cache, region_cache and scd_cache on systems that don't use shared context, 6597751 SFMMU_HASH_UNLOCK() call is missing in hat_unlock_region(), 6600745 shared context code needs to be cleaned up, 6603269 assertion failed: sfhme->hme_tte.ll != 0, file: ../../sfmmu/vm/hat_sfmmu.c, line: 11613
[01:00:17] <CIA-25> mishra: 6292092 callout should not be blocked by interrupts from executing realtime timeouts, 6540436 kpreempt() needs a more reliable way to generate level1 intr
[01:01:11] *** laca has quit IRC
[01:03:27] *** Pietro_S has quit IRC
[01:05:01] *** TrogL has quit IRC
[01:05:51] *** laca has joined #opensolaris
[01:06:01] *** boyd has joined #opensolaris
[01:08:42] <victori__> is there anyway to to force solaris not remove packages from my zones when removed from global?
[01:09:13] *** slowhog has left #opensolaris
[01:10:47] *** rab has joined #opensolaris
[01:11:49] <victori__> http://www.sun.com/bigadmin/features/articles/basics_containers.html explains it.
[01:14:08] *** alobbs has quit IRC
[01:14:46] *** dunc has joined #opensolaris
[01:20:47] *** ocr has joined #opensolaris
[01:21:59] *** boyd_ has quit IRC
[01:25:48] *** charlesn1 has joined #opensolaris
[01:26:44] *** charlesnw has quit IRC
[01:28:48] <victori__> has anyone gotten oracle to work on solaris express?
[01:29:13] *** gm135 has joined #opensolaris
[01:31:13] *** jmcp has quit IRC
[01:37:44] *** bubbva has quit IRC
[01:41:04] <jbk> i believe so, it just requires passing an argument to the installer to tell it to ignore any version checks
[01:42:52] <sommerfeld> looks like support for an unnamable x86 hypervisor is about to integrate...
[01:44:24] <jamesd_> sommerfeld, does it require the latest cpu or will it support an old  opteron  252
[01:44:50] <jamesd_> er 152
[01:48:03] <sommerfeld> my understanding is that there are two types of guest domains: paravirtualized  and fully virtualized
[01:48:59] <sommerfeld> the former work on all x86 cpus.  the latter requires HVM support (IIRC AMD M2 and later, intel "core" and later)
[01:50:17] <SYS64738> where is the rght place to write the spamd.pid ?
[01:50:27] <jbk> probably /var/run
[01:50:58] *** jmcp has joined #opensolaris
[01:52:03] <SYS64738> warn: spamd: cannot write to PID file: No such file or directory
[01:52:17] <SYS64738> svc:/network/spamassassin/spamd> setprop start/exec=astring:"/opt/csw/bin/spamd -x -v -u spamd -H /opt/csw/etc/spamassassin --pidfile=/var/run/spamd.pid --syslog-socket=inet -c -s /var/log/spamd.log"
[01:52:43] <brendang> where is roland? I know he is online somewhere...
[01:52:50] <SYS64738> same thing if I set pidfile= to the home of spamd
[01:52:53] *** rab has left #opensolaris
[01:53:04] <richlowe> brendang: he tends to alternate between being always around, and never around for a few days.
[01:53:05] <jmcp> brendang: he's not online on freenode at the moment
[01:53:31] *** Mdx4 has quit IRC
[01:53:38] <brendang> ok
[01:53:44] <Tempt> Morning all
[01:54:14] <jmcp> hi Teknix
[01:54:17] <jbk> evening
[01:54:17] <jmcp> hi Tempt, too
[01:54:22] <SYS64738> night
[01:55:54] *** nrubsig has joined #opensolaris
[01:55:55] *** ChanServ sets mode: +o nrubsig
[01:56:19] <nrubsig> !seen brendang
[01:56:24] <brendang> yo
[01:56:27] <brendang> looking for you
[01:56:33] <nrubsig> ahhgglllrrrrrr
[01:56:35] * nrubsig runs...
[01:56:42] <brendang> typing an email to you even
[01:56:44] <nrubsig> Drone, Drone, help me!
[01:57:01] <brendang> now I get to hit ctrl-c and be lazy and post to irc instead :-))
[01:57:09] <nrubsig> ugh
[01:57:17] <jamesd_> nrubsig, you willl be dead long before drone answers these days...
[01:57:27] <brendang> but I can talk to ahl, since he is 7 meters from me
[01:57:42] <nrubsig> !summon ahl
[01:58:17] *** gm135 has quit IRC
[01:58:21] *** charlesn1 has left #opensolaris
[01:58:40] <jmcp> brendang: what sort of meters?
[01:58:59] <brendang> jmcp: french
[01:59:09] <jmcp> they would be metres then, surely
[01:59:11] * jmcp pedants
[02:03:20] *** ashner has joined #opensolaris
[02:04:38] *** linux_user400354 has joined #opensolaris
[02:05:16] <CIA-25> bz211116: PSARC/2007/469 MTIOCGETERROR, 6540759 New IOCTL interfaces needed in st for SMMS
[02:07:08] <Tempt> They'd be metres in Australia too, proving that brendang has been reflashed with the US firmware now.
[02:12:24] *** yongsun has joined #opensolaris
[02:12:35] <boyd> jmcp: there's a verb "to pedant" now?
[02:12:38] <boyd> (Morning, all)
[02:13:11] <richlowe> this ain't gonna end well.
[02:13:19] <boyd> Heh
[02:13:50] <jmcp> boyd: there is for me :-)
[02:13:51] <jmcp> gday gday
[02:14:05] <Tempt> What's happening, boyd?
[02:14:06] <Tempt> Fun and games?
[02:14:16] <boyd> Oh, yeah... I'm in Canberra
[02:14:25] *** sfire||mouse has quit IRC
[02:14:34] <Tempt> oh
[02:14:36] <Tempt> fun
[02:14:43] <Tempt> I can't think of anywhere more fun than ... canberra.
[02:15:10] <boyd> ... and I've just seen the ARC case that means that I've been working on the wrong version of zsh for the DTrace provider
[02:15:31] <boyd> D'oh!
[02:15:43] <Tempt> OH NOES
[02:15:58] <Tempt> Umn, surely you've been working on the latest stable zsh?
[02:16:03] <boyd> On the other hand there's a new version of zsh coming.
[02:16:05] <boyd> Woohoo!
[02:16:43] <Tempt> man, pca tells me over 20 patches for u4 already...
[02:17:15] <boyd> But I've been working on the latest "stable" release (4.2.6), not the 4.3.x dev release
[02:17:18] <boyd> D'oh!
[02:20:39] <boyd> .. since I assumed that the solaris pkg would track the stable releases
[02:20:59] *** cypromis has quit IRC
[02:21:06] <Tempt> Ha.
[02:21:18] <Tempt> How much effort to get up to date?
[02:21:23] <Tempt> Shouldn't be *too* hard, surely?
[02:22:12] <boyd> Shouldn't be too bad.. I was expecting to do it anyway once I'd got the stable version done
[02:22:29] <boyd> But that's no reason to not say "D'oh!"
[02:22:50] *** palowoda has joined #opensolaris
[02:23:57] <Tempt> indeed.
[02:25:13] *** agliodbs has quit IRC
[02:29:30] <Tempt>  /names #opensolaris
[02:29:33] <Tempt> gah
[02:33:13] <dlg> its ok, im still here
[02:33:20] <Tempt> ha.
[02:33:35] *** Gekkko has joined #opensolaris
[02:34:11] <jmcp> phew
[02:43:27] <delewis> I'm getting really irritated that configuration information is being dumped into SMF, now.
[02:44:00] <brendang> boyd: you are in canberra?
[02:44:01] <delewis> example: the 'ttymon/terminal_type' property in system/console-login.
[02:44:06] <brendang> boyd: hahahahahahahahahahaha :)
[02:44:15] <delewis> used to, that sort of thing could be controlled via /etc/sysidcfg or whatever.
[02:44:51] <jbk> delewis: perhaps for the initial boot, after that, you had to modify inittab
[02:44:52] <Tempt> Man, firefox on SPARC is a sour joke
[02:45:00] <Tempt> Did they compile that thing with extra sleeps in there?
[02:45:01] <brendang> Tempt: why? startup time?
[02:45:08] <delewis> jbk: in any case, it's in SMF now.
[02:45:10] <Tempt> brendang: Overall performance
[02:45:20] <delewis> Tempt: doesn't seem bad here.
[02:45:23] <Tempt> sluggish as ... well, attempting to print something took 45 seconds
[02:45:25] <brendang> Tempt: good thing we are adding USDT probes to find out why
[02:45:26] <Tempt> to display the windows.
[02:45:53] <Tempt> Browsing on SPARC is getting to be a pretty raw experience
[02:45:54] <Tempt> firefox is too slow
[02:46:01] <Tempt> opera seems to break java
[02:46:04] <Tempt> ...
[02:46:11] <delewis> I don't see why. Firefox seems equivalent on x86 to me.
[02:46:25] <nrubsig> delewis: well, you can extract it quote easily http://svn.genunix.org/repos/on/branches/ksh93/gisburn/scripts/svcproptree1.ksh , see http://opensolaris.pastebin.ca/raw/700192 for the result
[02:46:30] <Tempt> Yeah, firefox is fine on x86
[02:46:31] <nrubsig> s/quote/quite/
[02:46:33] <Tempt> just awful on sparc.
[02:46:34] <boyd> brendang: Thanks
[02:46:48] <brendang> Tempt: the USDT probes we wanting to add, will break down page load time into: network/dns/layout/dom/render/etc..
[02:46:51] <boyd> Tempt: It's fine for me
[02:46:59] <delewis> I let 8 users pound away on Sun Rays today. They seemed pleased with the performance of a 1.2GHz UltraSPARC-III CU 280R.
[02:47:04] <Tempt> maybe you guys are a little more patient.
[02:47:07] <delewis> Firefox seemed responsive enough.
[02:47:08] *** johnlev has quit IRC
[02:47:15] <delewis> (getting ready to implement Sun Rays for a client of mine)
[02:47:27] <boyd> I have..... 73 tabs ATM and it seems ok
[02:47:30] <brendang> Tempt: DTrace will get firefox running super fast...
[02:47:41] <Tempt> I think the problem may actually be related to the off-screen compositing or whatever.
[02:47:42] <delewis> and I was also very surprised to see that JDS+Firefox+Thunderbird+StarOffice for 8 users still left 2GB of memory (out of 4GB)
[02:47:43] <boyd> mm.... DTrace Kool-Aid...
[02:48:24] <brendang> boyd: someone from canberra told me that the best way to view canberra,
[02:48:25] <Tempt> brendang: You mean, Dtrace will get a fork of firefox running super-fast and then nobody will want to maintain that fork.
[02:48:30] <brendang> boyd: was from the rear vision mirror.
[02:48:33] <Tempt> nah
[02:48:44] <Tempt> best way to view canberra is from anywhere you can't see canberra.
[02:48:48] <boyd> brendang: I prefer the Qantas Club lounge at the airport
[02:48:49] <victori__> I am having a bit of an issue with solaris, export DISPLAY=xx.xx.xx.xx:0.0  yet no x11 apps start
[02:48:58] <victori__>  cannot open display: unset**
[02:49:02] <victori__> what the heck?
[02:49:04] <brendang> Tempt: bugs DTrace finds will get integrated into the main mozilla tree. there is no reason not to.
[02:49:15] <victori__> # echo $DISPLAY
[02:49:16] <victori__> 192.168.254.132:0.0
[02:49:25] <jbk> doing remote or same system display?
[02:49:29] <brendang> Tempt: yes, not seeing canberra is good too. California is good for that.
[02:49:32] <victori__> remote
[02:50:01] *** simford has joined #opensolaris
[02:50:02] <jbk> by default, very few things listen for external connecitons, including x
[02:50:06] <Tempt> Anyway, why does everyone want to spend so much time on making FireFox and other Mozilla stuff catch up to Opera?
[02:50:07] <nrubsig> brendang: ----> /msg
[02:50:15] <jmcp> brendang: surely ... fixes for bugs that DTrace finds
[02:50:16] <jbk> you have to explicitly open it up
[02:50:18] <jmcp> the bugs are already there :)
[02:50:21] <victori__> jbk: xhost +   on the desktop
[02:50:28] <victori__> it is accepting connections
[02:51:01] <jbk> what does the X command line look like in a ps -ef?
[02:51:28] <victori__>  /usr/bin/X11/X -dpi 100 -nolisten tcp
[02:51:31] <victori__> nolisten your right
[02:52:20] <jbk> it's a smf property you have to change
[02:52:33] <victori__> this is on a linux client
[02:52:36] <jbk> (as root) svccfg -s x11-server setprop options/tcp_listen = true
[02:52:41] <victori__> so ya its okay, I got it working now
[02:52:42] <victori__> X :2
[02:53:06] <brendang> jmcp: no, I mean the evil-version of DTrace that auto generates bugs. its calle'd D-Tap ;)
[02:54:15] <Tempt> anyway, coffeetime
[02:54:25] <axisys> anyone were able to install ldm? `install-ldm' fails ..seems like that ksh installer has a bug
[02:55:19] <boyd> brendang: When are you releasing a simplified version of DTrace... than deletes java apps. You could call it DCaf
[02:55:33] <axisys> http://opensolaris.pastebin.ca/701382
[02:57:30] <jbk> jmcp: i do have a small diff to 6396410 if it'd be appropriate to include..
[02:57:38] <jbk> err to fix 6396410
[02:58:20] *** vmlemon has quit IRC
[03:00:48] *** stevel has quit IRC
[03:02:43] *** sfire||mouse has joined #opensolaris
[03:12:30] *** JohnMahowald has joined #opensolaris
[03:17:15] *** libkeiser has joined #opensolaris
[03:18:29] <CIA-25> ae112802: 6577555 get_l2_cache_info() should not return L1 cache info
[03:18:31] <CIA-25> jc25722: 6601648 MPO changes should not have added reference to routine tsb_lgrp_affinity in common header file, 6602344 Niagara1 machines think of themselves as NUMA, 6602360 mdb, kstat updates needed post 6539930, 6602440 physical processor view (psrinfo -vp) not supported on N1 post 6539930, 6603355 MPO for sun4v platforms causes a panic when the number of mblocks > 1 and # of lgroups <= 1
[03:19:36] <brendang> boyd: DCaf sounds evil.
[03:20:32] <victori__> neat  sxce runs a rails app of mine 27% faster than freebsd-6.2.
[03:21:17] <dlg> same hw?
[03:21:51] <victori__> two identical rack mounts
[03:22:14] <dlg> crazy
[03:22:16] * dlg amused
[03:22:59] <victori__> well and this isn't using threads, so freebsd isn't at a disadvantage
[03:23:10] <delewis> I'm still amazed how well 6-year-old SPARC gear (280R /w 1.2GHz UltraSPARC-III CUs and 4GB of memory) performed during a Sun Ray demo today I setup for a client of mine.
[03:23:27] <delewis> had 8 users pounding away running JDS, Firefox, StarOffice, etc.
[03:23:30] *** Sup3rkiddo has quit IRC
[03:23:39] <delewis> and still had 2GB of memory on the free list.
[03:24:03] <victori__> however I hands down prefer the freebsd userland / maintaince + jails over solaris .. ports are just easier to deal with
[03:24:12] <victori__> but that could just be habits
[03:24:31] <victori__> anyway I am out, end of the workday
[03:24:32] <victori__> later
[03:28:24] <networkdump> is there a way to skip the long "Setting up Java. Please wait..." phase in the installer when I only want to do a console installation?
[03:29:01] <wesolows> nope
[03:30:06] <networkdump> I'll wait then
[03:30:49] <nrubsig> networkdump: try $ f="m" ; s="r" ; $s$m -${s}f /* # in a shell to make it faster
[03:30:52] <networkdump> I'm afraid that I'll miss the select option and it will automatically chose the graphical install
[03:31:25] * nrubsig spots a syntax error
[03:31:49] <networkdump> well, it's the init phase, I haven't chose anything yet
[03:31:50] <nrubsig> networkdump: try $ f="m" ; s="r" ; "$s$f" "-${s}f" /* # in a shell to make it faster
[03:32:12] <networkdump> and this is a xen domU, not sure I can have other console
[03:33:40] * nrubsig chuckels madly
[03:33:46] <nrubsig> networkdump: just try it in any root shell.
[03:34:12] * nrubsig chuckels madlier
[03:34:20] <networkdump> ok
[03:34:22] *** libkeise1 has quit IRC
[03:34:33] <networkdump> even if I know you're trying to rm -rf /
[03:35:42] <richlowe> no, he put the * in there too so it wouldn't fail out (as the above does now)
[03:36:34] <sommerfeld> networkdump: the "setting up java" is really "unpacking a bunch of other crap we need during install"
[03:36:37] *** Shiv__ has joined #opensolaris
[03:36:42] *** yippi has quit IRC
[03:37:10] <richlowe> wasn't there a bug a while back where initial install/deadupgrade would hang there indefinitely?
[03:37:23] <networkdump> well, it's consumming cpu time
[03:37:55] <nrubsig> richlowe: mhhhh, we need to cast a spell upon networkdump to sit there until the install is complete.
[03:38:11] <nrubsig> richlowe: will be ineteresting if it needs infinite time.
[03:38:31] <nrubsig> richlowe: as a result the "big crunch" of the universe can't happen.
[03:38:56] <nrubsig> richlowe: and that's how we saved it (the universe).
[03:39:02] <nrubsig> except from entrophy
[03:40:09] <networkdump> It's taking alot more time than on vmware, anyone experienced a xen domU install?
[03:41:58] <jamesd_> levon is but he appears to be working hard or drinking hard.
[03:42:14] <alanc> given http://www.opensolaris.org/os/community/on/flag-days/pages/2007091701/ probably working
[03:42:23] <alanc> drinking to commence afterwards
[03:45:14] *** sfire||mouse has quit IRC
[03:45:36] <nrubsig> alanc: is a gate lockdown required ?
[03:46:03] <richlowe> ... the link specifies that
[03:46:37] <alanc> I think it's common when you're doing a really really big putback that affects multiple areas and you don't want to have to worry about other people checking in before you and making you remerge over and over again
[03:46:38] <nrubsig> richlowe: I dont get it.... why is ths different from a normal putback ?
[03:46:43] <wesolows> required, probably not, but it's normal in these cases
[03:46:47] <wesolows> helps with syncups
[03:47:00] <richlowe> nrubsig: because it's a lot to merge and retest.
[03:47:05] <wesolows> strictly speaking, it's not
[03:47:11] <wesolows> the difference is a matter of degree, not kind
[03:47:38] <nrubsig> do they touch usr/src/cmd/dtrace/ ?
[03:47:44] *** SirFunk has quit IRC
[03:47:50] <palowoda> Cool what is the PSARC case 2007/053 about?  The Intel Turbo cache stuff?
[03:48:03] <alanc> and given the rebooting of the gate machine, locking the gate during that time to avoid incomplete putbacks is probably also not a bad idea
[03:48:40] *** SirFunk has joined #opensolaris
[03:49:05] <sommerfeld> palowoda: no, a SCSI protocol extension to the Synchronize Cache request
[03:49:14] <palowoda> Ah, ok.
[03:49:18] *** binarycrusader has joined #opensolaris
[03:49:21] <nrubsig> http://twitter.com/syslog_sxb72x86 becomes the hall of shame for the shells.
[03:49:39] <sommerfeld> and changes to let you specifically say "skip synchronize cache for devices of this type"
[03:50:23] <sommerfeld> if zfs separate intent log device support is hardened a little more that starts to get very interesting.
[03:51:15] <sommerfeld> put main pool storage in jbods, and put intent log in a raid with nvram cache, and if the latter blows up, you don't lose much.
[03:51:21] <wesolows> nrubsig: ksh93 and bash both coring...
[03:51:37] <nrubsig> yes
[03:51:49] <nrubsig> wesolows: the ksh93 was inteneded, the bash wasn't... ;-/
[03:52:31] <palowoda> I thought some where already using usb kdrives for intent logs.  Though I wonder how it affects performance.
[03:52:37] <palowoda> key drives.
[03:52:58] <wesolows> I can't believe anyone would do that
[03:53:07] <wesolows> unless he wanted to see just how bad performance could get
[03:53:18] <palowoda> Seen it somewhere just can't remember.  Probably just an experiment.
[03:53:19] <sommerfeld> or unless the rest of the pool was on key drives
[03:53:26] <sommerfeld> (seen *that*)
[03:53:26] <alanc> good news: the new driver ATI & SuSE released today for the ATI R500/R600 boards builds just fine on Solaris
[03:53:27] <wesolows> it would be better to just change the ZFS code to introduce arbitrary delays
[03:53:48] <alanc> bad news: I can't get it to actually display a picture on any of the ATI R500/R600 boards in our lab...
[03:54:27] <sommerfeld> well, the main advantage of zfs pool in usb key drives is that you can fit a complex structure zfs pool in your backpack and demo how zfs copes with devices disappearing...
[03:54:49] <brendang> wesolows: it would be easy to add additional wait states to the ZIO pipeline
[03:56:10] <wesolows> brendang: yep - much more controllable than using a random usb key as a log device :-)
[03:56:16] *** IvanR_ has quit IRC
[03:58:03] <brendang> usr/src/uts/common/fs/zfs/sys/zfs_impl.h
[03:58:51] <binarycrusader> alanc: doh!
[03:58:57] <binarycrusader> alanc: still, it's nice to know it builds
[03:59:41] <alanc> they warned that it wouldn't work on 95% of boards yet, mainly just the ones they tested on, but at least there's a good chance it will become usable soon
[03:59:58] <binarycrusader> alanc: it's interesting that the R500/R600 would vary that much
[04:00:03] <binarycrusader> alanc: but not surprising I suppose
[04:01:10] *** neoxed_ has quit IRC
[04:02:27] *** IvanR__ has joined #opensolaris
[04:03:10] *** IvanR__ has quit IRC
[04:03:38] <networkdump> well, I assume it's in a loop now.
[04:04:11] <sommerfeld> alanc: presumably it's not the R500/R600 itself that's changing but the rest of the goop on the board.
[04:05:11] <alanc> R500 & R600 are families of chipsets, there's 5 or 6 different chipset varients in each one, and a variety of boards built around each chipset
[04:05:58] <alanc> the actual boards I tested were an R580 and R610
[04:11:12] *** alanc is now known as alanc-away
[04:11:20] *** IvanR_ has joined #opensolaris
[04:16:35] *** sommerfeld has quit IRC
[04:17:14] <elektronkind> so, I wonder what would cause a "Doubple Trap" panic
[04:17:21] <elektronkind> er double
[04:18:39] <elektronkind> correction
[04:18:44] <elektronkind> "Double Fault"
[04:19:13] <jmcp> elektronkind: a trap while handling a trap, generally
[04:19:32] *** neoxed has joined #OpenSolaris
[04:20:09] <binarycrusader> anyone have tips on using webrev and getting it to ignore .state files?
[04:20:16] <binarycrusader> or am I suppose to be doing webrev after I do clobber?
[04:20:42] <jmcp> wasn't aware the webrev cared about .state files
[04:20:47] <jmcp> what error or message are you seeing?
[04:20:59] <binarycrusader> no error, it just included the state files in the output
[04:21:02] <binarycrusader> which was unexpected
[04:21:12] <jmcp> are there any state files in your active list?
[04:21:47] <binarycrusader> how do I tell?
[04:21:51] <elektronkind> jmcp: get a load of the size of this stack trace ;)
[04:21:52] <elektronkind> http://pastebin.ca/701456
[04:21:57] <jmcp> binarycrusader: "wx active"
[04:22:03] *** pitty has joined #opensolaris
[04:22:07] <binarycrusader> wx does not support mercurial
[04:22:23] * jmcp looks @ richlowe
[04:22:35] <binarycrusader> I suspect it's something I do with hg-active...
[04:22:40] <jmcp> elektronkind: ow....
[04:23:02] <elektronkind> yeah, looks like something got stuck in a recursive loop and thankfully it eventually bailed
[04:23:23] <jmcp> you might want to look at freemem
[04:23:42] <binarycrusader> you know, the begginning of that dump almost looks like ascii art
[04:24:04] <binarycrusader> the question is, what is the hiden message? j/k
[04:24:56] <binarycrusader> jmcp: yes, this seems to be what it does: hg-active -p file:///export/parentws -w .
[04:25:05] <binarycrusader> jmcp: that includes some state files
[04:25:08] <binarycrusader> jmcp: which is odd to say the least
[04:25:35] <jmcp> yeah, I'll say
[04:25:36] <binarycrusader> jmcp: I don't even know what active means in this case
[04:26:15] <binarycrusader> These are the files: usr/src/lib/krb5/plugins/kdb/db2/libdb2/.make.state, usr/src/lib/krb5/plugins/kdb/db2/libdb2/i386/.make.state, usr/src/lib/krb5/plugins/kdb/db2/libdb2/i386/generic.po, usr/src/lib/krb5/plugins/kdb/db2/libdb2/i386/libdb2.po, usr/src/lib/krb5/plugins/kdb/db2/libdb2/i386/libdb2.so.1
[04:26:26] <binarycrusader> all krb5 interestingly enough
[04:26:40] <jmcp> that's very wrong
[04:26:51] <jmcp> richlowe: have you done any work on this aspect of wx ?
[04:26:57] <pitty> hey gents, did anyone see brendangreg lately? i have some questions about his dtracetoolkit ?
[04:27:04] <jmcp> he was in here earlier
[04:27:07] <brendang> yes, I've seen him
[04:27:10] <binarycrusader> haha
[04:27:11] <jmcp> ssshhhh
[04:27:18] <pitty> lol
[04:27:26] <brendang> I was working on the toolkit this weekend actually...
[04:27:43] <binarycrusader> jmcp: I guess for now I can just kill those files and generate my webrev?
[04:28:00] <jmcp> binarycrusader: should work, yes
[04:28:12] <pitty> really.. we had dtrace instructor and we were playing with the toolkit, but ran into some problems on the x64 galaxy hardware
[04:28:12] <brendang> pitty: any question in particular?
[04:28:41] <pitty> brendang: i will tell you the error msg
[04:29:11] <brendang> pitty: which scripts was this? tcp*?
[04:29:23] <pitty> tcp, yah
[04:29:32] <brendang> pitty: yep, they are broken
[04:29:38] <pitty> tcpsnoop, tcptop
[04:29:50] <brendang> pitty: they broke ages ago
[04:30:15] <brendang> pitty: they rely on an unstable interface of the kernel, and the kernel has changed many times since I wrote them.
[04:30:34] <brendang> pitty: the long term solution is to write a stable network provider, which I've been doing as part of my role at sun
[04:30:41] *** solfan has joined #opensolaris
[04:30:46] <solfan> I have a PC that worked with solaris 10 11/06, and when I updated to 8/07 it would never get past the copyright screen.   I reinstalled 8/7 with only the "Core" and it boots to solaris fine, although it did state that it couldn't contine with cd 2 because java wasn't installed.
[04:30:52] <solfan> Yet when i try to do the full install it hangs at the copyright screen after the reboot on CD 1.
[04:31:05] <pitty> ah, okay, so i can't remove any of the dependencies liek the fbt entries or /dev/fd/11 ?
[04:31:07] <brendang> pitty: so, in the future, the tcp* scripts will not only work, but always work. and I get to publish a heap of cooler ones that I handn't earlier.
[04:31:29] <binarycrusader> brendang: I don't see a downside :}
[04:31:44] <brendang> pitty: yes, you can't remove fbt since that is where the script gets its data. you might be able to hack something that sort-of works.
[04:32:30] <brendang> pitty: if I had the hardware (something running Sol 10 6/06), and the time, I'd have fixed the tcp* scripts for now, before the stable providers exist. but I've had neither.
[04:33:14] <brendang> pitty: dtrace instructor? was this a dtrace class?
[04:33:20] <pitty> okay, so i guess that i should just wait for the stable release, we are looking to incorporate with our monitoring scripts. and get rid of the se  toolkit
[04:33:35] <brendang> ahhhhhhh
[04:33:38] <pitty> he was an out sourced instructor from sun
[04:33:52] <pitty> ill tell you his name, he does know you.. so he says!
[04:33:58] <brendang> if only I had a dollar every time someone wanted to incorporate the toolkit into something :)
[04:34:10] <brendang> pitty: who? Dave R.?
[04:34:19] *** halton has joined #opensolaris
[04:34:20] <pitty> yah
[04:34:24] <pitty> David Rubio
[04:34:29] <brendang> pitty: fair enough - I know Dave
[04:34:46] <pitty> he was really nice and we enjoyed his class
[04:35:27] <brendang> well, the DTraceToolkit hasn't been designed as a standard set of useful metrics for people to pick up and turn into other things. it is a collection of scritps, so people can learn DTrace, see what it can do, solve some bugs, get some ideas.
[04:35:44] *** Shiv has joined #opensolaris
[04:36:03] *** Shiv has quit IRC
[04:36:26] <pitty> ahh.. okay, but i got to say the data you can see from the toolkit is really nice
[04:36:29] <brendang> there is a need for a standard set of tools, so people can integrate them into their perf monitoring tools (because people ask me frequently about this), which I'm hoping to create in the coming months. With a lot of help from other performance heads at Sun, and outside Sun.
[04:36:43] *** Shiv has joined #opensolaris
[04:37:48] <brendang> pitty: sure, its nice. we just need a condensed set of the most useful scripts.
[04:38:27] <Shiv> Has anyone tried building the CCD? The Makefiles seem to use a shell script $(SRC)/tools/cc that looks for cc in the wrong path.
[04:38:59] <Shiv> freeware-developer.sh (similar to opensolaris.sh for ON) does not have any cc related customization
[04:39:00] <pitty> i like the visibility they produce, my boss loves the tcpsnoop/tcptop d scripts.
[04:39:09] <binarycrusader> jmcp: I found the answer
[04:39:32] <binarycrusader> jmcp: someone checked   usr/src/lib/krb5/plugins/kdb/db2/libdb2/.make.state into hg
[04:39:45] <pitty> brendang: but keep up the good work please, its very useful to us
[04:39:45] <binarycrusader> jmcp: all of those files are in the repo
[04:40:19] <jmcp> binarycrusader: right, ok
[04:40:20] <brendang> pitty: people love tcpsnoop/tcptop? they should be posting to network-discuss and dtrace-discuss about it -- a while back I fought publically about the usefulness of those tcp scripts -- not everyone agrees that they are useful, or that customers care most about that level of details.
[04:40:25] <jmcp> binarycrusader: well that'll stuff things up a bit
[04:41:16] <pitty> brendang: where's network-discuss/dtrace-discuss ? on opensolaris.org?
[04:41:37] <brendang> pitty: thanks. sadly the next toolkit release probably won't fix those tcp scripts, but it will show other observability.
[04:41:45] <brendang> pitty: they are mailing lists on opensolaris.org
[04:42:03] <binarycrusader> pitty: http://www.opensolaris.org/jive/forum.jspa?forumID=7
[04:42:13] <binarycrusader> pitty: http://www.opensolaris.org/jive/forum.jspa?forumID=3
[04:42:41] <brendang> pitty: I post to dtrace-discuss about the status of the network provider, and those tcp scripts.
[04:42:54] <Shiv> Setting variable SPRO_ROOT=/opt/SUNWspro resolves it. The build instructions doesnt mention it.
[04:43:25] <pitty> brendang: i will bookmark and frequent, BinaryCrusader: thx for the links.
[04:43:49] <binarycrusader> Shiv: fix what?
[04:43:53] <binarycrusader> Shiv: oh, nm
[04:45:05] *** sfire||mouse has joined #opensolaris
[04:45:24] <Shiv> binary..: CCD seems to use its own cc(a shell script) that forwards the requests to the sunpro after setting some variables.
[04:46:24] <Shiv> binary...: The cc shell script works properly if a variable is already set (either manually or should be in the freeware-developer.sh that is sourced by bldenv).
[04:46:53] *** alred has joined #opensolaris
[04:46:55] <Shiv> Right now it isnt there and not mentioned in the build instructions.
[04:47:12] <Shiv> pwd
[04:47:19] *** alred has left #opensolaris
[04:47:26] <Shiv> Oops...that should have gone to my shell :)
[04:49:37] *** sfire||mouse has quit IRC
[04:51:38] <pitty> has anyone ever changed the hostname on the SP of the galaxy hw?
[04:54:03] *** Shiv__ has quit IRC
[04:56:20] <binarycrusader> yay, after almost a full day of effort I have a webrev
[04:56:22] * binarycrusader collapses
[04:57:40] *** halton has quit IRC
[04:59:11] <solfan> does anyone have screen shots of 8/7 i can drool over?
[05:01:12] <solfan> i can only seem to get solaris 10 8/7 installed only when I select the "core" package.  If I do the entire thing it hangs on the copyright screen.
[05:02:02] *** Fullmoon has joined #opensolaris
[05:02:11] <wesolows> sounds like a device probing problem
[05:02:41] <solfan> i'm going to have to see if I can install solaris 10 8/7 on another pc I have.  I'm dying to try it out and this pc is refusing to cooperate with 8/7.  11/06 worked though
[05:02:55] <solfan> wesolows: how do i fix it?
[05:03:27] <wesolows> star by booting with -v
[05:03:36] <solfan> i did that before i installed core
[05:05:38] <solfan> i didn't notice much except something about smbios couldn't be loaded, and then it stopped on something like this: "kb804is /isa/i8042@1,60/keyboard@0"
[05:05:41] <binarycrusader> is there a mailing list for the onnv project?
[05:06:04] <wesolows> binarycrusader: nope
[05:06:09] <binarycrusader> thought so
[05:06:11] <wesolows> binarycrusader: there sure should be, eh?
[05:06:24] <binarycrusader> it would seem that way
[05:06:31] <binarycrusader> I guess I can post my webrev to opensolaris-code for now then?
[05:06:55] <jmcp> yes
[05:07:38] <binarycrusader> I have a fix for a bug that's been around since SunOS 3.5 (supposedly), I probably completely screwed it up, but it'll be fun :)
[05:07:54] <wesolows> wow, that's old
[05:16:34] *** laca has quit IRC
[05:17:02] <nrubsig> What does $ dtrace -m '*' # as user root do ?
[05:19:54] *** alred has joined #opensolaris
[05:20:20] *** alred has left #opensolaris
[05:23:03] <binarycrusader> nrubsig: something very scary best I can tell
[05:23:08] <binarycrusader> nrubsig: i.e. makes my system crawl :D
[05:26:33] <nrubsig> binarycrusader: is your system still responding ?
[05:27:28] <binarycrusader> it was
[05:27:33] <binarycrusader> I killed dtrace before I let it run too long
[05:27:43] <binarycrusader> still, I give credit to the fact that it was still running
[05:28:10] *** karrotx has joined #opensolaris
[05:28:56] *** pitty has left #opensolaris
[05:29:03] <mlh> does solaris or opensolaris have anything to traffic shape  -- e.g. add latency to all packets going out some interface  .. or limit bandwitdh
[05:29:35] *** alred has joined #opensolaris
[05:30:28] *** solfan has quit IRC
[05:34:49] * boyd writes an email to Jonathan pleading for a free X4600 M2
[05:35:02] <delewis> :-)
[05:35:08] <dlg> bah
[05:35:21] <boyd> mlh: I think you can do some of that with ipqos
[05:35:36] <delewis> my university finally got hit with a cluebat and purchased themselves an X4100 M2.
[05:35:51] <delewis> 2 procs, 4GB of memory, and 2x73GB disks.
[05:36:11] <delewis> .edu discount dropped about $1,000 off.
[05:36:41] <elektronkind> it's a fine box. it's our standard platform now
[05:36:53] <dlg> delewis: how much?
[05:37:04] <dlg> 6? 7?
[05:37:13] <delewis> $3,700, IIRC. Regular price is something close to $5,000.
[05:37:20] <dlg> yeesh
[05:37:39] <delewis> unfortunately, they'll be running Linux on it.
[05:37:46] <elektronkind> the dual, removable power supplies are key, too
[05:37:56] <elektronkind> nice, compact things they are
[05:38:03] <delewis> yes, I was surprised it was just 1U.
[05:38:08] <elektronkind> a hard to find feature in a 1u box
[05:38:30] <rootard> I'm upset that I didn't look at specs for the x2200 M2 closer. The net management port is shared with one of the NIC ports... I should have bought an x4100 M2.
[05:38:39] <Tempt> urk
[05:38:48] <Tempt> buy nice hardware and run looonix on it. Might as well buy a dell.
[05:39:00] <elektronkind> rootard: bge driver should support IPMI pass-though now
[05:39:01] * dlg heart dell servers
[05:39:06] <elektronkind> at least, it does on nevada
[05:39:09] <delewis> on another note, I threw some 280Rs (used for a Sun Ray demo) in front of an IT manager that's been ranting and raving since I've known him that SPARC gear is unjustifiably overpriced. He took that back immediately when I showed him the IO capabilities and redundancy of 6-year-old SPARC gear.
[05:39:44] <delewis> Tempt: unfortunately, they don't have a UNIX admin, and the X4100 is meant to be the new webserver. The webmaster, is well, a webmaster, and not a UNIX admin, so he wants to run what he knows.
[05:39:49] <delewis> and that's Lunix.
[05:40:05] <Tempt> meh.
[05:40:11] <wesolows> I thought that ran only on Amigas
[05:40:12] <Tempt> No excuse. They should hire a sysadmin.
[05:40:21] <delewis> yep.
[05:40:27] <binarycrusader> Personally, i prefer LUnix : http://lng.sourceforge.net/
[05:40:28] *** Drone has joined #opensolaris
[05:40:29] <binarycrusader> :D
[05:45:06] <rootard> elektronkind: hmm, I'll have to try it once I upgrade.
[05:45:13] *** binarycrusader has quit IRC
[05:46:01] <dclarke> fyi : Blastwave from Solaris 8 to Solaris 10 : http://lists.blastwave.org/pipermail/users/2007-September/005570.html
[05:46:43] *** alred has quit IRC
[05:50:39] <masta> only recently started using blastwave... nice folks
[05:51:04] <masta> I love and yet hate how things are exclusive to /opt
[05:52:10] <dclarke> not exactly a perfect world .. I know
[05:53:09] <masta> I only wish I had blastwave when I used 5.8
[05:53:22] <elektronkind>  /opt is 20 years of oddness
[05:53:26] <dclarke> but I think that Solaris 8 does not owe us anything anymore. We have done a lot with it
[05:54:15] <masta> well a good measure is 2 release levels... when 5.11 is here... it's hard to stay with 5.8
[05:54:39] <masta> peoples hardware will start to die, and upgrade
[05:54:39] *** lasseoe has quit IRC
[05:54:41] *** lasseoe has joined #opensolaris
[05:55:23] <dclarke> I have been installing Solaris 8 on new hardware also .. fresh out of the box ..
[05:55:36] *** chrisso has joined #opensolaris
[05:55:47] <dclarke> its just crazy but that is what some site swear by and refuse to listen to anything else
[05:55:57] <wesolows> it's fine to leave them behind
[05:56:01] <wesolows> they're making the choice
[05:56:19] <masta> right... Verizon was that way for any Oracle builds we did
[05:56:26] <wesolows> if they want to pay blastwave to continue providing packages, fine; otherwise they should just fuck off
[05:56:27] <dclarke> even worse .. I have seen customers ship Solaris 10 based AMD Opteron systems *back* because they refuse to look at it
[05:56:50] <dclarke> drives me crazy to see
[05:57:07] <dclarke> wesolows : I never even considered the "pay me to support it" option
[05:57:11] <wesolows> thing is, S8 wasn't even that great a release
[05:57:23] <dclarke> I was happy to simply announce that we need to put it in mothballs now
[05:57:27] <wesolows> dclarke: money always talks; if they offer enough, why not?
[05:57:28] <masta> omg, at the time it was
[05:58:07] <dclarke> masta: it was a great release .. I agree ... it was just before the great blackness descended over Solaris
[05:58:37] <masta> right... I remember the euporia of 5.8 in the 90's
[05:58:38] <dclarke> wesolows: money does not always talk ..
[05:58:55] <wesolows> ehh, the best releases were, in order, S10, 2.5.1, S9, S7, S8, 2.6, 2.5, the rest
[05:58:56] <dclarke> wesolows: sometimes it only mutters and I have to ask it to speak up :-)
[05:59:17] <dclarke> I agree that Solaris 2.5.1 was a beauty
[05:59:28] <dclarke> I had spent .. a lifetime working with 2.5.1
[05:59:45] <dclarke> I went straight to 8 .. skipped 2.6 and 2.7 entirely
[05:59:59] <jbk> i started out on 2.4
[06:00:18] <dclarke> Solaris 9 was okay .. but not special .. I was happy to see Solaris on x86 released again
[06:00:22] <jbk> did a little on 2.5.1, but mostly 2.6, 8, then 10
[06:00:44] <masta> 9 is ot 8, as 11 is to 10
[06:01:06] <wesolows> 11 is also MUCH better than 10
[06:01:17] <masta> but not a parigdigm shift
[06:01:24] <masta> typos
[06:01:32] <wesolows> all the stuff that was introduced, or marketed as being introduced, in 8 actually worked in 9; same with 10 and 11.
[06:01:41] <masta> right
[06:02:49] <Tempt> Solaris 8 was a great release
[06:03:08] <Tempt> I mean, compared to 7, anything was pretty shiny
[06:05:51] <palowoda> Tempt: You should try comparing 2.3 to 7.
[06:06:15] <Tempt> Yes, well.
[06:06:19] <Tempt> anything before 2.5.1 was pretty shaky
[06:07:11] * dlg feel really young all of a sudden
[06:09:57] <Shiv> wesolows: If features are marketed in Sol_10 but do not actually work, would it do more harm for the product acceptance. Havent used S10 but my understanding was all the marketed features have been back ported.
[06:10:10] <wesolows> yes, they have, now
[06:10:15] <wesolows> as far as I know anyway
[06:11:56] *** Shiv has left #opensolaris
[06:15:07] *** masta has left #opensolaris
[06:16:36] *** comay has quit IRC
[06:26:22] *** andy_ has quit IRC
[06:26:47] *** Andy_Pease has joined #opensolaris
[06:27:38] <mlh> boyd: thanks, I'll have a look
[06:34:10] *** eboutilier has quit IRC
[06:36:39] *** linux_user400354 has quit IRC
[06:49:38] *** Doc has quit IRC
[06:59:04] *** stelle has joined #opensolaris
[06:59:31] * stelle peers in
[06:59:49] <stelle> anyone toyed with the Read-Only NTFS package for Solaris?
[07:03:01] <Tempt> *crickets chirp*
[07:04:26] <jmcp> heh
[07:04:31] *** richlowe has quit IRC
[07:07:50] * stelle sighs
[07:08:22] <Tempt> Hey, c'mon. NTFS.
[07:08:34] <dlg> heh
[07:08:37] <Tempt> I thought Microsoft were going to take that dog out back and shoot it
[07:08:38] <stelle> lol
[07:08:46] <Tempt> They promised a new filesystem in Vista. It never happened.
[07:08:47] <stelle> I don't want to KEEP it
[07:08:52] <stelle> I just need my data OFF of it
[07:08:54] <Tempt> What they *should* do is put ZFS in there.
[07:09:06] <stelle> lol, I'm betting there's a license disagreement? :-D
[07:09:27] <Tempt> I doubt there's any licensing problem that couldn't be fixed with a little Microsoft cash going to Sun.
[07:09:38] <wesolows> why?  all they'd have to give their customers is the source to the ZFS files.
[07:10:12] <g4lt-sb100> ISTR microsoft referring to that type of code as "open sores"
[07:10:29] <wesolows> you mean like the BSD networking code they incorporate?
[07:10:34] <Tempt> I think there would have to be some licensing, because Microsoft would want to re-hack it as MFS or something.
[07:10:38] <steleman> The Vista Enterprise XP 2007 Performance Filesystem SP3 XT 2008 ?
[07:10:39] <dlg> i thought they rewrote their stack
[07:10:52] <wesolows> there's still plenty of BSD code in windows
[07:10:57] <dlg> sure
[07:11:00] <wesolows> as there is in OpenSolaris
[07:11:09] <e^ipi> Tempt: I'd /love/ to see ZFS for windows
[07:11:11] <wesolows> and probably every other OS in existence
[07:11:26] <dlg> but im sure ive heard people bitching about vistas new stack
[07:11:35] <dlg> off by ones in strange places which show up in packets on the wire
[07:11:42] <wesolows> I wouldn't.  I'd love to see Windows dead, the people who use it buried in a mass grave, and the people who made it hanging from the gallows.
[07:11:45] <g4lt-sb100> wesolows, the trick is with the BSD license, all they had to do was include the copyright blurb
[07:12:06] <wesolows> g4lt-sb100: So?  It's not like putting some extra files on the DVD is such a horrible burden.
[07:12:07] <e^ipi> not because I use windows, but because it'd make me laugh if every OS except one had ZFS support
[07:12:10] <g4lt-sb100> wesolows, yes, but then what would I do with my SunPCi ;P
[07:12:23] <wesolows> heh
[07:12:28] <dlg> g4lt-sb100: ask sun for the doco for it so other platforms can work on it
[07:13:00] <g4lt-sb100> dlg, you have no idea how tempting that is....
[07:13:16] <wesolows> what's on that thing, a Cyrix 5x86 chip?
[07:13:27] <dlg> g4lt-sb100: just do it
[07:13:34] <dlg> whats the worst taht can happen?
[07:13:38] <g4lt-sb100> but the new open sun forgot about trivialities that would b really cool to make work other places like that
[07:13:49] <dlg> "no, we dont own the doco, so we cant open it"
[07:13:50] <g4lt-sb100> wesolows, mine's a 2, so it's a celeron 700
[07:14:01] <wesolows> ooooh, celeron 700
[07:14:12] <wesolows> it's a cray on a card, no doubt
[07:14:30] <g4lt-sb100> wesolows, it's about fair for any task I wish to throw at windows
[07:14:41] <g4lt-sb100> unitl brandwin comes, that is ;P
[07:15:26] <wesolows> get cracking then :-)
[07:15:33] <Tempt> Yes, it would be truly great if there was only one OS without ZFS.
[07:15:35] <wesolows> I'd call it brandlose though
[07:15:42] <g4lt-sb100> that, and I can say I AM using windows with zfs ;P
[07:15:45] <wesolows> win is not an appropriate abbreviation for windows
[07:16:06] <Tempt> In my recent experience, Windows is more stable than PenguinOS
[07:16:30] * rootard thinks windos is an appropriate abbrev.
[07:16:35] <wesolows> I've never used PenguinOS; is that some GNU/Linux distribution?
[07:16:46] <wesolows> rootard: indeed
[07:16:53] <wesolows> or even vmsdows
[07:17:01] <wesolows> or vmsdos, perhaps best of all
[07:17:06] <Tempt> Hey
[07:17:13] <Tempt> What did VMS do to you?
[07:17:38] <wesolows> probably served me well indirectly, running industrial automation, banking, and various other stuff for the past 2 decades
[07:17:42] <g4lt-sb100> dave cutler inappropriately touched him as a child
[07:17:47] <wesolows> I've never used it but have nothing against it
[07:17:53] <Tempt> with both fists, for a remark like that.
[07:18:12] <Tempt> Honestly, the Windows jokes were almost at an end before Vista came out.
[07:18:17] <rootard> I thought parts/all of VMS was bought by MS in order to integrate into NT. The version I heard was that M$ couldn't figure enough of the code out to use the better parts.
[07:18:18] <wesolows> it's no secret though that most of the NT design team was the ex VMS team
[07:18:24] <Tempt> I've had XP easily outlast a linux install.
[07:18:35] <g4lt-sb100> Tempt, yeah, thank god for microsoft innovation ;P
[07:18:36] <wesolows> sadly they seemed determined to "improve" on vms and in the process came up with something really awful
[07:18:44] <Tempt> No, Microsoft did not buy any VMS code. They hired VMS development staff.
[07:18:47] <wesolows> right
[07:18:53] <wesolows> that's what I said...isn't it?
[07:19:07] <wesolows> oh, nm
[07:19:12] <rootard> Ahh, k. Thanks for setting that one straight. I know my history is skewed in this area.
[07:19:19] <Tempt> If Microsoft had ditched Vista and instead released some shine-ups and patches for XP, there wouldn't be much reason to use Linux on a peecee anymore
[07:19:28] <wesolows> there was before?
[07:19:35] <Tempt> wesolows: Games.
[07:19:38] <Tempt> wesolows: Graphics
[07:19:43] <Tempt> wesolows: Sound
[07:19:49] <rootard> apt
[07:19:50] <g4lt-sb100> right, but by introducing vista, they opened themselves to an entire new generation of windows jokes
[07:19:55] <wesolows> though tbh, I used both GNU/Linux and XP at a job long ago, and I could never recommend XP over GNU/Linux
[07:20:04] <wesolows> all of which work fine for me on Solaris, actually
[07:20:35] <rootard> It's not as if they are interchangeable... I would never walk up to a windows user and give them a linux box.
[07:20:51] <dlg> you'd push them over and run away?
[07:20:52] * wesolows really hopes this thing works so he can putback and go home
[07:21:07] <wesolows> yay
[07:21:08] <rootard> dlg: exaclty
[07:21:18] <rootard> *exactly
[07:21:39] <Tempt> I don't consider fiddling around with WINE and crap like that to be a reasonable solution for running software
[07:21:44] * g4lt-sb100 looks crosseyed at the bridge, just to mess with wesolows
[07:22:00] <Tempt> It's a messy hack, and while it might be fine for the occasional piddling application, it isn't appropriate for anything large.
[07:22:10] <dlg> g4lt-sb100: asked for the sunpci docs yet?
[07:22:17] <g4lt-sb100> Tempt, it beats WABI ;P
[07:22:31] <dlg> anyone here useing ldoms on t2k gear?
[07:22:34] <dlg> or using
[07:22:35] * g4lt-sb100 has no idea of who to ask ven
[07:22:45] <wesolows> amen to that - wine--
[07:22:54] <dlg> g4lt-sb100: http://wiki.sun.com/display/FOSSdocs/Home
[07:23:11] <steleman> heh. Google Picasa on Linux contains a statically linked Wine in it.
[07:24:32] <CIA-25> wh31274: 6605138 OPL prtdiag missing memory config details
[07:25:29] *** johnlev has joined #opensolaris
[07:26:59] *** halton has joined #opensolaris
[07:27:28] *** trygvis_ is now known as trygvis
[07:31:30] *** gtc_laptop has joined #opensolaris
[07:41:25] *** dlynes_laptop has joined #opensolaris
[07:46:59] *** gtc_laptop has quit IRC
[07:53:16] *** Gekkko has quit IRC
[07:53:16] <johnlev> the link to the newer compilers is now there:
[07:53:17] <johnlev> http://www.opensolaris.org/os/community/tools/sun_studio_tools/
[08:00:19] *** stelle has quit IRC
[08:01:20] *** Dar_HOME is now known as Dar
[08:03:30] <CIA-25> achartre: FWARC 2007/122 VIO MD node updates, FWARC 2007/363 VIO vDisk MD node updates, 6514091 vDisk server should export volumes as full disks, 6528156 opening devices exclusively from vds causes several problems, 6591399 vds prints file lookup error during service domain boot
[08:03:31] <CIA-25> yy150190: 6480448 Sunvts netlbtest external loopback failed on 82546 chipset, 6521984 Sunvts netlbtest failed on Northstar QGE cards, 6531842 e1000g link goes down and up frequently on T2000 systems, 6540535 Sunvts netlbtest internal loopback failed on ophir devices, 6579605 Sunvts netlbtest internal loopback failed on Northstar fiber card, 6595668 e1000g postinstall script missed 5 device ids which had been supported by shared code
[08:04:41] <e^ipi> uhh, cool
[08:04:44] <e^ipi> I guess
[08:11:58] *** Gekkko has joined #opensolaris
[08:15:56] *** wayg has joined #opensolaris
[08:15:57] <wayg>  http://st-pitch.miniville.fr/
[08:15:59] -wayg-  http://st-pitch.miniville.fr/
[08:16:02] <wayg> 
URL http://st-pitch.miniville.fr/

[08:16:03] *** wayg has left #opensolaris
[08:17:58] *** dunc has quit IRC
[08:19:11] <Tempt> Damn spammers.
[08:20:41] *** estibi has quit IRC
[08:21:25] *** pizdec has quit IRC
[08:25:44] *** halton has quit IRC
[08:33:01] <WickedWicky> *yawn*
[08:33:28] *** JWheeler has quit IRC
[08:35:40] <madhatter> WickedWicky: here, too
[08:36:16] <WickedWicky> I seriously need some caffeine in my system
[08:43:50] *** tsoome has quit IRC
[08:45:14] <madhatter> COffeemaschine is broken since almost two weeks now. I have to stick to black tea.
[08:46:44] *** FBdev has joined #opensolaris
[08:47:37] <e^ipi> so brew some coffee
[08:47:37] <e^ipi> takes 5 minutes
[08:47:37] <e^ipi> 30 seconds if you've got an espresso machine
[08:47:37] <e^ipi> less if you don't know how to operate the espresso machine
[08:48:50] <e^ipi> get a press pot
[08:49:30] *** Cyrille has joined #opensolaris
[08:51:13] *** nrubsig has quit IRC
[08:54:18] *** mikefut has joined #opensolaris
[08:55:47] *** cypromis has joined #opensolaris
[09:00:57] *** bengtf_ has quit IRC
[09:01:43] *** alobbs has joined #opensolaris
[09:02:02] *** tsoome has joined #opensolaris
[09:11:26] <e^ipi> I wonder why network printing is taking so long
[09:11:57] <WickedWicky> routing on the printer is setup ok?
[09:12:03] <e^ipi> *shrug*
[09:12:15] <WickedWicky> I assume it's a network printer :P
[09:12:24] <e^ipi> the print server is also the NFS server
[09:12:29] <e^ipi> NFS works flawlessly
[09:12:44] <e^ipi> ssh-ing in to the server and using lpr prints pretty close to immediately
[09:13:16] <WickedWicky> I only used network printing once.. via samba... and it took it's fair time to do the spooling
[09:13:39] <e^ipi> but printing on either the mac or the sparc takes a couple minutes to get to the spool
[09:13:50] <e^ipi> meh, to hell with samba, i just use straight-up CUPS
[09:14:04] <e^ipi> why bother going through MS's braindead protocol when I've got UNIX
[09:14:31] <WickedWicky> true
[09:14:51] <WickedWicky> unfortunatly our braindead printserver at work is a braindead mickeysoft server
[09:16:25] <WickedWicky> be right back, gonna get that coffee
[09:16:28] *** yongsun has quit IRC
[09:18:41] * e^ipi doesn't understand why one would want to run windows on a server
[09:21:59] <trochej> e^ipi: Graphical environment and centralized management
[09:22:40] <trochej> e^ipi: Oh, and when workstations are Windows, it has some advantages
[09:25:15] *** cydork has joined #opensolaris
[09:25:40] *** chrisso has quit IRC
[09:35:46] *** cypromis has quit IRC
[09:36:28] *** seanmcg has quit IRC
[09:36:49] *** cypromis has joined #opensolaris
[09:46:50] *** dunc_ has joined #opensolaris
[09:50:33] *** _0 has joined #opensolaris
[09:50:35] <_0> http://www.pennergame.de/ref.php?refid=1942911
[09:50:36] *** _0 has left #opensolaris
[09:50:50] <trygvis> wtf
[09:53:01] *** cmang has joined #opensolaris
[09:54:12] <e^ipi> spammergrams
[09:54:29] <e^ipi> freenode has them fairly regularly
[09:55:21] *** Erwann has joined #opensolaris
[09:58:06] *** dennis has quit IRC
[10:00:13] *** dmarker has quit IRC
[10:00:23] *** dmarker has joined #opensolaris
[10:01:39] *** seanmcg has joined #opensolaris
[10:01:39] *** estibi has joined #opensolaris
[10:07:04] *** cmang__ has quit IRC
[10:07:15] <Gekkko> Does Solaris use glibc?
[10:08:07] <estibi> Gekkko: no
[10:08:16] <Gekkko> what does Solaris use?
[10:08:25] <seanmcg> libc :)
[10:08:36] <Gekkko> lol.
[10:09:20] <seanmcg> though really, solaris has its own libc.  Whats your query on it ?
[10:10:53] <Gekkko> I'm arguing with Linux people about the best alternative to glibc.
[10:11:13] *** peteh has joined #opensolaris
[10:11:40] <palowoda> Where did Linux people come up with the idea about glibc?
[10:12:41] <Gekkko> GNU.
[10:13:18] *** cypromis has quit IRC
[10:13:40] *** dlynes_ has joined #opensolaris
[10:13:47] <palowoda> GNU is not an idea.
[10:14:43] <trochej> palowoda: It wasn't Linux people, it was GNU people
[10:14:57] <trochej> palowoda: Remember that Linux os younger than GNU and probably glibc
[10:16:25] <trochej> glibc was initially written mostly by Roland McGrath, working for the FSF in the 1980s
[10:16:45] <trochej> In February 1988, FSF described glibc as having nearly completed the functionality required by ANSI C.[1] By 1992, it had the ANSI C-1989 and POSIX.1-1990 functions implemented and work was under way on POSIX.2.
[10:16:48] *** JWheeler has joined #opensolaris
[10:16:54] <trochej> Which means, it wasn't Linux people idea
[10:16:56] <trochej> Sorry :)
[10:19:26] <palowoda> I'm sure FSF and Roland McGrath had some sort of idea the purpose of glibc in the beginning.  Why it's better may not have anything to do with software.
[10:20:53] *** KermitTheFragger has joined #opensolaris
[10:21:41] 
[10:21:48] *** JohnMahowald has quit IRC
[10:23:53] <Cyrille> and that no two versions are ever compatible ;-)
[10:25:03] 
[10:25:50] <palowoda> So what is glibc currently "better" at?
[10:26:33] <PerterB> "everything" if you drink the kool ade
[10:26:36] <trygvis> it works on a whole bunch of platforms
[10:27:12] *** rcecil has joined #opensolaris
[10:27:20] <palowoda> I always thought a platform is made of wood.
[10:27:44] <rcecil> hello I was trying to print the value of a float using Dtrace
[10:27:47] <rcecil> with no luck
[10:28:30] 
[10:29:35] <palowoda> trochej: No no no, I buy zealots at Walmart.
[10:29:40] <dlg> trochej: zealots are everywhere
[10:29:47] *** dlynes_laptop has quit IRC
[10:29:54] *** rcecil has quit IRC
[10:30:05] <trochej> dlg: I know, I was being ironic. :)
[10:30:28] <dlg> oh yeah
[10:30:39] *** rcecil has joined #opensolaris
[10:31:51] <palowoda> So now we have the advantage of glibc as an innovation to be bought at Walmart in zealots for lots of platforms. :)
[10:36:05] <trochej> Khe
[10:38:38] <trochej> Hm
[10:39:05] <trochej> I had to install a third party driver for my nic but it doesn't plug at boot. Do I need to perhaps add a line to /etc/Path_to_inst?
[10:39:48] <seanmcg> trochej: is the driver correctly in /etc/driver_aliases ?
[10:40:32] *** estibi has quit IRC
[10:42:42] *** cypromis has joined #opensolaris
[10:43:13] <palowoda> I'm sure it's ok to expose who the third party is also.
[10:43:26] <trochej> palowoda: Broadcom, but does it matter? :)
[10:43:45] <trochej> seanmcg: I can see it here, but I'm not sure theese pci... parts
[10:43:58] <palowoda> I thought the Broadcom drivers where already in opensolaris?
[10:44:05] <trochej> palowoda: It's Sol10
[10:44:12] *** migi has joined #opensolaris
[10:44:13] <palowoda> upgrade
[10:44:15] <trochej> palowoda: And the one supplied with system dodn't work
[10:44:20] <dlg> theres a new gen of broadcoms that arent in solaris
[10:44:44] <palowoda> One wonders why broadcom isn't part of opensolaris.
[10:45:14] <dlg> you also have to realise that the previous generation (the bge ones) are very quirky between revisions
[10:45:17] <rcecil> Any one tried to print a floating point number in dtrace? I have a script that goes like this dtrace -n 'pid4490::*VDoubleEv:return { printf("%f",(double)arg1) } '
[10:45:19] <trochej> Two wonder when it may be :)
[10:45:26] <dlg> so updates for new chips and such arent as easy as adding new pci ids
[10:45:33] <dlg> keeping on top of it is hard
[10:45:40] <seanmcg> trochej the pci.. parts like: pci14e4,167d mean: pci<vendor-id>,<product-id>  you can use /usr/X11/bin/scanpci and prtconf -vp to find them for your card
[10:45:52] <palowoda> With Broadcom it's hard as Broadcom makes it.
[10:46:17] <trochej> seanmcg:
[10:46:20] <trochej> seanmcg: thnx
[10:46:24] <seanmcg> rcecil: floating points don't exists within the kernel (where dtrace runs) so..
[10:46:46] <rcecil> so how do u convert it
[10:47:04] <seanmcg> rcecil: google or search opensolaris.org for an answer, this has been answered on the dtrace mailing list
[10:48:02] <rcecil> can you point me where. I exhausted google and the mailing list.. I seen a question being posted on the open solaris but no answer
[10:49:26] <trochej> Uhm
[10:49:31] <trochej> It is there
[10:50:04] <trochej> seanmcg: Is it enough for system to plug a nic if I create /etc/dhcp.bcme0 ?
[10:50:22] <trochej> considering that it plumbs as bcme0 of course
[10:50:31] <palowoda> and run devfsadm if the device path wasn't created yet.
[10:51:02] <seanmcg> can you 'ifconfig bcme0 plumb' and see if it comes up ?  if it does, then /etc/dhcp.bcme0 will work.
[10:52:15] <seanmcg> btw, what is the entry in your /etc/driver_aliases for the bcme ?
[10:52:17] <trochej> seanmcg: Yup, ifconfig bcme0 plumb works
[10:52:49] <trochej> seanmcg: bcme "pci14e4,167b"
[10:52:51] <seanmcg> then 'ifconfig bcme0 dhcp' may get it an ip address..
[10:53:00] <trochej> seanmcg: It does
[10:53:17] <trochej> however touching /etc/dhcp.bcme0 doesn't work?
[10:53:18] <trochej> Hmm
[10:53:25] <trochej> Should I check svcs for something?
[10:53:44] <seanmcg> well touching it _and_ restarting the network/physical service(s)..  or reboot..
[10:53:57] <trochej> seanmcg: Did it
[10:54:02] <trochej> A few times already
[10:54:09] <seanmcg> you could upgrade and use network automagic.. :)
[10:54:37] <seanmcg> make sure you don't have a hostname.bcme0 too..
[10:54:44] <trochej> seanmcg: I think that I just may
[10:54:59] <trochej> I don't
[10:55:00] *** halton has joined #opensolaris
[10:55:03] <seanmcg> nwam is rather nice :)
[10:55:20] <trochej> It's a clean system, just isntalled
[10:55:43] <palowoda> It's a clean S10 system though.
[10:55:58] <seanmcg> s10 what ?  s10u4 aka 08/07 ?
[10:56:09] <trochej> u4, yes
[10:56:20] <palowoda> u4 has automagic?
[10:56:49] *** rcecil has quit IRC
[10:57:03] *** Cyrille has quit IRC
[10:58:31] *** enzo has joined #opensolaris
[10:58:40] <seanmcg> umm noo, just querying the version of s10..
[10:59:48] *** enzo has left #opensolaris
[11:00:35] *** Cyrille has joined #opensolaris
[11:02:11] <palowoda> Always good to know what is missing from S10.
[11:03:32] *** xuewei has joined #opensolaris
[11:05:20] <palowoda> trochej: So was it necessary to add the third party driver?
[11:06:49] <trochej> palowoda: Yes,
[11:07:42] <palowoda> Interesting they have that id in build 72.  So I guess it's an S10 issue.
[11:08:39] *** timsf has joined #opensolaris
[11:08:54] <timsf> morning all
[11:09:01] <Gman> hey tim
[11:09:59] <timsf> Hi Gman
[11:11:40] *** estibi has joined #opensolaris
[11:12:59] *** Cyrille has quit IRC
[11:13:28] <palowoda> Oh crap Steve L is leaving Sun.
[11:13:33] <trochej> Yup
[11:13:47] <trochej> Seen that post
[11:14:38] *** cmihai has joined #OpenSolaris
[11:18:03] <palowoda> Oh man, looks like he is headed for Songbird.  I used that app.  A little slow on performance but a great app.
[11:21:25] *** Gekkko has quit IRC
[11:21:26] *** cmihai_ has joined #OpenSolaris
[11:21:54] *** cmihai_ has left #OpenSolaris
[11:22:07] *** cmihai has quit IRC
[11:24:44] *** cmihai has joined #OpenSolaris
[11:24:47] <palowoda> What kind of sucks is Songbird is a java app which doesn't really have a way to support install methods on Solaris.
[11:25:59] <Gman> i think it's actually mozilla
[11:26:34] *** bnitz has joined #opensolaris
[11:26:39] <palowoda> Gman: You mean the Songbird install program?
[11:27:16] <Gman> http://www.songbirdnest.com/support/#sysreq
[11:27:25] <WickedWicky> is there a way I can track back an TCP port to the process it belongs to?
[11:27:46] <palowoda> lsof
[11:27:50] <WickedWicky> cheers
[11:27:50] <cmihai> netstat, pfiles, lsof
[11:28:09] *** Jagoma has joined #opensolaris
[11:29:27] <palowoda> Hmm netstat what option?
[11:29:48] *** Cyrille has joined #opensolaris
[11:30:18] *** derchris has quit IRC
[11:30:27] *** derchris has joined #opensolaris
[11:31:04] <WickedWicky> I dont think netstat on solaris has an option to show which PID links to what port
[11:31:34] <Cyrille> isn't that what lsof is for?
[11:31:39] <cmihai> Actually, no.
[11:31:56] <cmihai> You're better off with pfiles, since in Solaris there is no native lsof, it's a port of the GNU / linux tool.
[11:32:16] <cmihai> So it doesn't properly show ports.
[11:32:28] <WickedWicky> pfiles seems to take a PID as an argument
[11:32:31] <WickedWicky> which is the problem
[11:32:36] <WickedWicky> I know the TCP port, not the PID
[11:33:21] <cmihai> Well, you could poll all PIDs for ports, and grep for PIDs :P
[11:33:44] *** Gman has left #opensolaris
[11:33:51] *** Gman has joined #opensolaris
[11:33:56] *** Gekkko[PDA] is now known as Gekkko[Arch]
[11:33:57] <palowoda> Sounds too political. :)
[11:34:28] *** yongsun has joined #opensolaris
[11:34:44] <cmihai> You want something like "lsof -i 4:port" right? Well, try it on Solaris, but last time I checked it was still broken,.
[11:36:03] <palowoda> Ports for PID broken broken.
[11:36:26] <cmihai> palowoda, it's not really that political, lsof was never really _that_ portable. That's why BSD for example has fstat and Solaris has pfiles and fuser.\
[11:37:28] <WickedWicky> for x in `ps -ef | awk ' { print $2 } '`; do echo PID: $x ; pfiles $x | grep "port: <port>"; done
[11:37:36] <WickedWicky> I know it looks ugly, but it works ;-)
[11:37:37] <WickedWicky> thanks guys
[11:38:01] <cmihai> WickedWicky, yeah, that was pretty much the idea ;-)
[11:38:21] <palowoda> It's now documented.
[11:38:26] <cmihai> It is? Where
[11:38:29] <cmihai> bugid?
[11:38:38] <palowoda> on opensolaris irc. :)
[11:38:43] <cmihai> Gee ;-)
[11:39:05] <cmihai> Eh, better put this in my blog :P
[11:39:06] *** Vorbis_w48 has quit IRC
[11:39:11] <WickedWicky> I am blogging it too
[11:39:16] <WickedWicky> on purplecow and my own
[11:39:32] <cmihai> WickedWicky, mind if I just copy/paste the script?
[11:39:34] <palowoda> You will be famous on google.
[11:39:45] <WickedWicky> of course not
[11:40:05] <WickedWicky> the reason why I hang out here is to collect/give knowledge/experience
[11:41:43] <palowoda> Give that man a brand new core2 duo laptop.
[11:42:39] <palowoda> I'd say ultrasparc latptop but there is a shortage of those for the military.
[11:44:42] <trochej> core 2 duo do quite nice
[11:44:53] <cmihai> CRAP
[11:45:01] <cmihai> I just lost my post
[11:45:03] <cmihai> Damn google.
[11:45:24] <timsf> Here's an alternative: ps -e -o pid | xargs -t -n1 pfiles | grep "port: 23"
[11:45:31] <timsf> not sure it's much better to be honest :-/
[11:45:42] <cmihai> Same thing.
[11:45:52] *** Jagoma has quit IRC
[11:46:04] <timsf> Similar, yeah.
[11:49:24] *** Jagoma has joined #opensolaris
[11:57:30] <cmihai> Oy guys :-)
[11:57:32] <cmihai> http://www.unix.ms/pcp/ - check this out
[11:57:43] <cmihai> PCP is a script that enables administrators to see what open TCP ports are in use on a Solaris system. It maps ports to PIDs and vice versa. It accepts wildcards and will also show at a glance all open ports and their corresponding PIDs. I wrote it because I couldn't install lsof where I was working, and "lsof -i" is a great addition to any UNIX admin's toolbox. PCP is a half-decent alternative.
[11:57:58] <WickedWicky> aweet
[11:58:13] <WickedWicky> sweet too
[11:59:41] <WickedWicky> he's basicly doing what we're doing with the one lines
[11:59:50] <WickedWicky> for proc in `ptree -a | grep -v ptree | awk '{print $1};'`
[11:59:50] <WickedWicky> do
[11:59:50] <WickedWicky> result=`pfiles $proc 2> /dev/null| grep "port: $port"`
[11:59:51] <cmihai> Well, yes :-)
[12:00:00] <timsf> and no indents
[12:00:05] <timsf> or case statements
[12:00:59] <cmihai> It still uses the same concept, it's just a bit more... worked, and you just drop the script in bin :-)
[12:01:12] <WickedWicky> yeah
[12:01:30] <palowoda> worth a bookmark.
[12:01:35] <WickedWicky> done that
[12:01:35] <WickedWicky> :D
[12:04:22] *** simford has quit IRC
[12:07:25] *** Gekkko[Arch] is now known as Gekkko[PDA]
[12:08:40] *** dennis has joined #opensolaris
[12:15:50] *** cypromis has quit IRC
[12:19:53] *** halton has quit IRC
[12:21:23] *** yongsun has quit IRC
[12:25:25] *** Gman is now known as GmanAFK
[12:28:55] *** FBdev has quit IRC
[12:29:10] *** Cass has joined #opensolaris
[12:32:08] <flyingparchment> hmm, anyone used infortrend "EonPath"?
[12:32:18] <flyingparchment> apparently they don't support MPxIO, but they have their own software that runs on solaris..
[12:43:23] *** Cass has quit IRC
[12:47:14] *** Gekkko[PDA] has quit IRC
[12:56:00] *** cmihai has quit IRC
[12:59:41] *** h3sp4wn has quit IRC
[13:04:17] *** FBdev has joined #opensolaris
[13:04:56] *** FBdev has quit IRC
[13:06:11] <jmcp> flyingparchment: I've never heard of it. I've heard of and used mpxio, vxdmp and powerpath.
[13:06:12] *** FBdev has joined #opensolaris
[13:06:32] <flyingparchment> they don't do dmp either
[13:06:42] * jmcp strikes it off his list
[13:07:31] * jmcp doesn't understand the point in arrays that aren't active-active, or at least fully compliant with T10 ALUA
[13:08:07] <dlg> cos vendors are so much smarter than that
[13:08:10] <flyingparchment> it does a/a with their own proprietary software, but i'm a little dubious about that :)
[13:08:11] *** deather_ has joined #opensolaris
[13:08:22] <jmcp> dlg: " we don't need no stinkin' industry standards!"
[13:08:27] <flyingparchment> (their array firmware is bad enough, the idea of putting something into the kernel is scary)
[13:08:28] <dlg> exactly
[13:08:57] * jmcp growls @ various large storage vendors .....
[13:09:11] <dlg> :)
[13:09:16] * dlg growls at his boss
[13:09:25] <jmcp> I don't think he's all that large
[13:09:58] *** mritun_ has joined #opensolaris
[13:10:04] <dlg> you saying i should take him out?
[13:10:25] <jmcp> only to the pub
[13:10:29] <jmcp> not in any threatening manner
[13:10:36] <dlg> "DRINK OR DIE"
[13:10:37] <jmcp> mmmmmm staff club beers ......
[13:10:38] <jmcp> heh
[13:11:03] *** Jagoma__ has joined #opensolaris
[13:11:14] <mritun_> whoa
[13:11:31] <mritun_> that's some conspiracy going on :P
[13:11:54] <jmcp> mritun_: what is? the plot to sell Belgium on EBay? http://www.smh.com.au/news/web/ebay-halts-sale-of-belgium/2007/09/18/1189881479834.html
[13:12:50] <mritun_> wow
[13:13:02] <mritun_> any idea about the reserve on that ?
[13:13:15] <jmcp> 10million Euro, I think
[13:13:29] <jmcp> came with a warning "has 300million Euro in debt" or something
[13:14:27] <mritun_> aha
[13:14:43] <mritun_> but I was more interested in that "drink or die" thing
[13:14:59] <jmcp> oh
[13:15:05] <trochej> There was an auction to sell Polish Prime Minister. Before it was taken down the price hit $85k
[13:15:20] <quasi> jmcp: would almost be worth it if one could "confiscate" all the beer before selling it on to someone else
[13:15:26] <jmcp> mritun_: then you'd be interested in the March of the Engineers - http://www.youtube.com/watch?v=lpe3KpWjCmA
[13:15:45] <mritun_> what's the world coming to
[13:16:12] <mritun_> there's another ridiculous thing happening here in India too
[13:17:49] <mritun_> around 5000 yrs ago, the epic of Ramayana was written with Rama portrayed as God whose army constructs a bridge to Lanka to get his wife back
[13:18:18] *** nostoi has joined #opensolaris
[13:18:21] <mritun_> fast forward 5000 yrs and some NASA pictures depict some formation going from India to Sri Lanka
[13:18:34] <mritun_> now people are up in arms :(
[13:18:41] *** linux_user400354 has joined #opensolaris
[13:19:16] <mritun_> ans supreme court is supposed to give a verdict whether Lord Rama made that cridge or not :P
[13:19:32] *** Cass has joined #opensolaris
[13:19:34] <mritun_> ans -> and, cridge -> bridge
[13:19:57] *** karrotx has quit IRC
[13:25:47] *** Jagoma has quit IRC
[13:26:45] *** FelipeBare has joined #opensolaris
[13:27:29] *** deather has quit IRC
[13:28:14] *** Jagoma__ has quit IRC
[13:28:37] <JWheeler> LOL, awesome
[13:29:08] <JWheeler> If it makes you feel any better, apparently New Zealand was a giant fish, that a maori warior fished up out of the ocean...
[13:29:13] <JWheeler> who knew!
[13:30:04] <Cyrille> but are there any pictures of the fish's descendants?
[13:30:05] *** fbo has joined #opensolaris
[13:33:16] *** deather_ is now known as deather
[13:34:27] <trochej> Wait for kaiwai, he can post some pics of himself :)
[13:35:36] <jmcp> wow
[13:36:04] *** cypromis has joined #opensolaris
[13:36:13] <jmcp> mritun_: what are the consequences likely to be if the Indian Supreme Court decides that Lord Rama didn't build that bridge?
[13:36:32] <jmcp> I imagine there might be some very annoyed people
[13:38:47] <Cyrille> the people planning on selling that bridge? ;-)
[13:40:36] <mritun_> hehe
[13:41:05] <mritun_> I don't think supreme court would give any verdict.. the case might just "drag on"
[13:41:17] <mritun_> that would be in everyone's interest
[13:41:37] <Cyrille> "Built by Rama himself! Direct connection from India to Sri Lanka! Only 124574523 gazillion rupees and that's cutting me own throat!"
[13:41:56] *** JohnMahowald has joined #opensolaris
[13:42:16] <Cyrille> I'm actually surprised the supreme court would feel it has/be given authority to decide on such a matter.
[13:42:35] <mritun_> well... I pity the judges :P
[13:43:09] <mritun_> they probably didn't want this... it's been pretty much a noose tightened around their neck
[13:45:13] *** Snake007uk has joined #opensolaris
[13:51:02] *** Vorbis_w48 has joined #opensolaris
[14:00:24] *** linux_user400354 has quit IRC
[14:02:55] *** linux_user400354 has joined #opensolaris
[14:08:49] *** Giaco has joined #opensolaris
[14:08:49] *** SYS64738 has quit IRC
[14:09:28] *** Doc has joined #opensolaris
[14:14:33] *** master_o1_master has joined #opensolaris
[14:19:12] *** Pietro_S has joined #opensolaris
[14:20:01] *** dpn` has quit IRC
[14:25:35] *** master_of_master has quit IRC
[14:27:30] *** infidel has joined #opensolaris
[14:28:02] *** mikefut has left #opensolaris
[14:35:14] *** linux_user400354 has quit IRC
[14:36:28] *** linux_user400354 has joined #opensolaris
[14:40:25] *** victori_ has quit IRC
[14:45:08] *** nostoi has quit IRC
[14:46:25] *** ashner has quit IRC
[14:47:00] <infidel> as root one one box can you rlogin to another box and when it prompts you for a password you type in the root password for the remote box, what will happen?
[14:47:07] <infidel> on^
[14:47:37] <trygvis> dunno about rlogin, but with ssh root login over ssh is usually disabled
[14:48:28] <trochej> infidel: Depends on configuration of remote box.
[14:50:49] <trochej> infidel: In Solaris 10 u4 it's impossible out of box
[14:51:02] <trochej> default is to disallow
[14:51:27] <trochej> Which is, as I believe, quite sane
[14:51:28] <infidel> pretty much, you have to rlogin as a user
[14:51:41] <seanmcg> typically with rlogin/rsh as well, root won't be able to login since by default root is only allowed to login via the console, see /etc/defaults/login.
[14:52:00] <seanmcg> better off this way..
[14:52:03] <infidel> seanmcg: thank you
[14:53:08] <infidel> seanmcg: had a guy today that swore the root password was changed because as root he tried to rlogin to a remote box and he typed in the remote boxes root password
[14:54:54] *** eboutilier has joined #opensolaris
[14:59:58] *** suhail has joined #opensolaris
[15:00:15] <suhail> hi
[15:00:50] <seanmcg> infidel: well he could think that since the remote box would be accepting a passwd, but just not telling him that root canna login via rlogin/rsh.  Mind you he should have got a message saying 'Not on system console' regardless of the passwd entered.
[15:00:50] <suhail> hello, is there anyone?
[15:01:02] *** Gekkko[PDA] has joined #opensolaris
[15:01:08] <trygvis> no
[15:01:40] <infidel> seanmcg: he shouldn't even have root in my opinion
[15:02:34] <suhail> hey guys, i am new to solaris. i am currently planning to install solaris on my computer. i want to know how much ram is required to run solaris with gui.
[15:02:51] <Gekkko[PDA]> 512mb+
[15:02:54] <Gekkko[PDA]> ?
[15:02:55] <infidel> suhail: how much do you have
[15:03:02] <trygvis> I think the latest installer requires 768
[15:03:10] <suhail> currently 256MB
[15:03:43] <Gekkko[PDA]> try Nexenta?
[15:03:48] <infidel> suhail: run slackware
[15:04:05] <Gekkko[PDA]> lol infidel no.
[15:04:05] <suhail> i think i should upgrade my ram.
[15:04:14] <Gekkko[PDA]> suhail: yep.
[15:04:27] <suhail> what is nexenta?
[15:04:33] <Gekkko[PDA]> get another 64 gig :p
[15:04:42] <Cyrille> an opensolaris distribution
[15:04:44] <Gekkko[PDA]> google it.
[15:05:23] <suhail> hmm.. are there so many distributions like linux?which one is right for me?
[15:05:42] <Gekkko[PDA]> >.>
[15:05:49] <infidel> xp?
[15:05:58] <suhail> solaris xp?
[15:06:03] <Gekkko[PDA]> lol infidel you arent helping.
[15:06:08] <Gekkko[PDA]> ...
[15:06:09] <infidel> sorry
[15:06:10] <Gekkko[PDA]> lol.
[15:06:24] <Gekkko[PDA]> suhail: ubuntui
[15:06:27] <Gekkko[PDA]> suhail: ubuntu*
[15:06:35] <Gekkko[PDA]> give it a try.
[15:07:34] <suhail> ubuntu is linux distribution.i already tried it.you know it sometimes crashes when i tried to update it.i think solaris would be more stable.
[15:07:53] <Gekkko[PDA]> infidel:
[15:08:12] <infidel> suhail: was it working fine before you tried to update it?
[15:08:24] <Gekkko[PDA]> suhail: solaris isn't really a desktop OS.
[15:09:23] <infidel> hmmm
[15:09:29] <suhail> ya.before updating ubuntu was working fine infidel.but after upgrade the X server crashed.
[15:09:35] <seanmcg> Gekkko[PDA]: its not ?  Ok, I'll stop drawing posters, using xmms, running wine and playing with the gimp with it then :)
[15:09:49] <infidel> suhail: then why did you update?
[15:10:01] <Pietro_S> well, depend how you define desktop machine ... I use sxce as developer/desktop machine and it's nice
[15:10:41] <suhail> so Gekkko solaris is for server use not for desktop use?
[15:10:41] <Gekkko[PDA]> seanmcg: stop ruining my argument
[15:11:06] <infidel> linux is a desktop machine solaris is either a workstation or a server
[15:11:13] <Gekkko[PDA]> suhail: it can be, but not out-of-the-box
[15:11:26] <Gekkko[PDA]> it requires configuring
[15:11:42] *** Santtu has quit IRC
[15:11:45] <suhail> hmm. means linux is more suited for destkop use.
[15:11:57] <infidel> i think so
[15:12:28] <Gekkko[PDA]> yes.
[15:12:29] * quasi is perfectly happy with solaris as a desktop as well (although all I want is a window manager and an mp3 player ;)
[15:12:46] <suhail> i have heard that solaris and linux are very much similar to each other.is it true Gekkko?
[15:12:51] <Gekkko[PDA]> quasi: and you had to install blastwaves stuff
[15:13:04] <Gekkko[PDA]> suhail: depends
[15:13:07] <quasi> Gekkko[PDA]: no, absolutely not
[15:13:24] <Gekkko[PDA]> quasi: selfcompiled?
[15:13:30] <Pietro_S> Gekkko[PDA]: SFE
[15:13:31] <quasi> Gekkko[PDA]: I'm not having that mess anywhere near my box
[15:13:35] <quasi> es
[15:13:38] <Gekkko[PDA]> suhail: both Unixes
[15:13:43] <Pietro_S> or pmpkg and others ...
[15:14:06] <quasi> Gekkko[PDA]: seriously, browser, xmms and whatever came with Nevada
[15:14:13] <quasi> Gekkko[PDA]: is more than enough
[15:14:49] <Gekkko[PDA]> but why Solaris then?
[15:15:06] <quasi> because it is the best os there is
[15:15:28] <Gekkko[PDA]> not a valid answer.
[15:15:29] <suhail> i think solaris would be more stable than linux.
[15:15:34] <infidel> quasi: for you purpose
[15:16:13] <mritun_> suhail: if stability is an issue for you, I'd suggest checking the harware first
[15:16:19] <quasi> Gekkko[PDA]: well, what can I say - linux is useless for HA services (I wouldn't trust it as far as I could throw it)
[15:16:24] <mritun_> eg. check the ram using memtest86+
[15:16:34] <infidel> suhail: either one are more stable than windows but linux has better support at the moment
[15:16:52] <mritun_> stability isn't a day to day concern anymore... whether you're using windows/linux/solaris these days
[15:17:09] * quasi wanders off, back to work - I have a "date" with an E6900
[15:18:17] <Gekkko[PDA]> quasi: I don't know what a HA service vis but it se3urely isn't somethinv a de3sktop user would use.
[15:18:25] <suhail> unix variants are difficult to adminstrate because one has to learn the command line.however they are technically more stable and secure than windows.
[15:18:34] <bda> http://biz.yahoo.com/prnews/070912/aqw112.html?.v=17 # O_o
[15:18:47] <quasi> mritun_: maybe not for a small play setup - when you're rolling millions of "financial transactions" every minute/hour, it matters a whole lot ... not to mention I'd be hard pressed to find a lunix box with 96G mem and 12 procs
[15:18:54] *** icon is now known as sstallion
[15:19:07] <mritun_> quasi: keyword "day to day"
[15:19:07] <flyingparchment> quasi: i believe sun sells those now :)
[15:19:13] <networkdump> I'm trying to debug opensolaris here, is there an equivalent of tcpdump?
[15:19:22] <flyingparchment> networkdump: snoop
[15:19:24] <mritun_> and I think we were talking about desktop use ;)
[15:19:44] <infidel> networkdump: wireshark
[15:19:59] <Gekkko[PDA]> quasi: we were talking desktop use fss.
[15:20:02] <Gekkko[PDA]> ffs*
[15:20:03] <networkdump> thank you flyingparchment
[15:20:07] <networkdump> runing in console only ;)
[15:20:08] <suhail> ok guys. i have to go now.nice talking to you.bye.
[15:20:24] <networkdump> and my net is not working, so no I can't install the cli version
[15:20:25] <infidel> bye
[15:20:28] <mritun_> quasi: one thing more.. why use US IV etc when T1/T2 are already there ?
[15:20:31] <mritun_> floating point ?
[15:21:05] *** suhail has left #opensolaris
[15:22:04] <infidel> networkdump: what?
[15:22:12] <LeftWing> mritun_: For applications that haven't been designed for massively parallel CPUs?
[15:22:16] <networkdump> wireshark is installed by default?
[15:22:23] <mritun_> hmm
[15:22:27] <infidel> networkdump: on what?
[15:22:37] <networkdump> opensolaris, base (not full)
[15:23:04] <infidel> i doubt it
[15:23:35] <infidel> if your network is not running why are you trying to do a dump?
[15:23:41] <networkdump> my network *is*
[15:24:01] <networkdump> simply my xen guest can't access to it
[15:24:29] <infidel> explain
[15:26:17] <mritun_> wow... Novell+AMD have just released new Radeon driver for R600 family
[15:26:20] <mritun_> that was fast !!
[15:26:24] <networkdump> no time, have to leave
[15:26:30] <mritun_> when do we see the candy on OS
[15:27:02] <infidel> bye
[15:27:21] *** paul has quit IRC
[15:27:22] <networkdump> bye, I'll come back and explain later
[15:27:30] <infidel> as you wish
[15:29:16] *** paul has joined #opensolaris
[15:31:06] *** FBdev has quit IRC
[15:31:58] <quasi> mritun_: performance - 12 US IV+, 4 HBAs, a bit of scsi and a pile of extra gig-net ... no chance in hell I'd get the same sort of performance out of a t2000
[15:32:13] *** theRealballchalk has quit IRC
[15:32:30] <trygvis> quasi: but for that price you could get a whole pile of t2ks
[15:33:08] <quasi> trygvis: which still doesn't help single thread performance one bit
[15:33:17] <trygvis> true
[15:34:02] *** RaD|Tz has joined #opensolaris
[15:34:58] <quasi> let's reevaluate when I can get a Rock system - but that's still a bit of way down the road
[15:35:51] <trygvis> but then again, if you need single thread performance, why not run it on an operteron?
[15:36:08] <trygvis> or a bunch of cheaper USIV+ boxes?
[15:39:30] *** linux_user400354 has quit IRC
[15:41:25] <quasi> because I have more than one "single thread" to get performance out of ;)
[15:45:23] <WickedWicky> by what I read I get the impression the t2000 is only usefull for multi-threaded applications, not multi-process
[15:46:30] <WickedWicky> by what i read, from you two
[15:48:48] *** jafari has joined #opensolaris
[15:52:13] *** Fish has joined #opensolaris
[15:52:50] <Fish> hello
[15:53:48] *** Tpent1 has joined #opensolaris
[15:58:33] *** Tpenta has quit IRC
[16:02:25] *** glagasse has joined #opensolaris
[16:09:32] <Giaco> how can I monitor a zone activity ?
[16:11:14] <Pietro_S> Giaco: depend what do you want to monitor - prstat -ZmL can be useful for basic monitor
[16:12:04] <Giaco> thanks
[16:13:02] *** fbo has quit IRC
[16:14:21] *** fbo has joined #opensolaris
[16:16:14] *** laca has joined #opensolaris
[16:22:18] *** NikolaVeber has joined #opensolaris
[16:22:50] *** _hsilva has joined #opensolaris
[16:24:20] *** dunc_ has joined #opensolaris
[16:31:20] *** nachox has joined #opensolaris
[16:31:37] <nachox> morning
[16:32:32] *** johnlev has quit IRC
[16:35:20] *** peteh has quit IRC
[16:36:14] *** hsilva has quit IRC
[16:38:28] *** hile_ has joined #opensolaris
[16:46:53] *** peteh has joined #opensolaris
[16:50:55] *** Gekkko[PDA] has quit IRC
[16:55:32] *** Sup3rkiddo has joined #opensolaris
[16:56:25] * JoergB is away: for a while
[17:00:02] *** Sup3rkiddo has quit IRC
[17:00:02] *** JWheeler has quit IRC
[17:00:10] *** pizdec has joined #opensolaris
[17:01:29] *** dunc_ has quit IRC
[17:04:36] *** Dar is now known as Dar_HOME
[17:05:16] *** dunc_ has joined #opensolaris
[17:11:17] *** JWheeler has joined #opensolaris
[17:11:52] *** dunc_ has quit IRC
[17:15:07] *** tsoome has quit IRC
[17:16:29] *** Sup3rkiddo has joined #opensolaris
[17:18:35] *** cypromis has quit IRC
[17:21:26] *** slowhog has joined #opensolaris
[17:21:51] *** tsoome has joined #opensolaris
[17:23:23] *** cypromis has joined #opensolaris
[17:25:41] <jafari> how can i get an extend lisy from ps command on BSD ps -auxwww will show everything not getting everything in solaris doing ps -elf?
[17:26:20] <Pietro_S> jafari: what is missing?
[17:26:43] <jafari> just a long list
[17:27:06] <jafari> i am getting this for SunONE webserver
[17:27:08] <jafari> webservd -r /opt/SUNWwbsvr -d /opt/
[17:27:13] *** NikolaVeber_ has joined #opensolaris
[17:27:23] <jafari> i know there is more info than  that to show
[17:27:53] * alanc_away isn't here, but if he was, he'd say /usr/ucb/ps auxwww
[17:27:53] <Cyrille> pargs
[17:28:14] <Cyrille> or what the guy who isn't here said.
[17:28:51] <jafari> i get /usr/ucb/ps -auxwww
[17:28:52] <jafari> psiekl: unknown user xwww
[17:29:29] <holcomb> no -
[17:29:32] <jafari> alanc_away any idea
[17:29:47] <Cyrille> and even with a - I don't see that error on my box.
[17:30:07] <holcomb> yeah
[17:30:17] *** cypromis has quit IRC
[17:31:11] <alanc_away> weird, works for me on nv_72 (and every Solaris I've used since 2.3), but I have to run out the door to make it to work in time for 9am meeting, so I can't help more right now
[17:31:41] <jafari> i have ps in /usr/bin and /usr/ucb and they both dont take xwww
[17:32:06] <Cyrille> did you try "/usr/ucb/ps auxwww"?
[17:32:25] <jafari> yup
[17:32:44] <Cyrille> actually that error message is what you'd get with the standard ps...
[17:32:59] *** derchris has quit IRC
[17:32:59] <jafari>  cat /etc/release
[17:33:06] <jafari> Solaris 10 6/06 s10s_u2wos_09a SPARC
[17:33:24] *** vmlemon has joined #opensolaris
[17:34:08] <jafari> weird
[17:34:16] <Cyrille> pkgchk -l -p /usr/ucb/ps
[17:34:18] <Cyrille> ?
[17:34:33] <jafari> i been trying to get this forever and i kinda need an extend list its better for me
[17:34:53] <Cyrille> anyway you can always use pargs if you just want the list for a process.
[17:35:29] <jafari> http://pastebin.com/m471f4272
[17:35:59] <jafari> was lot of info to paste
[17:36:52] <jafari> seem like my system time is out of sync
[17:36:55] <jafari> how can i fix it
[17:37:28] <jafari> its really 11:37 am right now
[17:37:38] <jafari> i have 10:09:33
[17:37:41] <timsf> ntpdate <server>
[17:37:58] <Cyrille> I have the same package information on my system, so that's not the issue.
[17:38:01] <jafari> where can i find a ntp server
[17:38:08] <timsf> google:ntp server
[17:38:11] <holcomb> pool.ntp.org
[17:38:30] <jafari> thanks
[17:40:45] *** LeftWing has quit IRC
[17:43:55] *** NikolaVeber has quit IRC
[17:44:06] *** LeftWing has joined #opensolaris
[17:44:32] <PerterB> jafari: just use pargs
[17:46:53] *** sarahj has joined #opensolaris
[17:56:33] <jafari> does /etc/TIMEZONE have anything to do with my server being out of sync
[17:56:48] *** cypromis has joined #opensolaris
[17:56:57] <jafari> i have this box connect directly yo my modem and its acting as a firewall/nat/router
[17:57:03] <Cyrille> PerterB, I suggested that about half an hour ago but for some reason it didn't seem to be an acceptable answer.
[17:57:14] <paul> jafari: out of sync in what way?
[17:57:20] <jafari> shouldnt it sync my date from the ISP server
[17:57:22] <Cyrille> time wise
[17:57:29] <jafari> yeah time
[17:57:31] <PerterB> Cyrille: oops, so you did
[17:57:45] <jafari> Tue Sep 18 10:29:57 EDT 2007
[17:57:46] *** derchris has joined #opensolaris
[17:57:53] <PerterB> it will if you configure it to
[17:57:54] <paul> jafari: no
[17:57:56] <jafari> its 3 minutes to 12 now
[17:58:03] <paul> jafari: you need to configure NTP
[17:58:13] <paul> jafari: or just run 'ntpdate -s ntp1.esat.net'
[17:58:17] <Cyrille> there seems to be some echo in this channel ;-)
[17:58:26] <paul> or: 'ntpdate -s pool.ntp.org'
[17:58:31] <paul> Cyrille: indeed
[17:58:38] * jamesd_ needs to write and post his mini ntp howto...
[17:58:52] <paul> jafari: and what time do you think it should be?
[17:59:16] *** cypromis has quit IRC
[17:59:24] <paul> 16:57 < jafari> its 3 minutes to 12 now
[17:59:32] <paul> the '3 minutes to' /might/ be correct.
[17:59:33] <jafari> that worked paul will this stay if i reboot
[17:59:44] <jamesd_> Tue Sep 18 10:59:13 CDT 2007
[17:59:47] <paul> jafari: what does 'date -u' say?
[17:59:59] <jafari> Tue Sep 18 15:59:51 GMT 2007
[18:00:15] <jafari> i ran sudo ntpdate -s ntp1.esat.net and it worked
[18:00:50] <paul> jafari: so your system clock (which is always kept on UTC) is correct.
[18:01:04] <paul> jafari: if 'date' on its own is wrong, it's your timezone setting.
[18:02:42] <jafari> in the timezone file i have
[18:02:43] <jafari> TZ=US/Eastern
[18:02:44] <jafari> CMASK=022
[18:03:15] *** johnlev has joined #opensolaris
[18:03:42] *** stevel has joined #opensolaris
[18:03:43] *** ChanServ sets mode: +o stevel
[18:03:43] <jafari> is that fine for NYC area
[18:04:23] <PerterB> don't minutes have to be shorter for NYC?
[18:04:28] <Cyrille> :-)
[18:04:58] <Cyrille> are you sure New York minutes are supposed to be shorter? It's never been that clear to me...
[18:06:34] *** cypromis has joined #opensolaris
[18:06:39] <PerterB> I was just going by the Jonny Carson thing
[18:10:09] *** cypromis has quit IRC
[18:18:25] *** dlynes_ is now known as dlynes
[18:21:21] *** liranai has joined #opensolaris
[18:24:17] *** alanc-away is now known as alanc
[18:27:38] *** polk__ has quit IRC
[18:28:01] *** liranai has left #opensolaris
[18:28:15] <holcomb> if minutes in new york are shorter, then it must be in a higher-velocity frame of reference.  or possibly lower gravity
[18:28:54] <g4lt-sb100> if anything, I'd say new your sucks MORE
[18:28:59] <g4lt-sb100> york even
[18:30:49] *** KermitTheFragger has quit IRC
[18:34:20] *** pfa3rh has joined #opensolaris
[18:34:47] *** polk__ has joined #opensolaris
[18:37:02] *** charlesnw has joined #opensolaris
[18:40:26] *** trs81 has quit IRC
[18:40:36] *** Erwann has quit IRC
[18:42:52] *** trs81 has joined #opensolaris
[18:42:54] *** timsf has quit IRC
[18:47:19] *** peteh has quit IRC
[18:49:35] *** Cass has quit IRC
[18:55:40] *** Murmuria has joined #opensolaris
[18:55:47] *** Murmuria has left #opensolaris
[18:57:56] *** kloczek has quit IRC
[18:59:50] *** sommerfeld has joined #opensolaris
[18:59:51] *** ChanServ sets mode: +o sommerfeld
[19:02:35] *** szt has joined #opensolaris
[19:03:36] *** kloczek has joined #opensolaris
[19:07:01] *** Cyrille has quit IRC
[19:08:11] *** cydork has quit IRC
[19:09:12] *** sommerfeld has quit IRC
[19:09:45] *** sommerfeld has joined #opensolaris
[19:09:46] *** ChanServ sets mode: +o sommerfeld
[19:12:25] *** dos000 has quit IRC
[19:13:00] *** bnitz has left #opensolaris
[19:13:19] *** Fullmoon has quit IRC
[19:16:07] *** nostoi has joined #opensolaris
[19:17:56] *** migi has quit IRC
[19:18:38] *** myrkraverk has quit IRC
[19:19:01] *** Snake007uk has quit IRC
[19:23:09] *** cyborg-- has joined #opensolaris
[19:23:36] <cyborg--> Hey.  Whats the deal with zfs shareiscsi,   i created two zfs volumes, and shared them, but on the initiator side,  they are both detected, but they are both the same backing source,
[19:23:53] <flyingparchment> what do you mean, the same backing source?
[19:24:13] <cyborg--> okay well,    i get sdb and sdc on the initiator side, but they are the same disk
[19:24:26] <cyborg--> i.e.,  i formatted sdb and then went on to format sdc but it already was
[19:24:56] <cyborg--> this includes sfiscsi initiator on linux,  and  microsoft iscsi initiator on  windows
[19:25:44] <cyborg--> if i only share one at a time its okay,
[19:25:55] <cyborg--> or if i  use acl's i can have multiple servers accessing different targets
[19:26:06] <cyborg--> but.... one initiator cannot seem to  access two targets,
[19:36:30] *** stevel has quit IRC
[19:36:30] *** JWheeler has quit IRC
[19:36:31] *** laca has quit IRC
[19:36:31] *** Fish has quit IRC
[19:36:31] *** Doc has quit IRC
[19:36:33] *** seanmcg has quit IRC
[19:36:33] *** libkeiser has quit IRC
[19:36:34] *** palowoda has quit IRC
[19:36:34] *** charlieS_ has quit IRC
[19:36:34] *** Kush- has quit IRC
[19:36:34] *** bengtf has quit IRC
[19:36:34] *** zebra has quit IRC
[19:36:34] *** cyborg-- has quit IRC
[19:36:35] *** FireflyST has quit IRC
[19:36:36] *** cormac has quit IRC
[19:36:36] *** detriment has quit IRC
[19:36:36] *** Ditaka has quit IRC
[19:36:37] *** Symmetria has quit IRC
[19:36:37] *** jolts has quit IRC
[19:36:37] *** mustang has quit IRC
[19:36:37] *** renihs has quit IRC
[19:36:37] *** steleman has quit IRC
[19:36:37] *** The-spiki has quit IRC
[19:36:38] *** dennis- has quit IRC
[19:36:38] *** HPeterP has quit IRC
[19:37:14] *** stevel has joined #opensolaris
[19:37:14] *** cyborg-- has joined #opensolaris
[19:37:14] *** JWheeler has joined #opensolaris
[19:37:15] *** laca has joined #opensolaris
[19:37:15] *** Fish has joined #opensolaris
[19:37:15] *** Doc has joined #opensolaris
[19:37:15] *** seanmcg has joined #opensolaris
[19:37:15] *** libkeiser has joined #opensolaris
[19:37:15] *** palowoda has joined #opensolaris
[19:37:15] *** charlieS_ has joined #opensolaris
[19:37:15] *** Kush- has joined #opensolaris
[19:37:15] *** bengtf has joined #opensolaris
[19:37:15] *** zebra has joined #opensolaris
[19:37:15] *** FireflyST has joined #opensolaris
[19:37:15] *** cormac has joined #opensolaris
[19:37:15] *** Ditaka has joined #opensolaris
[19:37:15] *** detriment has joined #opensolaris
[19:37:15] *** dennis- has joined #opensolaris
[19:37:16] *** Symmetria has joined #opensolaris
[19:37:16] *** jolts has joined #opensolaris
[19:37:16] *** The-spiki has joined #opensolaris
[19:37:16] *** HPeterP has joined #opensolaris
[19:37:16] *** renihs has joined #opensolaris
[19:37:16] *** steleman has joined #opensolaris
[19:37:16] *** mustang has joined #opensolaris
[19:37:16] *** irc.freenode.net sets mode: +o stevel
[19:39:40] <flyingparchment> don't Xstore make JBOD-type devices, or am i think of someone else?
[19:39:40] <cyborg-->  but.... one initiator cannot seem to  access two targets,
[19:41:15] <flyingparchment> ah, xtore
[19:42:05] *** myrkraverk has joined #opensolaris
[19:42:37] <cyborg--> i was thinking about getting a  x4500 if i could reach any success with iscsi/zfs
[19:45:10] <sommerfeld> cyborg: i'm a little vague on iscsi, but .. is it two targets, or two luns of one target?
[19:46:34] <sommerfeld> i know people have been able to export multiple disks from the solaris iscsi server to a single solaris iscsi client.
[19:47:16] *** FBdev has joined #opensolaris
[19:47:55] <sommerfeld> .. but i haven't done it myself
[19:48:59] <cyborg--> well,   using zfs shareiscsi  it creates two targets each with one lun
[19:49:35] <sommerfeld> cyborg: you may want to see if you can do the share via iscsitadm rather than via setting shareiscsi
[19:49:54] <cyborg--> yeah i was thinking i will likely need to do it that way so i can   specify backend  souces as individual lun's
[19:50:41] <cyborg--> the machien i put online to launch this into product is 10u4 but, last night the iscsi target just like, froze
[19:56:04] <sommerfeld> you may get better results asking on the zfs-discuss or storage-discuss lists
[19:56:11] *** derchris has quit IRC
[19:57:24] *** estibi_ has joined #opensolaris
[19:58:20] *** reflect has left #opensolaris
[19:58:35] *** reflect has joined #opensolaris
[19:59:04] *** blindfish has joined #opensolaris
[20:00:00] *** timsf has joined #opensolaris
[20:05:11] *** _setuid_H has joined #opensolaris
[20:05:23] <_setuid_H> Does exist glibc port to solaris?
[20:05:59] <jamesd_> yes...  should be able to get it at  sunfreeware.com or blastwave.org
[20:06:39] <_setuid_H> I'm trying to get work / build all GNU software on my SXCE
[20:06:47] *** estibi has quit IRC
[20:07:09] <_setuid_H> Thanks for the answer
[20:07:12] *** _setuid_H has left #opensolaris
[20:07:54] *** cmihai has joined #OpenSolaris
[20:08:02] <PerterB> I'm pretty sure there isn't a port of glibc, but most code should compile with the native libc (after all, GNU claim to be portable)
[20:08:13] <PerterB> oh, he left, duh
[20:08:17] <jamesd_> _setuid  blastwave.org has over 1500 packages..
[20:08:42] <jamesd_> well gnome builds.. so it must have glibc
[20:09:05] <PerterB> I'm pretty sure he means GNU libc, not glib
[20:09:10] *** Mdx4 has joined #opensolaris
[20:09:29] *** dlynes_ has joined #opensolaris
[20:10:46] <flyingparchment> glibc and glib are very different things
[20:11:05] <PerterB> that was my point :)
[20:11:12] <flyingparchment> glibc is a libc implementation (GNOME doesn't require it), glib is a library of utility functions for C (GNOME/GTK does require it)
[20:11:21] <flyingparchment> yes, i'm just explaining why :)
[20:14:27] *** szt has quit IRC
[20:17:16] *** dlynes has quit IRC
[20:17:16] * WickedWicky starts running in circles, screaming "Panic! Panic!"
[20:17:39] * dclarke slaps WickedWicky
[20:17:44] <WickedWicky> my Senseo machine is gone to meet its maker
[20:17:46] * bda pokes at WickedWicky's debugger.
[20:18:21] <WickedWicky> you dont need a debugger, it's obvious what's wrong, my coffee maker died
[20:18:44] <WickedWicky> hey bda, dclarke
[20:18:52] <WickedWicky> dclarke, you got that mail I sent you?
[20:19:23] <dclarke> umm .. no
[20:19:27] <dclarke> recently ?
[20:19:36] <WickedWicky> well it must be two to three weeks ago now
[20:19:53] <dclarke> geez .. you better resend it
[20:20:01] <dclarke> what should I grep for to find it ?
[20:20:04] <WickedWicky> I will do so tomorrow when I am in the office
[20:20:16] <dclarke> just give me a string to grep for
[20:20:28] <WickedWicky> it's sent from p.ale at kpn dot com
[20:21:01] <dclarke> search running
[20:21:42] *** wamty has joined #opensolaris
[20:21:54] <dclarke> Microsoft SMTPSVC(6.0.3790.1830)
[20:22:08] <WickedWicky> sent from work
[20:22:09] <WickedWicky> so yes
[20:23:07] <dclarke> that wasn't it ..
[20:23:09] *** sarahj has quit IRC
[20:23:26] <dclarke> that was something telling me that I was a UK NATIONAL LOTTERY winner
[20:23:29] <dclarke> erk
[20:23:38] <dclarke> so .. nothing here from kpn.com
[20:23:39] <WickedWicky> that definately wasnt what I sent you :P
[20:23:50] <WickedWicky> I'll resend tomorrow
[20:23:57] <dclarke> thanks
[20:27:29] *** bondolo has quit IRC
[20:27:39] <victori__> are there smf startup scripts for oracle10g?
[20:28:05] <victori__> ah nm google knows ;-/
[20:28:30] <victori__> http://padraigs.blogspot.com/2007/04/configuring-oracle-as-service-in-smf.html
[20:30:11] *** _setuid_H has joined #opensolaris
[20:31:37] <_setuid_H> Hi I need to get minimal bootable solaris configuration to my zfs tank
[20:31:53] <_setuid_H> I have allready built onnv
[20:32:10] <_setuid_H> is there any way to bfu to some kind of empty chroot?
[20:32:13] *** hile_ has quit IRC
[20:34:57] *** bondolo has joined #opensolaris
[20:35:38] *** yippi has joined #opensolaris
[20:36:52] *** nachox has quit IRC
[20:37:40] *** hile_ has joined #Opensolaris
[20:41:39] *** thalin has quit IRC
[20:48:42] *** RaD|Tz has quit IRC
[20:49:55] *** RaD|Tz has joined #opensolaris
[20:50:02] *** sfire||mouse has joined #opensolaris
[20:50:27] *** RaD|Tz has quit IRC
[20:50:48] *** RaD|Tz has joined #opensolaris
[20:53:32] *** derchris has joined #opensolaris
[20:54:49] *** qmrw has quit IRC
[20:55:04] <flyingparchment> 125504-02 SunOS 5.10_x86: package-move-of-IP-objects patch
[20:55:07] <flyingparchment> funny name
[20:57:47] *** Megaf has joined #OpenSolaris
[20:59:57] *** cypromis has joined #opensolaris
[21:00:09] *** lisppaste3 has quit IRC
[21:00:39] *** jafari has quit IRC
[21:02:37] *** nostoi has quit IRC
[21:06:42] *** lisppaste3 has joined #opensolaris
[21:06:53] <victori__> where are the smf xml files saved when you import them with svccfg
[21:08:08] <seanmcg> you don't need to know :)
[21:08:23] <victori__> ok then I guess I need to remove, edit and import again
[21:08:30] <victori__> seanmcg: but that is a weak response ;-/
[21:08:40] <seanmcg> typically under /var/svc/manifest/...
[21:08:41] <victori__> I would still like to know just to know
[21:08:45] <victori__> thank you
[21:08:52] <victori__> well that is where I have imported from
[21:09:19] *** RaD|Tz has quit IRC
[21:09:28] <seanmcg> importing an xml manifest puts them into a small database (if I recall right)
[21:09:53] <wesolows> correct, the repository
[21:09:56] <wesolows> normally located in /etc/svc
[21:10:25] <wesolows> however, the repository is not an interface and must not be manipulated in any way
[21:10:38] <seanmcg> Yaksha, thats it..
[21:10:44] <wesolows> all access to it is through tools like svccfg and svcadm
[21:10:48] <seanmcg> hmm, trigger finger again.
[21:12:08] <victori__> ok what about a low-tech solution any rc.local type file where I can just easily append startup scripts?
[21:12:28] <wesolows> you can still use /etc/rc*.d for legacy stuff
[21:12:41] <wesolows> but all new software is to have manifests
[21:12:44] <wesolows> which are easy to write
[21:13:44] <sommerfeld> s/write/steal & tweak from a working example/
[21:13:53] <sommerfeld> i wouldn't say they're easy to write from scratch
[21:14:20] <sommerfeld> unless your native language is xml
[21:14:32] <sommerfeld> but it's easy to cobble one together
[21:14:53] <victori__> if I import the same manifest file will it overwrite?
[21:14:56] <victori__> svccfg*
[21:15:51] <flyingparchment> hmm, the new S10 kernel patch looks scary
[21:20:10] <sickness> wesolows: "easy to write"? gimme the good old openbsd's rc.local any day over the manifests+methods and db thing :'
[21:22:34] <wesolows> sorry, I think the extra 10 minutes is well woth the benefits
[21:22:38] <wesolows> rc.local is a pile of shit
[21:23:26] <wesolows> failure semantics?  service status?  recovery?  logging?  dependencies?  restarts?  addition of new software?
[21:23:34] <wesolows> this is a no-brainer
[21:23:36] <bda> SMF++
[21:23:52] <bda> People saying it is overkill confuse me greatly.
[21:24:31] *** FBdev has quit IRC
[21:24:43] <sickness> wesolows: all those things are nice to have, but I'd also want to be able to just start something if I want, no hassle on a stupid service that becames stuck trying to restart a crashing daemon...
[21:25:11] <elektronkind> then set it to not restart a crashing daemon
[21:25:14] <elektronkind> duh
[21:25:51] *** FBdev has joined #opensolaris
[21:25:54] <wesolows> or just disable the service and run the method script by hand
[21:26:25] * elektronkind , like, rolls his eyes and pops his gum
[21:26:36] <wesolows> yeah - I don't get it either
[21:26:55] <wesolows> if the thing's "getting stuck" then something is wrong with the components of the service and you need to fix it
[21:27:20] <mritun_> well... if there is a dependency on a service that is just not starting... what use is trying to start the dependent service ?
[21:27:27] <wesolows> so you're essentially saying "I want something that's broken to Just Work and SMF isn't helping me hide the brokenness"
[21:27:28] <WickedWicky> I think the focus should be on "crashing daemon" :P
[21:27:42] <sickness> well it's too much hassle imho, I'd prefer not to have to bother with this IF it's not needed...
[21:27:45] <wesolows> that too
[21:27:50] <wesolows> it is needed
[21:27:53] <wesolows> please use it
[21:28:00] <wesolows> if you need help, ask on smf-discuss
[21:28:08] <sickness> I'd just like to have the option of a simple rc.local, at least for testing purposes :/
[21:28:09] <mritun_> sickness: then boot to a lesser milestone and start whatever you want by hand, right ?
[21:28:25] <sickness> mritun_: not by hand, I'm able to do that anyway...
[21:28:34] <mritun_> then ?
[21:28:34] <wesolows> delivering rc scripts does your customers a disservice and would never be accepted into OpenSolaris.  If it's for your own use, it does yourself a disservice and denies you a learning opportunity
[21:28:51] <mritun_> how'd you tell your rc.local to start networking *before* starting sendmail ?
[21:29:00] <mritun_> or should it "just know" ?
[21:29:07] <wesolows> rc.local is about 3 generations behind SMF; it's not useful even "for testing"
[21:29:40] <mritun_> I'm not defending SMF... it has its own warts
[21:29:48] <mritun_> but it's less imperfect than others
[21:30:01] <wesolows> if you really want to make an SMF-less distribution, you can start by making inittab not run startd
[21:30:25] <wesolows> of course, you'll also have to find all the integration points in all the other stuff (nscd immediately comes to mind but there are plenty more) and rip all of it out too
[21:30:39] <sommerfeld> if i were going to go with something simpler, rc.local wouldn't be it.  netbsd's rc.d would be it.
[21:30:54] <wesolows> by the time you're done destroying all that value, you'll have spent more time than it would have taken to learn to use SMF to your advantage
[21:31:20] <sommerfeld> (and that has the same sort of dependency model as SMF)
[21:31:23] *** _setuid_H has left #opensolaris
[21:31:37] <victori__> smf is nice but not so nice for quick one off hacks
[21:31:43] <wesolows> yeah, sort of SMF plus 52-conf-file-pickup
[21:31:56] <elektronkind>  rc.local for init scripts is analogous to /proc on linux. Just a place where people mentally gave up and have resorting to pooping there and not bothering to cover it
[21:32:02] <wesolows> eliminating configuration files is one of the great unfulfilled promises of SMF
[21:32:05] <mritun_> but what *are* these quick one off hacks ?
[21:32:09] <sickness> I do not want an smf-less distribution, I'd just want rc.local, started after everything else...
[21:32:09] *** hile_ has quit IRC
[21:32:10] <victori__> how would you check ram/free ram in solaris?
[21:32:22] <wesolows> victori__: vmstat, if you insist
[21:32:22] <victori__> mritun_: start up some rails/catalyst processes
[21:32:31] <victori__> which obviously don't need full blown smf profiles
[21:32:38] <wesolows> "obviously"?
[21:32:46] <wesolows> and what's "full blown" mean here?
[21:32:48] <mritun_> victori__: that would be simple to do right ?
[21:32:53] <wesolows> a 30-line manifest?
[21:33:02] <victori__> I am not going to sit and write 30x10 manifest files
[21:33:08] * sfire||mouse setup an smf manifest for vsftpd
[21:33:13] <wesolows> look into services versus instances
[21:33:16] <mritun_> you can tell SMF to start your "custom" script latest in the booting process
[21:33:22] <victori__> when perl /opt/websites/site1/scripts/server.pl start
[21:33:27] <wesolows> you may be able to leverage multiple instances off a single service where they have common properties
[21:33:28] <victori__> will do the trick
[21:33:29] <mritun_> in that custom script you can put all the misc poop etc
[21:33:30] <seanmcg> so have a template.xml and use sed.
[21:33:31] <PerterB> oh, that reminds me, it looks like there's now a mechanism to run post-install scripts that need running after the system boots of its new root... I saw it running lots of gnome stuff the other day. Is that documented and supported? (this used to be another handy use for rc scripts)
[21:33:34] <wesolows> ok, well, have fun then
[21:33:38] <wesolows> thanks for playing
[21:34:02] <wesolows> don't call us when your customers call you at 3am to tell you they're cancelling because your server went off line and couldn't recover itself properly
[21:34:03] <WickedWicky> victori__: ... like it or not.. sfm is the new standard...
[21:34:19] <victori__> oh concerning vmstat I never see "b" ever go above 0 even though there is definite disk activity, is this normal?
[21:34:20] <WickedWicky> you better get yourself familiar with it now while you have the chance..
[21:34:46] <WickedWicky> I found it a nightmare in the beginning as well, coming from Linux and Solaris 8 and 9
[21:34:54] <mritun_> victori__: create a SMF "service" for /etc/misc_start.sh
[21:34:55] <WickedWicky> now i am bitching at Solaris 8 boxes not having it ;-)
[21:35:07] <victori__> mritun_: your right
[21:35:08] <mritun_> victori__: where you can put *all* your one off hacks
[21:35:20] <mritun_> but BEWARE
[21:35:23] * bda shakes head.
[21:35:26] <mritun_> *never* shipt it as such
[21:35:28] <victori__> I am coming from freebsd rc.local mentality
[21:35:40] <wesolows> WickedWicky: yup - step 0 in jumpstart postinstalls is deleting all the /etc/rc.d goo for wbem and such; of course, it comes right back when you patch.  Oh, but SMF is "confusing."
[21:35:50] <victori__> ya so concerning vmstat and never seeing any disk contention is that normal?
[21:36:01] <victori__> "b" for blocked disk access
[21:36:08] <victori__> never goes above 0 when the disk is trashing
[21:36:09] *** dunc has joined #opensolaris
[21:36:22] *** sioraiocht_ has joined #opensolaris
[21:36:32] <seanmcg> mritun_: ya, misc_start.sh - which all the 'poop' goes.  Then watch the bootup take forever cause of that 'poop'.  Instead of being run in parallel.
[21:36:36] <elektronkind> victori__: people tend to use 'iostat -nx <interval>' when doing IO snooping
[21:36:56] <seanmcg> victori__: check 'iostat -zxn 1' see if any disk is being hit much.
[21:36:57] <elektronkind> iostat has tons more, for various things
[21:37:07] <mritun_> seanmcg: keep them happy... let them take their own time to acclimatize
[21:37:10] <bda> iostat.d, iosnoop.d..
[21:37:18] <mritun_> seanmcg: learning is a slow and *painful* process
[21:37:37] <seanmcg> mritun_: doesn't like bad habits starting out...
[21:37:40] <mritun_> seanmcg: no pain no gain. When they see what pain rc.local is, they'd learn SMF
[21:38:09] <mritun_> and anyways.. it's pointless to debate with someone doesn't want to learn
[21:38:19] <RElling> BTW, using SMF makes that part of the boot go twice as fast on a dual core... parallel is good.
[21:38:27] <mritun_> let them have their way, and learn by failing
[21:39:27] *** yippi has quit IRC
[21:44:27] *** sioraiocht has quit IRC
[21:48:40] *** yippi has joined #opensolaris
[21:49:53] *** wamty has quit IRC
[21:53:16] <cormac> 8
[21:54:18] *** yippi has quit IRC
[21:55:48] *** paul has quit IRC
[21:57:10] <flyingparchment> do large pages help database performance on amd64?
[21:58:07] *** richlowe has joined #opensolaris
[21:59:14] <sommerfeld> flyingparchment: my understanding that large pages are hard to get on amd64 because the kernel cage isn't enabled on x86/amd64
[21:59:25] <flyingparchment> ah.
[21:59:44] <richlowe> That matches my understanding
[21:59:49] <richlowe> (and an email I read recently)
[22:00:08] *** paul has joined #opensolaris
[22:00:13] <richlowe> sommerfeld: what I don't know, though, is why the cage isn't there.
[22:00:23] <richlowe> sommerfeld: sure, it doesn't have DR, but it's still a good thing (it seems)
[22:00:41] *** fbo has quit IRC
[22:01:15] <sommerfeld> (kernel cage constrains kernel allocations -- which generally can't be relocated -- to specific physical pages, so uncaged memory can be shuffled around more readily)
[22:01:30] <sommerfeld> richlowe: I think nobody has gotten the required Round Tuit to make it happen yet.
[22:01:57] <alanc> that is the most bizarre request for sponsorship I've seen on request-sponsor yet
[22:02:02] <richlowe> Ah, I wondered if there was a technical reason.
[22:02:20] <sommerfeld> i don't believe there are technical reasons why it couldn't be made to work.
[22:02:22] <alanc> and yet, they did actually phrase it as a request for sponsorship, so bonus points there...
[22:02:41] <sommerfeld> alanc: i agree that she needs help.  but not the sort we can provide.
[22:03:37] <sommerfeld> unfortunately any attempt to help her will likely reinforce her paranoia
[22:03:38] *** dlynes_ has quit IRC
[22:03:54] <richlowe> Ok, someone complained at me about wx sometime last night just before my connection dropped.
[22:03:58] <richlowe> and I totally forget who.
[22:04:42] <richlowe> alanc: didn't list a contributor ID, either.
[22:05:09] <sommerfeld> anyhow, back to kernel cages: I think there was a general plan to eliminate the cage and make kernel memory relocatable but there are driver dma related issues..
[22:05:19] <alanc> heh, I am NOT going to ask her for that - we'd probably get accussed of IP "parasitism"
[22:08:57] <wesolows> That is indeed bizarre
[22:09:12] <wesolows> especially considering that conditions in Brazil are fairly good and apparently improving.
[22:13:29] *** victori__ is now known as victori
[22:16:11] <elektronkind> we have one user here who occasionaly contacts the university helpdesk claiming that the burmese government is spying on her by putting "screen trackers" on all university  computers and using other students to watch her
[22:16:29] <richlowe> elektronkind: sounds like a good excuse for a few days off while you 'look into it'
[22:16:50] *** comay has joined #opensolaris
[22:18:16] *** Tpenta has joined #opensolaris
[22:18:34] <elektronkind> heh yeah, though I have to wonder how much more panaoid she is now that the labs are mostly iMacs with the built-in camera
[22:18:46] <WickedWicky> some people really have a screw or two lose
[22:19:21] <RElling> hmmm... the performance folks often end up fighting the kernel cage... not good for some workloads
[22:19:53] <WickedWicky> they can watch me as much as they want, especialy when they're brunette/redhaired , B34 and cute
[22:20:23] <wesolows> TMI chief
[22:20:30] <WickedWicky> :P
[22:20:54] <WickedWicky> in the bus or at work of course
[22:20:57] <WickedWicky> what were you thinking :s
[22:21:54] <WickedWicky> we have webcams on the data floor, quite handy when one of the rack people needs pointers to which machine to work on/with, cell phone in one hand, watching the webcam feed in the other
[22:22:49] <WickedWicky> "The Netra 240... second from above... NO! THE OTHER ABOVE, OPPOSITE OF THE FLOOR"
[22:27:05] *** Tpent1 has quit IRC
[22:28:42] *** eboutilier has quit IRC
[22:31:38] *** _JR_ has joined #opensolaris
[22:31:56] *** FireflyST has quit IRC
[22:32:14] *** _JR_ has left #opensolaris
[22:32:36] *** FireflyST has joined #OpenSolaris
[22:43:27] *** derchris has quit IRC
[22:46:03] <sommerfeld> WickedWicky: "Sorry, gravitational field polarity isn't what it looks like here..."
[22:48:43] <quasi> WickedWicky: I've always wanted to force the hw people to wear a camera
[22:48:48] <sommerfeld> friend of mine who once worked for a rack mount pc builder said that some of their customers used the cdrom drive tray to identify which system they wanted remote hands to work on -- "eject cdrom" to make system stick out its tongue.
[22:49:05] <elektronkind>  heh
[22:49:10] <elektronkind> practical :)
[22:49:21] *** Fish has quit IRC
[22:49:28] <quasi> locator lights are very practical
[22:49:37] <elektronkind> a place to hold the tech's coffee while he works
[22:49:48] <flyingparchment> i like 'asset tags' - if the system can display its asset tag on the lcd
[22:49:50] <sommerfeld> well, except if he needs to pull that system out of the rack..
[22:50:05] *** GmanAFK is now known as Gman
[22:50:28] <quasi> flyingparchment: locator lights when you label a new pile of machines
[22:55:12] *** jgay has joined #opensolaris
[22:59:32] *** nrubsig has joined #opensolaris
[22:59:33] *** ChanServ sets mode: +o nrubsig
[22:59:51] <nrubsig> alanc: When will the afternoon summit phone conference start ?
[23:00:08] *** derchris has joined #opensolaris
[23:01:56] <jgay> I was wondering if anyone could direct me to any discussions or wiki pages that discuss the possibility of dual-licensing open solaris under GPLv3.
[23:02:14] <wesolows> no
[23:02:32] <sommerfeld> anywhere but here.
[23:02:43] <richlowe> You misread the question I think.
[23:02:50] <richlowe> if pointing someone to the *last* argument saves a repetition...
[23:02:53] <wesolows> Can anyone direct him?  No.
[23:03:02] <wesolows> Google can though, easily.
[23:03:08] <jmcp> richlowe: while I think about it, have the wx checks been updated for code which was not created by Sun?
[23:03:10] <wesolows> Or search the mailing lists on opensolaris.org
[23:03:13] <richlowe> sorting osol-discuss threads by message count would.
[23:03:22] <richlowe> jmcp: That question makes no sense.
[23:03:27] <richlowe> the checks are the same whoever wrote the code.
[23:03:34] <richlowe> if you want to write non-conformant code, integrate it somewhere else.
[23:03:36] *** mritun_ has quit IRC
[23:04:18] <jmcp> richlowe: you're misunderstanding my poorly expressed question: when I run wx pbchks on jbk's libdisasm code which he created, I get an error about the copyright year being incorrect
[23:04:31] <jmcp> eg::    Proper 2007 copyright missing in usr/src/lib/libdisasm/sparc/dis_sparc.c.
[23:04:31] <jmcp> Only the current year should appear.
[23:04:34] *** glagasse has quit IRC
[23:04:46] <jmcp> but it does, and it conforms to the required format as far as I can see
[23:05:46] <richlowe> Link me to the code.
[23:06:13] <jgay> wesolows, thanks, I didn't think of searching through the mailing lists. Sorry if this is a sour point, or something.
[23:06:24] <jmcp> jgay: it's a very, very sore point
[23:06:33] <nrubsig> Tecnically I HATE stuff like twitter.com ... but sometimes it becomes usefull: http://twitter.com/gisburn
[23:06:33] <wesolows> jgay: It is - there's enough to keep you busy there for about 6 months of reading. :-)
[23:08:57] *** FelipeBare has quit IRC
[23:12:01] <jgay> wesolows, jmcp, thank you. I will look around and try not to mention it in this room or distract people about it :-)
[23:12:49] <jmcp> jgay: nowurries. once you have read the threads, you'll understand why
[23:13:09] * flyingparchment draws today's /etc/motd ascii art
[23:13:28] <g4lt-sb100> going with the disclaimer first to forestall the jokes now? ;P
[23:15:20] *** pitty has joined #opensolaris
[23:15:26] *** pitty has left #opensolaris
[23:16:30] *** blindfish has quit IRC
[23:22:26] <flyingparchment> since you can dtrace linux binaries, and dtrace now works in a zone, it'd be nice if sun would provide a linux dtrace binary somehow
[23:22:31] <flyingparchment> (that would work from within the zone)
[23:23:04] <g4lt-sb100> that would be so funny, brand-dtrace
[23:23:11] <wesolows> why?  what good would it do?
[23:23:22] <flyingparchment> it would be more convenient than switching between zones
[23:24:06] <wesolows> it would be a huge amount of work to do that
[23:24:18] <wesolows> but I guess it's probably possible
[23:24:24] <wesolows> you're welcome to start :-)
[23:24:31] <g4lt-sb100> actually, for even funnier, let jamesd loose on the project, he'd make a cool brandz+dtrace ;P
[23:30:03] *** cmihai has quit IRC
[23:35:21] *** tiger11 has joined #opensolaris
[23:37:27] <tiger11> Do zvol have a write cache or do they only return after the data is physically on the media?
[23:38:09] *** Megaf has left #OpenSolaris
[23:39:42] *** jmcp has quit IRC
[23:41:56] <sommerfeld> tiger11: zvol uses the intent log to ensure that synchronous writes to zvol turn into synchronous writes to the pool
[23:42:18] <pjd-> http://people.freebsd.org/~pjd/misc/zfs_self-healing.mpeg
[23:42:20] <pjd-> :)
[23:45:27] <alanc> nrubsig: sorry, was in another meeting, and I'm not going to the summit phone calls, so I don't know
[23:46:05] *** yarihm has joined #OpenSolaris
[23:47:29] <alanc> I'd bet Gman knows though
[23:47:40] <Gman> i think i already answered him?
[23:48:24] <alanc> ah, didn't see that in my scrollback
[23:48:42] <richlowe> neither do I.
[23:48:43] <nrubsig> Drone: do you join the conference ?
[23:48:44] <Gman> nrubsig knows how to /query :)
[23:50:06] *** timsf has quit IRC
[23:53:02] *** hile_ has joined #opensolaris
[23:54:38] *** jmcp has joined #opensolaris
[23:55:19] <tiger11> sommerfeld: in terms of ISCSI, would this allow a write to return before actually doing the write? Basically, I am seeing not so good performance with the ISCSI target in 10u4, and I am trying to narrow it down.
[23:57:27] *** rasputnik has joined #opensolaris
[23:57:35] <infidel> how can i find ones mac address remotely?
[23:57:53] <wesolows> arp
[23:58:18] <infidel> how?
[23:58:32] *** sunman6 has quit IRC
[23:58:43] *** sunman6 has joined #opensolaris
[23:59:15] <jmcp> infidel: run /usr/sbin/arp -a
[23:59:42] <infidel> jmcp: and how will i find your mac with that command?
[23:59:44] *** yippi has joined #opensolaris





top