[00:00:16] <CIA-26> prasad: 6524011 pcie slots show up in cfgadm as "Slot<n>" vs. "pcie<n>" in Nevada/s10u4
[00:02:40] *** hohum has quit IRC
[00:03:51] *** Arnald has quit IRC
[00:05:35] *** wnorrix has joined #opensolaris
[00:06:07] <wnorrix> what is the /export dir for on osolaris ?
[00:07:06] <flyingparchment> NFS exports
[00:09:50] <wnorrix> cool
[00:09:56] <wnorrix> is it exported by default ?
[00:10:21] <jmcp> no
[00:10:28] <wnorrix> (sorry new to solaris, but old school linux geek)
[00:10:38] <flyingparchment> it's just a convention, there's nothing special about it
[00:10:40] <jmcp> no need to apologise
[00:11:12] <jmcp> generally you'll find your actual physical homedir under /export/home/$username, and the automounter will mount it under /home/$username
[00:11:21] <jmcp> but you need to explicitly share it out
[00:11:22] *** Mdx4 has quit IRC
[00:11:32] <mw46> II'll ask my question again now that more people look alive;-) Does somebody have a suggestion for ZFS+NFS?
[00:11:54] <wnorrix> :)
[00:11:55] <wnorrix> thanks
[00:12:24] <wnorrix> so i just do a share -F nfs -o rw=engineering -d 'appdata' /exports/appdata
[00:12:53] *** sahafeez has joined #opensolaris
[00:13:19] <flyingparchment> you'll probably want to adjust the values
[00:13:27] <flyingparchment> mw46: what kind of suggestion?
[00:15:18] *** bitrot has quit IRC
[00:18:12] <flyingparchment> hm.. apparently mysql.com's binaries explicitly link with mtmalloc.  which is rather unfortunate, as it wastes lots of memory
[00:18:20] *** myrkr has joined #opensolaris
[00:18:43] <myrkr> hi all
[00:19:36] <myrkr> having a qt app in a dlopen()ed .so file, I get the following error (at a specific event) if my driver app isn't linked against libX:
[00:19:54] <myrkr> ld.so.1: qt: fatal: relocation error: file /usr/openwin/lib/locale/comon/xomTRTTB.so.2: symbol XrmStringToQuark: referenced symbol not found
[00:20:04] <myrkr> would anyone here have clue as to why?
[00:20:33] <mw46> flyingparchment: we want to set up a svn repository on a NFS client (the NFS server has a big ZFS FS which we want to share). But ti looks like the writes are ... slow:-(
[00:20:34] <myrkr> that is, the .so file is (of course) linked against libX, but why does the driver app need it too?
[00:21:45] *** Pedro1 has quit IRC
[00:21:53] <myrkr> btw, the following seems (a bit) relevant:
[00:22:00] <myrkr> http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6584998
[00:23:17] <myrkr> would any of you know how to comment on that bug?
[00:24:25] *** yippi has quit IRC
[00:24:31] <flyingparchment> heh, never saw this scsi error before: "address mark not found for data field"
[00:24:56] <wnorrix> http://pastie.caboo.se/102681
[00:25:18] <wnorrix> what does cannot share filesystem in non global zones mean ?
[00:25:19] *** yippi has joined #opensolaris
[00:26:15] <flyingparchment> wnorrix: local zone cannot be an nfs server
[00:26:42] <wnorrix> zero idea?
[00:26:46] <wnorrix> what that means :)
[00:26:56] <flyingparchment> do you know what a zone is?
[00:27:15] <wnorrix> is it the Solaris Virtulization thing ?
[00:27:37] <alanc> myrkr: bugs.openslaris.org doesn't allow commenting on bugs unfortunately (one of the reasons it's being replaced with something far less broken)
[00:28:02] <wnorrix> flyingparchment: or is it something else?
[00:28:23] *** cypromis has quit IRC
[00:30:32] <myrkr> alanc: since 1) my problem seems excacly the same and 2) I have a workaround, it might be worthwhile to get my comment, so if you have an email for me (which won't be sent tonight though) that might help
[00:30:41] <alanc> zones is a form of virtualization
[00:31:02] <alanc> myrkr: you can mail me at alanc at sun dot com and I can add it to the bug report
[00:31:12] <myrkr> alanc: ok, thanks
[00:31:14] <wnorrix> ah so im right on that one
[00:31:20] <wnorrix> and zones dont support nfs
[00:31:42] <jmcp> wnorrix: not exactly correct
[00:31:58] <jmcp> as just mentioned, you cannot have a non-global zone as an nfs server
[00:32:02] <jmcp> it can be a client, no worries at all
[00:32:04] <jmcp> just not a server
[00:32:08] <wnorrix> hm
[00:32:23] <wnorrix> i am hosting with joyent
[00:32:37] <wnorrix> so i will have to get them to change these settings ... hmm
[00:33:02] <jmcp> wnorrix: Ben Rockwood is Da Man @ joyent, as far as I understand things
[00:33:15] <jmcp> he occasionally comes into this channel too
[00:33:28] <wnorrix> whats his irc nic?
[00:33:35] <jmcp> benr, iirc
[00:33:56] <jbk> evening
[00:34:00] <wnorrix> cool
[00:34:00] <jmcp> hi
[00:34:03] <wnorrix> will make note of it
[00:34:11] *** KoTeR1 has joined #opensolaris
[00:34:20] *** mikefut has quit IRC
[00:36:27] *** KoTeR1 has left #opensolaris
[00:36:28] *** WickedWicky has quit IRC
[00:43:03] *** gisburn has joined #opensolaris
[00:43:11] <gisburn> alanc: ping!
[00:43:20] <alanc> gisburn: pong!
[00:43:58] <gisburn> alanc: Do you have German+Japanese Sun USB keyboards floating around somewhere, preferably with all keys still in place ?
[00:44:39] <alanc> they were still in the lab last I looked - I took the pile from my office down there a while ago
[00:44:55] <gisburn> alanc: ---> /msg
[00:45:06] <alanc> all I have in my office now are US/Unix Type 7 & a Danish Type 6/USB
[00:45:32] <e^ipi> sun should make a dvorak board
[00:45:45] <e^ipi> or at the very least, someone should shove the xmodmap files in to X11
[00:45:47] <flyingparchment> they should support dvorak in solaris first
[00:45:49] <gisburn> Aren't dvoark people extinct ?
[00:45:58] <e^ipi> in fact, alanc ... put the xmodmap files in Xorg
[00:45:59] <e^ipi> *nod*
[00:46:08] <flyingparchment> xmodmap sucks, just add the proper keymaps
[00:46:36] <e^ipi> gisburn: myself & brendang would disagree with that statement
[00:46:50] <flyingparchment> <- dvorak model m
[00:46:52] <alanc> aren't there dvorak keymaps in the Xorg XKB files?
[00:46:55] <jmcp> gisburn: brendang isn't extinct
[00:47:00] <alanc> thought the community did that long ago
[00:47:05] * gisburn rembers what brendang did to drone and won't forget it...
[00:47:27] <flyingparchment> alanc: console keymaps still missing ;(
[00:47:50] <alanc> ah, nothing I can do there
[00:48:19] <alanc> get someone who cares to submit one to ON, maybe you can get brendang as your sponsor
[00:48:20] *** vmlemon has quit IRC
[00:48:29] <gisburn> alanc: just send the email of the "offender" who owns OS/Net keymaps...
[00:48:42] <alanc> gisburn: /dev/null
[00:48:43] <flyingparchment> did xvm integrate into nevada?  if so, does it include the vm bits?  so i can replace vmware with xvm?
[00:48:53] * gisburn attacks /dev/null
[00:49:06] <jmcp> flyingparchment:  yes, at the start of snv_75
[00:49:18] <alanc> like much of OS/Net, there is no owner, just people who put in changes when needed once every decade or two
[00:49:24] <e^ipi> flyingparchment: a model m has no keymap, you just use a butter knife to pop the keycaps off & change it however you like
[00:49:40] <flyingparchment> e^ipi: uhm, yes, but then i type ' on my keyboard and solaris thinks i typed q
[00:49:51] <flyingparchment> e^ipi: moving the keyscaps doesn't change what the keys mean ;)
[00:50:01] <e^ipi> that's why you reassign them
[00:50:12] <flyingparchment> e^ipi: i can't, because there's no dvorak keymap for the console
[00:50:17] <e^ipi> loadkeys(1) & dumpkeys(1)
[00:50:17] <flyingparchment> (not X)
[00:50:28] <flyingparchment> jmcp: where can i get that?  hg only, for now?
[00:50:35] <e^ipi> dumpkeys > keymap.dvorak; edit it, loadkeys dvorak.keymap
[00:50:40] <jmcp> flyingparchment: probably
[00:50:46] <flyingparchment> hm.
[00:50:49] <jmcp> the build closes today US/Pacific time
[00:50:50] <e^ipi> that's what I've done so far, but it's an unnecessary extra step
[00:51:08] <flyingparchment> i'd like to have solaris on my desktop, i think vmware is the only software i use that's missing
[00:51:22] * gisburn pokes alanc
[00:51:43] <e^ipi> my desktop is a sparc, I don't have much of a choice anyways
[00:51:53] <trs81> didn't vmare open source their tools recently?
[00:51:57] <e^ipi> i suppose i could use BSD or linsucks, but why bother
[00:52:06] <flyingparchment> trs81: that doesn't help me run vmware server on solaris..
[00:52:19] <e^ipi> or NeXTStep, apparantly
[00:52:23] <gisburn> DId anyone see johnlev recently ?
[00:52:28] <flyingparchment> e^ipi: 4.3BSD!
[00:52:38] <movement> gisburn: nope.
[00:53:08] <e^ipi> umm... it's a sparc
[00:53:20] <flyingparchment> 4.3 ran on sparc.  (or was it 4.4?)
[00:54:15] *** sommerfeld has quit IRC
[00:55:25] <flyingparchment> hm, also is there a way to make solaris ignore a particular cpu when booting?
[00:56:40] <jmcp> blacklist it
[00:58:26] <jbk> isn't blacklisting specific to a particular model/product line?
[00:58:30] <Gman> heh, thomas packaged my bro's zfs usb backup foo
[00:58:46] <jmcp> jbk: it can be, but the ideas carry over
[00:58:57] <jmcp> jbk: flyingparchment was vague
[00:59:07] <flyingparchment> i have two cpus.  i want cpu1 to be not used
[01:00:16] <CIA-26> danmcd: 6608255 zero_spi_check() doesn't properly handle split-mblk data packets
[01:00:17] <CIA-26> fr41279: 6558449 Add ECC support to NCP
[01:00:18] <CIA-26> johnlev: 6590353 ancient mtrr crud in i86pc/os/startup.c must die, 6604314 lock ordering problem can cause deadlock at bootup, 6604381 typo in LEVEL_SIZE call in i86pc's hat_share(), 6604444 HCK_PARTIALCKSUM packets from dom0 to domU fail checksum test in domU, 6605202 domU panics with 'bad mutex' from a freed data structure after 1h of xmstress, 6605536 xvdi_ring_has_unconsumed_responses() panic during suspend, 6606142 page_numtopp_alloc() uses page_re
[01:00:50] <jmcp> flyingparchment: man psradm
[01:02:06] <flyingparchment> jmcp: i don't see anything there about doing it automatically on boot
[01:02:49] <jmcp> flyingparchment: what sort of system do you have?
[01:03:02] <flyingparchment> jmcp: standard ia-32
[01:03:43] <e^ipi> you can probably turn it off in the bios too
[01:03:49] <e^ipi> but... why do you want to disable a CPU?
[01:03:56] <flyingparchment> i don't remember seeing it, i'll have another look
[01:04:00] <flyingparchment> e^ipi: it's broken :)
[01:04:12] <jmcp> flyingparchment: then you're outta luck unless you do it in the bios
[01:05:17] <flyingparchment> i guess i could replace the cpu.. i suspect athlon mps aren't sold anymore, though
[01:05:26] *** Gekz[sleep] is now known as Gekz
[01:05:51] *** Shiv_1 has left #opensolaris
[01:06:11] *** benr has joined #opensolaris
[01:06:11] *** ChanServ sets mode: +o benr
[01:07:29] <e^ipi> flyingparchment: an athlonXP and a piece of solder can become an athlonMP with very little difficulty
[01:08:00] <flyingparchment> will it actually work?  i seem to recall the mp being very picky about which cpus worked together
[01:08:07] <e^ipi> the difference between the XP and the MP was a single small laser cut in one of the leads on top
[01:08:33] <e^ipi> well, if it's a ceramic package
[01:09:02] <e^ipi> if it's a fibreglass package you'll need to fill inbetween the two with a not-conductive epoxy or glue of some sort, then reconnect them
[01:09:31] <e^ipi> since they cut down to the ground bus in the glass ones to prevent you from unlocking it
[01:09:47] <flyingparchment> hmm, i have a same clock speed XP here i think, maybe i'll have a fiddle
[01:10:18] <e^ipi> consult with the Google for more precise instructions
[01:16:17] *** charlesnw has quit IRC
[01:30:29] *** NikolaVeber has quit IRC
[01:39:39] <alanc> nv_75 X changelogs posted to http://opensolaris.org/os/community/x_win/changelogs/changelogs-nv_70/
[01:40:04] *** Teltariat has joined #opensolaris
[01:40:31] *** myrkr has quit IRC
[01:41:18] <gisburn> alanc: nv-70 ?
[01:41:35] <alanc> the page is nv_70-79
[01:41:47] <alanc> the URL just lists the first build in the set though
[01:43:04] <e^ipi> I melted the lid to my popcorn popper :-/
[01:45:12] <e^ipi> I guess it wasn't designed to have it's safety features removed & roast coffee
[01:45:46] <Gman> haha
[01:45:51] <jmcp> e^ipi: my mother-in-law found that using a pizza tray in a fan-forced over worked ok
[01:46:07] <jmcp> otoh, istr bonwick saying that he used a popcorn popper
[01:46:17] <jmcp> he didn't mention having neutered its safety features first though
[01:48:35] <e^ipi> I had trouble getting anything darker than a very light cinnamon roast so I googled it, and it turns out there's a thermal cutoff switch to prevent fires
[01:48:42] <e^ipi> so I had to bypass that & then it worked well
[01:48:50] <e^ipi> though, evidently it worked well enough to melt the lid
[01:50:28] <Gman> e^ipi, can you light fires where you live? :)
[01:51:33] *** Tpent1 has quit IRC
[01:52:17] *** yippi has quit IRC
[01:52:49] <Gman> if you can, i'd suggest a oil can style roasting mechanism
[01:52:55] <Teltariat> Hello folks.  I'd like to setup an x86 Solaris machine with root on ZFS.   Is this possible?
[01:53:02] <e^ipi> Gman: I don't think my landlord would appreciate having a campfire in the apartment
[01:53:13] <Gman> but think of the smell!
[01:54:38] * Gman heads for lunch
[01:54:39] *** Gman is now known as GmanAFK
[01:55:23] *** gavagai__ has joined #opensolaris
[01:55:36] <jbk> Teltariat: you can with recent solaris express releases, however it's a manual process at the moment
[01:55:39] <jmcp> Teltariat: there's doco on how to do it at http://opensolaris.org/os/community/zfs/boot/
[01:56:33] <Teltariat> jbk, jmcp: Thank you kindly.  I'm trying to do this on top of software RAID, so I'll have to figure that one first.  Again, thanks
[01:57:07] <gisburn> Did anyone saw comay recently ?
[01:57:10] <jmcp> raid1 is ok, but not raidz/raid5
[01:57:14] <jbk> one other thing, is your boot pool can only be mirrored or a single disk
[01:57:19] <jmcp> gisburn: he's a figment of your imagination
[01:57:26] <flyingparchment> /usr/sbin/kclient[14]: 101159 Segmentation Fault(coredump)
[01:57:27] <gisburn> jmcp: ?!
[01:57:34] <jmcp> gisburn: I'm teasing
[01:58:00] <Teltariat> jbk, jmcp: I'm shooting for 2 drives mirrored.  RAID 1 setup.  You guys mentioned that RAID-Z requires more?
[01:58:20] <jbk> you cannot currently boot off a raid-z pool
[01:58:29] <jmcp> and yes, raidz requires at least 3 drives
[01:58:40] <jbk> you can however have multiple pools on a given system
[01:58:42] <Teltariat> But a RAID 1 metadevice should be OK, right?
[01:59:05] *** Tpenta has joined #opensolaris
[01:59:10] <flyingparchment> raidz on 2 drives doesn't make sense anyway.  it'd just be a slow mirror
[01:59:50] * coffman puts some more needles in the doll
[01:59:59] <coffman> gar
[02:00:04] <coffman> gn8
[02:00:10] * gisburn twitches
[02:00:16] *** coffman is now known as coffman_zzz
[02:00:17] <gisburn> coffman: STOP IT
[02:00:20] <CIA-26> udpa: 6554813 Possible deadlock when enabling/disabling soft rings
[02:00:21] <CIA-26> michen: 6299080 nscd initialises name_service_door file with wrong permissions after JASS, 6587024 nscd fails to notice nsswitch.conf changes within a second of its start time, 6608351 nss2 test cases expose memory problems in the nss_dns and nss_nisplus backends, 6609869 nscd has problems if an ip address resolves to an empty hostname
[02:00:34] <coffman_zzz> gisburn: heh
[02:04:59] <gisburn> Who is benoit at sun dot com ?
[02:05:13] <jmcp> some bloke
[02:06:05] <gisburn> erm
[02:06:10] <gisburn> "bloke"=?!
[02:06:26] <alanc> "bloke" is australian for "dude"
[02:06:31] <jmcp> or guy
[02:06:51] <gisburn> positive, negative, neutral or GTH (go to hell) ?
[02:07:01] <jmcp> neutral
[02:07:11] <gisburn> erm
[02:07:24] <jmcp> "which one of you bastards called this bastard a bastard?!?!?!"
[02:07:35] <kjetilho> Benoit is a fairly common French boy's name
[02:07:35] <gavagai__> whats goin on
[02:07:36] <gisburn> I always imagined that if english-speaking people call you "dude" then it's time to dig a hole...
[02:08:16] <alanc> dude, you should not come to california then
[02:08:19] <alanc> 8-)
[02:09:04] <gisburn> alanc: I don't have to... I can send people to your lab to harvest stuff in it... much easier... :-)
[02:09:07] *** uebayasi has joined #opensolaris
[02:09:26] <alanc> too late, I already pulled the keyboards out of the lab
[02:09:36] <gisburn> grumpf
[02:09:42] * flyingparchment stabs kerberos
[02:10:03] <gisburn> $50000000/Inf+ for alanc's head on a plate!
[02:10:10] <alanc> at least I think they're german ones - they have keys like "Vordergrund" and "Ausschneiden"
[02:10:32] <alanc> (which really doesn't fit well on a keyboard keycap)
[02:10:49] <gisburn> alanc: what about the japanese thing ?
[02:11:01] <alanc> got that too
[02:12:09] <alanc> though those just say "Front", "Copy", "Paste", etc. in english
[02:12:14] * gisburn imagines the SMF people shivering in fear and dispear... =:-)
[02:14:48] <gisburn> dispear or despair ?
[02:15:30] <jmcp> despair
[02:18:52] <flyingparchment> hm.. the JDS fonts in b7something i have installed here look much uglier than in S10 on another system
[02:19:22] *** HCoyote has joined #opensolaris
[02:20:16] <flyingparchment> although changing to FreeSans is much better than the default Sans
[02:21:14] <HCoyote> Hello.  Anyone here have an x4500 and use it for serving NFS?  I'm trying to find out what others are seeing for average write performance is for streaming and random i/o to a ZFS backed filesystem to compare with what I'm seeing on mine.
[02:22:08] <HCoyote> s/is//
[02:24:30] <mw46> what are you seeing?
[02:25:12] <HCoyote> About 60MB/s with the backend setup as a zpool of mirrored zvols.
[02:25:53] <HCoyote> when I do iozone directly from the thumper in that same configuration (eg, ruling out the network and NFS), I get about an order of magnitude better, on average.
[02:26:23] <mw46> you're lucky:-(
[02:26:30] <HCoyote> I am?
[02:27:12] <HCoyote> (for the record, read performance is awesome out of this thing ... I can build a 4Gb LACP aggregate and almost get wirespeed for reads)
[02:27:36] <mw46> I have to find a solution for a SVN server (NFS client to a ZFS+NFS thumper) :-(
[02:27:58] <alanc> flyingparchment: check your hinting/anti-aliasing settings in the font preferences dialog in JDS
[02:28:56] <mw46> My first tests untar 32MB worth of small files on the NFS client takes 1:38 min. On the thumper directly < 1 sec.
[02:29:15] <flyingparchment> alanc: i did, it looked sane
[02:29:30] <HCoyote> I'm also going to be testing this with iscsi setups to see how it performs there.  And lustre, whenever the backend comes out.  Thing just sucks for NFS writes (which is bad because I'm trying to target this for replacing Netapps in certain situations)
[02:29:45] <flyingparchment> HCoyote: zvolz for nfs?  why?
[02:29:55] <HCoyote> flyingparchment:  hmm?
[02:30:26] <HCoyote> I don't understand that question.
[02:30:47] <flyingparchment> HCoyote: you said "About 60MB/s with the backend setup as a zpool of mirrored zvols."
[02:30:53] <flyingparchment> what are you using the zvols for?
[02:31:52] <HCoyote> I'm just describing how the nomenclature does.  a zpool is made up of a number of zvols that actually contain the physical disks.
[02:32:02] <flyingparchment> i think you mean zdevs
[02:32:13] <flyingparchment> or vdev rather
[02:32:23] <flyingparchment> a zvol is a block device backed by a zpool
[02:32:33] <jbk> are you using a nevada built on the box? if so, which build?
[02:32:47] *** sparvu has joined #opensolaris
[02:32:53] <jbk> errr build
[02:33:13] <HCoyote> I've used 10u3, 10u4, and build 70.  Performance actually seems worse the farther I go forward.
[02:34:11] *** hile_ has quit IRC
[02:34:14] *** GmanAFK is now known as Gman
[02:34:16] <HCoyote> flyingparchment:  hmm.  let me see if I can dig up where I found zvol.
[02:34:29] <HCoyote> (I very well may be using that term wrong)
[02:34:39] <jbk> that seems wrong.. have you posted anything to zfs-discuss?
[02:34:48] <jbk> i'm sure someone on there might be interested
[02:35:01] <HCoyote> I'm in contact with one of the NFS developers
[02:35:07] <HCoyote> inside sun
[02:35:14] *** Hunger- has quit IRC
[02:35:32] <HCoyote> though, I probably should post to zfs-discuss.
[02:36:58] *** Hunger- has joined #opensolaris
[02:38:22] <HCoyote> flyingparchment:  ah, yes, I guess I've been using the wrong term.  thanks for correcting me. :-)
[02:38:42] * HCoyote got it from someone inside sun.  <cough>  :-)
[02:38:45] *** estibi has joined #opensolaris
[02:39:38] *** Teltariat has quit IRC
[02:39:38] *** sahafeez has quit IRC
[02:39:38] *** peemus-home has quit IRC
[02:39:39] *** sickness has quit IRC
[02:39:39] *** estibi_ has quit IRC
[02:39:39] *** mw46 has quit IRC
[02:39:40] *** lasseoe has quit IRC
[02:39:40] *** MattAFC has quit IRC
[02:39:40] *** bondolo has quit IRC
[02:39:40] *** insomnia has quit IRC
[02:39:41] *** shou has quit IRC
[02:39:41] *** gdamore has quit IRC
[02:39:41] *** postwait has quit IRC
[02:39:41] *** Fullmoon_ has quit IRC
[02:39:41] *** jcea has quit IRC
[02:39:42] *** neoxed has quit IRC
[02:39:42] *** Gekz has quit IRC
[02:39:42] *** bda has quit IRC
[02:39:43] *** Auralis_ has quit IRC
[02:39:43] *** sparvu_ has quit IRC
[02:39:44] *** cmang has quit IRC
[02:39:44] *** djgregor has quit IRC
[02:39:44] *** jacobs has quit IRC
[02:39:45] *** tomww has quit IRC
[02:39:45] *** elektronkind has quit IRC
[02:39:45] *** moazamraja has quit IRC
[02:39:45] *** e^ipi has quit IRC
[02:39:47] *** pjlv has quit IRC
[02:39:47] *** AtomicPunk has quit IRC
[02:39:48] *** snuff-away has quit IRC
[02:39:48] *** quasi has quit IRC
[02:39:48] *** trs81 has quit IRC
[02:39:49] *** pjd- has quit IRC
[02:39:49] *** SunTzuTech has quit IRC
[02:39:49] *** dennis has quit IRC
[02:39:49] *** AtomicPnk has quit IRC
[02:39:50] *** sniffy has quit IRC
[02:39:50] *** the-decider has quit IRC
[02:39:50] *** mikaeld has quit IRC
[02:39:50] *** sparkleytone has quit IRC
[02:39:51] *** vipe has quit IRC
[02:39:51] *** Bartman007 has quit IRC
[02:39:52] *** Fetch has quit IRC
[02:39:52] *** coffman_zzz has quit IRC
[02:39:52] *** rootard has quit IRC
[02:39:53] *** xinkeT has quit IRC
[02:40:42] *** snuff-home has joined #opensolaris
[02:40:42] *** Teltariat has joined #opensolaris
[02:40:42] *** sahafeez has joined #opensolaris
[02:40:42] *** peemus-home has joined #opensolaris
[02:40:42] *** sickness has joined #opensolaris
[02:40:42] *** mw46 has joined #opensolaris
[02:40:43] *** lasseoe has joined #opensolaris
[02:40:43] *** MattAFC has joined #opensolaris
[02:40:43] *** bondolo has joined #opensolaris
[02:40:43] *** insomnia has joined #opensolaris
[02:40:43] *** shou has joined #opensolaris
[02:40:43] *** gdamore has joined #opensolaris
[02:40:43] *** postwait has joined #opensolaris
[02:40:43] *** Fullmoon_ has joined #opensolaris
[02:40:43] *** jcea has joined #opensolaris
[02:40:43] *** neoxed has joined #opensolaris
[02:40:43] *** Gekz has joined #opensolaris
[02:40:43] *** bda has joined #opensolaris
[02:40:43] *** Auralis_ has joined #opensolaris
[02:40:43] *** sparvu_ has joined #opensolaris
[02:40:43] *** cmang has joined #opensolaris
[02:40:43] *** djgregor has joined #opensolaris
[02:40:44] *** jacobs has joined #opensolaris
[02:40:44] *** tomww has joined #opensolaris
[02:40:44] *** elektronkind has joined #opensolaris
[02:40:44] *** moazamraja has joined #opensolaris
[02:40:44] *** e^ipi has joined #opensolaris
[02:40:44] *** pjlv has joined #opensolaris
[02:40:44] *** AtomicPunk has joined #opensolaris
[02:40:44] *** snuff-away has joined #opensolaris
[02:40:44] *** quasi has joined #opensolaris
[02:40:44] *** trs81 has joined #opensolaris
[02:40:44] *** pjd- has joined #opensolaris
[02:40:44] *** SunTzuTech has joined #opensolaris
[02:40:44] *** dennis has joined #opensolaris
[02:40:44] *** AtomicPnk has joined #opensolaris
[02:40:44] *** the-decider has joined #opensolaris
[02:40:44] *** rootard has joined #opensolaris
[02:40:44] *** sniffy has joined #opensolaris
[02:40:45] *** mikaeld has joined #opensolaris
[02:40:45] *** Fetch has joined #opensolaris
[02:40:45] *** xinkeT has joined #opensolaris
[02:40:45] *** coffman_zzz has joined #opensolaris
[02:40:45] *** Bartman007 has joined #opensolaris
[02:40:45] *** vipe has joined #opensolaris
[02:41:52] *** sparkleytone has joined #opensolaris
[02:42:01] *** mw46 has quit IRC
[02:42:10] *** neoxed_ has joined #OpenSolaris
[02:42:19] *** mw46 has joined #opensolaris
[02:43:49] *** e^ipi has quit IRC
[02:43:50] *** cmang has quit IRC
[02:43:53] *** e^ipi has joined #opensolaris
[02:44:08] *** cmang has joined #opensolaris
[02:45:45] *** quasi has quit IRC
[02:45:53] *** quasi has joined #opensolaris
[02:46:31] <alanc> hmm, just noticed SPARC graphics delivered OpenGL for Xorg in nv_74
[02:49:12] *** ichigo has quit IRC
[02:51:12] *** salamanders has joined #opensolaris
[02:54:13] *** sparvu_ has quit IRC
[02:56:17] <flyingparchment> how many people is 'sparc graphics' nowadays?
[02:56:23] <gisburn> 0.1
[02:56:26] <jmcp> alanc: 3?
[02:56:27] <gisburn> 0/Inf-
[02:56:52] <alanc> jmcp: about that
[02:57:01] <flyingparchment> full time?
[02:57:04] <alanc> yes
[02:57:16] <flyingparchment> huh
[02:57:21] <flyingparchment> didn't know there was that much to do ;)
[02:57:34] <alanc> in the actual driver/OpenGL team, not counting the 3Dserver/virtualization guys
[02:57:54] <HCoyote> graphics virtualization?
[02:57:59] <alanc> yes
[02:58:19] <HCoyote> thin client type stuff?  or something else?
[02:58:20] <alanc> put graphics cards in your servers in the backroom, doing the heavy 3-D rendering, display to thin clients
[02:58:24] <alanc> or PC's
[02:58:29] <HCoyote> ah
[02:58:30] *** paulf has joined #opensolaris
[02:58:37] <paulf> Hello
[02:58:49] <alanc> nvidia QuadroPlex type thingys
[03:00:22] <CIA-26> mlf: 6600169 CD/DVD Drive not detected in B72, 6600948 remove probe() entry point from ata driver
[03:02:32] *** neoxed_ has quit IRC
[03:02:44] *** snuff-away has quit IRC
[03:09:20] <paulf> Hmmm...that ZFS Automatic Backup thing looks pretty neat
[03:12:08] *** salamanders has quit IRC
[03:13:57] *** neoxed has quit IRC
[03:14:11] <Teltariat> Can you set everything up on a single drive, and then add another drive later, and have that automatically mirrored?
[03:21:30] <jamesd> yes
[03:23:01] <holcomb> dmesg
[03:23:04] <holcomb> doh
[03:23:05] *** rubymonk has quit IRC
[03:30:14] <jmcp> jbk: there's just been a putback (should show up here soonish) which your/our stuff will have to get re-whacked with
[03:30:23] <jmcp> jbk: I'll handle it, don't you worry about it
[03:34:22] <Teltariat> If RAID-Z requires a minimum 3 disks, must all of them be the same size?
[03:36:57] <paulf> I think it will use only the size of the smallest disk
[03:37:37] <paulf> but I'm definitely no expert here
[03:38:03] <jbk> jmcp: really? i'm surprised there's been any changed to that code..
[03:38:12] <jbk> but thanks..
[03:38:48] <jmcp> jbk: I had no idea there was anything else coming along in that area
[03:38:59] * jmcp grumbles
[03:39:04] <jmcp> for the moment, back to parallel scsi
[03:39:05] <jmcp> sigh
[03:39:29] <jbk> *shrug* if you get busy with other work, let me know and I'll take a look..
[03:39:45] <jmcp> ok
[03:40:37] <jmcp> this || scsi stuff seems to come and go - at the moment I've just had a realisation about the likely root cause of a nasty issue - but I need a scsi trace to confirm ... that could take a while
[03:40:38] <jmcp> sigh
[03:45:57] <jbk> heh.. this is amusing
[03:46:16] <jbk> i need to get my personal laptop serviced
[03:46:19] <jbk> so i call the support line
[03:46:26] <jbk> they give me a 'fedex drop off number'
[03:46:32] <jbk> go to a fedex store
[03:46:35] <jbk> no idea what that is
[03:46:44] <jbk> call fedex's 800 number, they have no idea either
[03:46:47] *** gisburn has quit IRC
[03:54:06] *** gisburn has joined #opensolaris
[04:00:19] <CIA-26> th199096: 6523051 root crontab should only execute nfsfind if it is present and executable (fix packaging)
[04:00:21] <CIA-26> dm120769: backout 6558449: breaks build
[04:02:38] <gisburn> Another beheading my dmarker ...
[04:02:46] <gisburn> s/my/by/
[04:03:08] * gisburn curses his typos and all the jokes which were screwed-up by them...
[04:03:17] *** lloy0076 has joined #opensolaris
[04:05:46] <jbk> who needs ecc support anyway? :)
[04:06:21] <dlg> me
[04:06:53] <jbk> (elliptic curve cryptography)
[04:07:12] <dlg> not me then
[04:10:02] *** alobbs has quit IRC
[04:10:03] *** MattAFC has quit IRC
[04:10:23] *** alobbs has joined #opensolaris
[04:10:32] *** MattAFC has joined #opensolaris
[04:12:18] *** binarycrusader has joined #opensolaris
[04:15:40] *** peemus-home has quit IRC
[04:16:33] *** lloy0076 has left #opensolaris
[04:19:26] *** rootard has quit IRC
[04:19:58] <elektronkind> hmm
[04:20:37] <elektronkind> could someone here give me a quick refresher (or point me to one) for panic mesgbuf output?
[04:20:44] <binarycrusader> ::msgbuf ?
[04:20:56] <elektronkind> yeah, that kind of stuff :)
[04:21:07] <binarycrusader> what do you mean?
[04:21:11] <binarycrusader> ::msgbuf gives you the output
[04:21:14] <binarycrusader> in mdb
[04:21:28] <elektronkind> I'm trying to figure out what the  %l0-3 and %l4-7 fields are in the output below each function in a stack trace
[04:21:50] <binarycrusader> http://blogs.sun.com/dmick/entry/diagnosing_kernel_hangs_panics_with
[04:22:28] <binarycrusader> Dunno if that will help or not.
[04:23:19] <elektronkind> oh, that stuff I'm already familiar with.... it's actually groking the output is what I'm after ;)
[04:26:00] <binarycrusader> Unfortunately, I don't have that answer.
[04:26:23] <libkeiser> look at ::dis for the symbols that correlate to the struct frames you're interested in
[04:26:44] <movement> elektronkind: you need to sit down with the SPARC architecture manual, the results of ::dis and some caffeine
[04:27:52] <elektronkind> figured that would come into the picture ;)
[04:28:00] <elektronkind> trying to track down a null ptr deref
[04:29:11] <jbk> those are probably gonna be function specific
[04:29:34] <flyingparchment> one day, kernel people will learn how to use source-level debuggerss
[04:31:13] <jbk> elektronkind: do you have a pastebin?
[04:31:41] <jbk> basically those are gonna correspond to whatever local variables are defined in the particualr function
[04:32:34] <elektronkind> jbk: yeah, I do... I only have a the meshbuf from the panic. This crash wasn't mine, so I'm waiting to hear from the guy that this did happen to if I can get a hold of his crash dumps.
[04:32:35] <elektronkind> http://opensolaris.pastebin.ca/722524
[04:33:23] <elektronkind> at the minimum, I need to prove that it wasn't my code that caused it ;)
[04:33:53] <jbk> do you have anywhere where the source for the 'nfs3_to_afs_call' function can be viewed?
[04:33:54] <libkeiser> oh jesus. a crash in the translator?
[04:34:09] <elektronkind> jbk: yeah, hold on
[04:34:24] <jbk> how many parameters does it take?
[04:34:29] <elektronkind> 4
[04:34:39] <libkeiser> jbk: http://www.openafs.org/cgi-bin/cvsweb.cgi/openafs/src/afs/afs_nfsdisp.c
[04:34:43] <jbk> ok, that rules out the obvious one
[04:34:52] <elektronkind> libkeiser is faster than me
[04:35:05] <jmcp> cpu_update_pct should have a (kthread_t *) as its first arg
[04:35:06] <libkeiser> i can type those urls off the type of my head anymore
[04:36:34] <jbk> i'm not even seeing where nfs3_to_afs_call is invoking cpu_update_pct
[04:36:40] <elektronkind> me neither
[04:37:04] <elektronkind> even in the ON UTS code it's called in the dispatcher only it seems
[04:38:27] <jbk> i don't suppose you know which op has a value of 3?
[04:38:30] *** Gman has quit IRC
[04:40:00] <libkeiser> elektronkind: what version of solaris had the panic?
[04:40:25] <elektronkind> NFSPROC3_LOOKUP according to nfs/nfs.h
[04:40:55] *** paulf has quit IRC
[04:41:22] <jbk> hmm...
[04:41:42] <jbk> this is already getting beyond my extremely limited kernel knowledge :)
[04:41:46] <jbk> though i'm wondering
[04:42:00] <jbk> i see no way nfs3_to_afs_call can invoke cpu_update_pct
[04:42:41] <elektronkind> no telling how that got on the stack
[04:42:49] <libkeiser> elektronkind: reason i ask is i was able to repro this exact stack trace about a year ago when i was trying to stress the nfs translator.  iirc, my conclusion was the code needed a rewrite, and i couldn't be bothered to spend time on it
[04:43:08] <elektronkind> libkeiser: this is s10u4
[04:43:14] <elektronkind> such a problem child its been
[04:43:29] <libkeiser> frankly, anyone attempting to run the translator on s10 has a death wish
[04:43:34] <elektronkind> but the reporter is saying that he's also having other issues with this particular box
[04:44:31] <libkeiser> first things first, he should start deploying the libafs64.nonfs.o kernel module
[04:45:09] <libkeiser> right now you're getting blamed for known bugs that have existed forever
[04:45:23] <elektronkind> probably.
[04:45:23] <jbk> it would be useful to see a disassembly of nfs3_to_afs_call to figure out what corresponds to 1b4
[04:45:54] <elektronkind> jbk: that's the position the fault occurred?
[04:46:30] <libkeiser> that's the offset in the symbol where the pc was last before the panic
[04:46:36] <elektronkind> gotcha
[04:48:27] <libkeiser> it's definitely having trouble correctly walk the frames.  afs_xserver+ad650c4 is a good example. that symbol isn't even executable
[04:48:28] <elektronkind> I'll sit tight and see if the guy is willing to send me dumps. if not, then "caveat translat0r"
[04:49:34] <elektronkind> I guess memory might have gotten stomped on
[04:49:37] *** alanc is now known as alanc-away
[04:50:44] <elektronkind> in the mean time, I'll advise the man to move away from the afs/nfs translator
[04:50:48] *** Shiv_1 has joined #opensolaris
[04:52:25] *** rubymonk has joined #opensolaris
[04:53:01] *** Highlander has joined #opensolaris
[04:53:50] <Shiv_1> I am trying to created a webrev. Am new to this tool & to hg. How to get the comments to appear in the html for the change entries
[04:53:57] <Shiv_1> http://opensource.shivakumar.info/webrevs/sfw-expect_webrev/
[04:54:57] *** linux_user400354 has quit IRC
[04:55:15] <binarycrusader> I had to edit the files manually
[04:55:30] <binarycrusader> It would be nice to know if there is a better way
[04:55:37] *** neoxed has joined #OpenSolaris
[04:56:07] *** lloy0076 has joined #opensolaris
[04:56:22] <Shiv_1> Oh, I was expecting that the nifty webrev tool picks comments applied to the files via hg. But the "hg add" did not provide a -m option
[04:56:27] *** lloy0076 has left #opensolaris
[04:56:34] <Shiv_1> and I was thinking that I might be missing something here.
[04:56:46] <Shiv_1> For now, I will hand edit.
[04:59:07] <movement> Shiv_1: you need to commit for webrev to pick up comments
[05:00:12] <binarycrusader> movement: how does it know which revision to use for the webrev then?
[05:00:25] <Shiv_1> Once I commit, I would probably need additional webrev command line options. Which are these options?
[05:00:34] <movement> binarycrusader: via hg outgoing basically
[05:00:52] <movement> Shiv_1: you don't. as long as you're using onnv-scm's webrev, a simple 'webrev' will do
[05:00:55] <binarycrusader> movement: ahh....
[05:04:10] <Shiv_1> Am getting a "no output"  with "webrev -p . -o ../test_webrev". Using onnv-scm that I pulled yesterday.
[05:04:53] <Shiv_1> This is after I do the commit.
[05:05:34] *** linux_user400354 has joined #opensolaris
[05:14:17] *** __hsilva has joined #opensolaris
[05:18:06] *** sfire||mouse has joined #opensolaris
[05:23:49] *** axisys has joined #opensolaris
[05:24:07] *** wnorrix_ has joined #opensolaris
[05:24:56] *** wnorrix has quit IRC
[05:27:05] *** _hsilva has quit IRC
[05:27:09] *** __hsilva is now known as _hsilva
[05:30:44] *** karrotx has joined #opensolaris
[05:32:35] <movement> Shiv_1: uh, that's because you're comparing the workspace against itself
[05:32:50] <movement> Shiv_1: I don't know why you added options when I already said you didn't need any.
[05:33:01] <movement> seems like you're just making work for yourself :)
[05:37:36] *** sfire||mouse has quit IRC
[05:46:00] *** Highlander is now known as WickedWicky
[05:46:24] <WickedWicky> Howedy
[05:54:26] *** binarycrusader has left #opensolaris
[05:54:29] <Shiv_1> movement: Thanks for the inputs. Got it to work.
[05:55:06] *** neoxed has quit IRC
[05:55:14] *** kaiwai has quit IRC
[05:56:14] *** edwardoc has joined #opensolaris
[05:56:37] <edwardoc> hey people, hows things?
[05:57:26] <edwardoc> ping
[05:57:51] *** movement has quit IRC
[06:01:50] *** movement has joined #opensolaris
[06:13:15] *** rubymonk has quit IRC
[06:24:52] *** tcuji`e has quit IRC
[06:26:45] <Doc> hmm.. so apparently pluggable security modules are good, but pluggable schedulers are bad
[06:26:46] <Doc> go Linus!
[06:33:58] *** ferret_0567 has quit IRC
[06:34:19] *** Shiv__ has joined #opensolaris
[06:35:23] *** wnorrix__ has joined #opensolaris
[06:36:45] *** shou has quit IRC
[06:45:09] *** wnorrix_ has quit IRC
[06:45:38] *** gisburn has quit IRC
[06:52:23] * psn2 tries to imagine replugging his scheduler
[06:54:30] *** Shiv_1 has quit IRC
[06:59:41] <elektronkind> doc: Linus is a cynical SOB. I can't count more than a day or two before he dereides something as crap
[07:00:56] <elektronkind> Linux "Sweeping generalism du jour" Torvalds
[07:03:03] <elektronkind> haha
[07:03:04] <elektronkind> "If LSM remains, security will never be a first class citizen of the
[07:03:05] <elektronkind> kernel."
[07:03:29] <elektronkind> no, that's because Quake 3 frames rates will always be the first class citizen of the linux kernel
[07:04:00] <WickedWicky> loading smashing pumpkins on my mp3 player and then off to the office for some bitch slapping
[07:05:06] <WickedWicky> ow and Reservoir Dogs soundtrack.. that's suitable for today
[07:06:28] *** mw46 has quit IRC
[07:06:30] <WickedWicky> pronto
[07:06:33] <WickedWicky> see you all later
[07:08:34] <elektronkind> good movie
[07:18:51] <edwardoc> ha ! so ture !
[07:19:03] *** edwardoc has quit IRC
[07:24:34] *** laca has quit IRC
[07:27:38] *** WickedWicky has quit IRC
[07:30:31] *** WickedWicky has joined #opensolaris
[07:30:33] <WickedWicky> b00
[07:46:48] <quasi> morning
[07:46:49] *** jcea has quit IRC
[07:50:41] *** slowhog has left #opensolaris
[07:53:17] *** gavagai__ has quit IRC
[07:53:30] *** baijiutong has joined #opensolaris
[07:53:41] *** WickedWicky has quit IRC
[07:56:35] *** estibi has quit IRC
[07:59:18] *** derchris^ has quit IRC
[07:59:33] *** derchris^ has joined #opensolaris
[08:00:23] <CIA-26> yc148097: 6567838 panic in nxge_start when dupb fails, 6573600 nxge reports ereport.io.device.intern_uncorr errors, 6599061 With jumbo frame enabled in nxge, the incoming packet count is not right for large pkts, 6599105 Jumbo does not work for packats larger than 8144 bytes when bcopy is forced
[08:00:24] <CIA-26> markfen: 6610537 ipseckey error output can get mangled on x86, 6610538 ipseckey can core dump with truncated input
[08:04:11] *** m0le has joined #opensolaris
[08:07:55] <sickness> morning all
[08:11:31] <jmcp> jbk: no need to worry about that putback / merge I mentioned - I misread the headers in the email - it went into the s10 feature gate, not nv
[08:11:32] <jmcp> :-)
[08:11:33] <jmcp> hi sickness
[08:11:36] *** Dar_HOME is now known as Dar
[08:13:26] <sickness> hey :)
[08:14:39] *** JWheeler has quit IRC
[08:17:42] <palowoda> I see the CDrom bug 6600169 is going in build 75.  Anybody know what the fix was.  Just curious to what the problem really was.
[08:20:55] *** sioraiocht has quit IRC
[08:24:49] *** JWheeler has joined #opensolaris
[08:26:58] <jmcp> palowoda: adds a check for statuses of 0x0, 0x7f and ATS_DF, basically
[08:28:19] *** tsoome1 has quit IRC
[08:28:37] <palowoda> jmcp: Ok thanks. Just for grins I switch my configuration around to the one that was failing when b75 hits.
[08:28:41] *** Shiv__ has quit IRC
[08:29:08] *** neoxed has joined #OpenSolaris
[08:29:41] <jmcp> palowoda: you like causing yourself pain? :-)
[08:31:24] <palowoda> Nah, I just couldn't believe it when I moved the cdrom to a different channel.  I'm just really curious that the status bits can fix it.  I'm sure it does I just wonder if it's my case.
[08:32:20] <palowoda> That and I like pain.  I do use Solaris you know. :)
[08:33:06] <jmcp> hehe
[08:33:07] <jmcp> :-)
[08:42:07] *** mikefut has joined #opensolaris
[08:51:42] *** myrkraver1 has joined #opensolaris
[08:54:36] *** Cyrille has joined #opensolaris
[08:54:46] *** Chihan has joined #OpenSolaris
[08:59:07] *** renihs has quit IRC
[08:59:22] *** cydork has joined #opensolaris
[08:59:47] *** tsoome has joined #opensolaris
[09:00:18] *** myrkraverk has quit IRC
[09:09:55] <trochej> Coffee?
[09:10:05] <trygvis> I'm good, thanks
[09:12:04] *** sahafeez has quit IRC
[09:13:49] *** benr has quit IRC
[09:13:57] <Chipdancer> yea, just had one ;)
[09:15:34] *** Snake007uk has joined #opensolaris
[09:16:32] *** sioraiocht has joined #opensolaris
[09:21:37] *** mustang has joined #opensolaris
[09:24:32] *** Erwann has joined #opensolaris
[09:36:15] *** WickedWicky has joined #opensolaris
[09:39:20] <trochej> :)
[09:44:30] *** sparvu has quit IRC
[09:55:05] *** peteh has joined #opensolaris
[09:57:33] *** NikolaVeber has joined #opensolaris
[09:59:02] *** MattAFC is now known as MattMan
[10:00:07] *** dmarker has quit IRC
[10:00:26] *** dmarker has joined #opensolaris
[10:02:10] *** sioraiocht has quit IRC
[10:09:49] *** Kernel86 has quit IRC
[10:09:58] <trochej> Hi, anyone here has iPhone?
[10:13:05] *** KermitTheFragger has joined #opensolaris
[10:14:28] *** sahafeez has joined #opensolaris
[10:14:29] *** JWheeler has quit IRC
[10:14:54] *** ICU has joined #opensolaris
[10:15:09] *** Kernel86 has joined #OpenSolaris
[10:15:48] <quasi> "Hi, anyone here has been screwed by jobs?" ;)
[10:16:13] <sahafeez> is that steve jobs or work in general
[10:16:40] <quasi> s/jobs/steve jobs/
[10:17:20] <sahafeez> hum, nope. do not have any problems. would like to beable to update my iphone to 1.1.1 but working on that...
[10:17:41] <e^ipi> i find it disconcerting that my lights flicker when i print something
[10:18:02] <sahafeez> laser printer?
[10:18:09] <e^ipi> yes
[10:18:56] <sahafeez> fuser drawing power - light on the same circut. move the light or laser and it will not happen
[10:19:17] <quasi> don't worry, your wiring might be close to burning, but it hasn't caught fire yet ;)
[10:20:30] <sahafeez> quasi what has poor steve done now
[10:24:52] <e^ipi> aside from intentionally bricking hacked iphones, added checksumming to prevent non-itunes media apps from accessing new ipods, all the while trying to convince media companies to not do that sort of thing to their own stuff?
[10:24:58] <trochej> I have a quesion born out of curiosity
[10:25:17] <trochej> Is this AT&T lock in a forever thing, or for a time (say two years)
[10:25:48] <sahafeez> bah, they did not go out of their way to brick the phone. heck they warmed everyone many times. the fact of the matter is that the sim unlock does screw up the state of the baseband...
[10:25:52] <Doc> like you'd by apple
[10:26:09] <sahafeez> if you want to really understand it come over to iphone-dev on irc.osx86.hu
[10:26:33] <sahafeez> the installed update warns you before you apply it.
[10:26:40] *** JWheeler has joined #opensolaris
[10:26:56] <sahafeez> they do not format the flash when the update is put on. it over writes the cfg files.
[10:27:24] <sahafeez> the iphonesimfree lock still works on the 1.1.1 code -the phone is unlocked....the anysimm is borked
[10:27:43] <sahafeez> one is hardward + software and the other is a buffer overflow
[10:28:44] <trochej> But I am not interested in theese, I just want to know if after some time said in agreement I can safely unsimlock the iPhone or if this is a forever thing
[10:29:07] <sahafeez> no idea. the talk is that at the end of 2 years you are good to go
[10:29:21] <sahafeez> or just get a 1.1.1 and downgrade to 1.0.2
[10:29:36] <trochej> HMmm
[10:31:28] <sahafeez> if you get a phone from the store now - you can check it. not all have the 1.1.1 on it
[10:33:17] <trochej> I don't. I'm in Europe
[10:33:19] <trochej> :)
[10:36:39] <coffman_zzz> haha
[10:36:45] *** coffman_zzz is now known as coffman
[10:36:55] *** uebayasi has quit IRC
[10:37:01] <coffman> thats the right punishment for buying an iphone
[10:40:46] <trochej> coffman: Eh? I didn't buy it. I'm just curious
[10:40:51] <e^ipi> it becoming an overpriced brick?
[10:41:32] <coffman> e^ipi: yes
[10:42:03] *** ludovicp has joined #opensolaris
[10:42:14] <e^ipi> once the initial "wow, teh flashy!!!11" wears off, I really wonder about the iphone's longevity
[10:42:19] *** Cass has left #opensolaris
[10:43:32] *** Cass has joined #opensolaris
[10:45:13] *** timsf has joined #opensolaris
[10:45:48] <timsf> hi all
[10:46:06] <trochej> 'lo
[10:46:11] <trochej> timsf: Nice backup tool
[10:46:49] <timsf> Thanks trochej
[10:48:23] <timsf> Planning on fixing the zfs auto snapshot service to take advantage of user-properties as well (so that the default instance will actually do something useful, rather than expecting people to create their own individual instances)
[10:48:30] <timsf> Just not enough hours in the day!
[10:48:52] *** migi has joined #opensolaris
[10:49:36] *** rasputnik has joined #opensolaris
[10:50:30] <quasi> timsf: 26 isn't enough for you?
[10:50:51] <e^ipi> replace sleep with caffeine
[10:50:54] <e^ipi> 'swhat I do
[10:52:45] *** sahafeez has quit IRC
[10:53:04] <rasputnik> if I've got a zone on ufs on a zvol, will Bad Things happen if I set shareiscsi=on ?
[10:54:29] <quasi> sounds like something that wouldn't work
[10:55:12] <rasputnik> quasi: I'd stop the zone and umount the zvol first. Then I should be able to float the zone across front end servers, right?
[10:55:41] *** snuff-away has joined #opensolaris
[10:55:50] * rasputnik wishes he hadn't given that second test machine awap
[10:56:01] <rasputnik> s/awap/away/
[10:56:29] <tsoome> iscsi will share an block device, so unless you gonna mount ufs on client system, it will destroy your data
[10:57:13] * quasi nods
[10:57:50] <rasputnik> tsoome: yeah, that was the plan. Should work, shouldn't it?
[10:58:23] <tsoome> well, you can mount ufs if its not mounted already
[10:58:44] <quasi> as long as you've shut down anything that uses it locally, it probably won't kill your data ;)
[10:58:45] <tsoome> and if architecture will match
[10:59:14] <rasputnik> tsoome: sure, it'd only be on one frontend at a time. Endianness is a bitch, but until zones on 'real' zfs is supported, waddayagonnodo
[10:59:38] <rasputnik> think I'll give it a go.
[11:01:03] <rasputnik> I already do something similar with a bunch of zones on an iscsi-backed storage pool. this will be a lot clankier but might actually be supported (since it's zones on ufs)
[11:01:56] <tsoome> why dont use zfs on that iszsi device?
[11:02:15] <rasputnik> tsoome: because then I'd have to run zones on zfs
[11:02:23] <tsoome> yes?
[11:02:48] <tsoome> if you have solaris 10, it's just fine
[11:03:35] <rasputnik> tsoome: still not supported according to all the update 4 docs I read
[11:03:40] <tsoome> ?
[11:04:17] <rasputnik> well, to be specific, updating isn't supported. and zoneadm clone doesn't take advantage of ZFS clones
[11:04:36] *** Fullmoon_ has quit IRC
[11:04:38] <tsoome> ah, ok i see
[11:05:11] <rasputnik> and more to the point, I couldn't install a zone on ZFS zonepath yesterday. the postgresql packages crap out because their checkinstall scripts are b0rked. installing to UFS-on-ZFS works, however
[11:05:12] *** Arnald has joined #opensolaris
[11:05:26] <rasputnik> I'd raise a bug but I don't have a support contract for s10 on x86
[11:05:55] <rasputnik> to clarify, I'm talking about s10u4. good old Nevada is still fine.
[11:11:06] *** snuff-home has quit IRC
[11:13:06] * rasputnik clones zones by hand
[11:15:25] <tsp> is anyone running php5 with the solaris provided apache?
[11:22:15] *** bnitz has joined #opensolaris
[11:28:17] *** Fullmoon has joined #opensolaris
[11:28:46] *** kloczek has joined #opensolaris
[11:30:38] *** sahafeez has joined #opensolaris
[11:35:02] *** kloczek has quit IRC
[11:35:49] *** kloczek has joined #opensolaris
[11:37:06] *** crib has quit IRC
[11:37:26] *** kloczek has joined #opensolaris
[11:37:42] *** kloczek has quit IRC
[11:37:51] *** crib has joined #opensolaris
[11:40:10] *** kloczek has joined #opensolaris
[11:43:43] <tomww> tsp: yes, it's compiled as module against the Apache from Nevada build sometink
[11:43:48] <tomww> *g
[11:44:45] <tomww> Look at http://tinyurl.com/2lafpo
[11:46:11] <ofu> ah, sun has learned things... http://www.sun.com/servers/x64/x4150/wp.pdf shows chipset diagrams and is not even sun-internal documentation ;-)
[11:47:37] <e^ipi> don't tell them
[11:47:57] <e^ipi> or it'll get locked down faster than you can say "hey, where'd it go?"
[11:49:18] <ofu> its a pity sun thinks this kind of documentation should remain internal
[11:52:19] *** kloczek has quit IRC
[11:52:52] *** kloczek has joined #opensolaris
[11:55:20] <ofu> there is a sata option for x4150/x4450? cool %-)
[11:57:05] <ofu> oh, the raid-controller is an adaptec-model? interesting
[11:57:15] <dlg> really?
[11:57:19] * dlg look unimpressed
[11:58:01] <ofu> i never used adaptec-controllers with solaris... dont know their problems
[11:59:04] * dlg prefer lsi gear
[12:00:49] <Berny> some adaptecs just work... some don't some work halfways
[12:09:31] *** wnorrix__ has quit IRC
[12:16:59] *** yess has joined #opensolaris
[12:17:16] <yess> Hello.
[12:20:23] *** Teltariat has quit IRC
[12:21:30] <rasputnik> is it normal to be getting lots of 'showmount: xxx: program not registered' errors rebooting a zone, or have I been a bit over-zealous in disabling things?
[12:23:34] <yess> I'm experimenting with priocntl(2) opensolaris system call. What does realtime time quantum signal mean? That's one of the values that can be set with that system call.
[12:25:51] *** kaiwai has joined #opensolaris
[12:27:31] <kaiwai> hmm
[12:29:58] *** tcuji`e has joined #opensolaris
[12:31:55] *** redbeard0x0a has quit IRC
[12:32:23] *** alanbur has joined #opensolaris
[12:34:51] *** yarihm has joined #OpenSolaris
[12:38:24] <yess> I see, it is the signal that is send when the realtime time quantum expires.
[12:44:54] <jmcp> dlg: the embedded SATA controller is from Intel, the two SAS hbas are X-options, one from LSI the other from Adaptec
[12:45:25] <dlg> wow
[12:46:07] <dlg> ahci?
[12:52:36] *** The-spiki has joined #opensolaris
[12:55:00] <jmcp> I assume so
[12:59:13] <jmcp> wow
[12:59:18] <jmcp> serious swapping going on for me then
[12:59:19] <dlg> ?
[12:59:50] <dlg> ah
[13:00:45] <Doc> evening jmcp
[13:01:21] *** jonkri has joined #opensolaris
[13:01:26] <jmcp> hi Doc
[13:02:35] *** tazz has joined #opensolaris
[13:03:12] <kaiwai> hi jmcp
[13:03:19] * kaiwai gives jmcp a high five
[13:03:38] <jmcp> hi kaiwai
[13:03:41] * jmcp hi-5s back
[13:03:57] * alanbur wishes the colonials would go to bed
[13:03:59] <alanbur> ;-)
[13:04:16] <trochej> :)
[13:04:18] <kaiwai> wish I could shout a really helpful engineer a case of beer
[13:04:20] *** Gman has joined #opensolaris
[13:04:21] <jmcp> yeah, we need our beauty sleep to thump you lot in the match on Saturday/Sunday !
[13:04:38] <jmcp> hi alanbur
[13:04:40] <jmcp> :-)
[13:04:41] <alanbur> hi
[13:04:41] <kaiwai> lol
[13:04:58] <Doc> kaiwai: i'm a really helpful engineer - i'll send you my address
[13:05:03] <kaiwai> lol
[13:05:16] <kaiwai> just an engineer who fixed the cd/dvd drive issue
[13:05:37] <kaiwai> now, I'll give the man who supports my webcam - $500 cash and a case of beer :)
[13:05:44] <Doc> i'm getting very close to the point where i'm going to get all of the people that have promised me a beer together and get them to buy be a brewery
[13:05:55] <kaiwai> lol
[13:06:03] <alanbur> There's a brewery on the street where I live...
[13:06:04] <Doc> s/be /me /
[13:06:22] <jmcp> I used to live next to a brewery
[13:06:23] <WickedWicky> kaiwai is scoring potential dates I see
[13:06:28] <alanbur> http://www.howardtownbrewery.co.uk/
[13:06:33] <kaiwai> na, just my webcam to get working
[13:06:43] <Doc> alan: it's perfect - it's already named after me!
[13:06:43] <kaiwai> I'm a simple man with simple needs
[13:06:53] <alanbur> ;-)
[13:07:16] <alanbur> It used to be in Howard Town, but it burnt down.
[13:07:23] <WickedWicky> depends what you define simple... casual sex can be a real gift to some ppl
[13:07:34] <WickedWicky> for some ppl
[13:07:56] <kaiwai> WickedWicky: for me, simple is; working webcam, beer and a shag
[13:07:59] <kaiwai> nothing more :)
[13:08:16] <WickedWicky> Well Doc, nevermind that address I guess :P
[13:09:12] *** Fullmoon has quit IRC
[13:16:37] *** ichigo has joined #opensolaris
[13:21:46] <Gman> http://www.geekz1.com/images/blog/moderating.jpg [chuckle]
[13:22:10] * Gman wants one of those for the opensolaris lists
[13:22:28] <alanbur> oss-app1 is titsup again, sigh...
[13:24:22] *** FelipeBare has joined #opensolaris
[13:33:14] *** redbeard0x0a has joined #opensolaris
[13:35:41] *** tazz has quit IRC
[13:35:42] *** gaurav__ has joined #opensolaris
[13:36:37] *** jonkri has quit IRC
[13:39:47] <kaiwai> lol
[13:39:50] <kaiwai> what a cute pussy
[13:40:30] *** rasputnik has quit IRC
[13:40:45] *** cmihai has joined #OpenSolaris
[13:44:41] *** Snake007uk has quit IRC
[13:46:59] *** Snake007uk has joined #opensolaris
[13:48:52] <Vanuatoo> rhytmbox crashes trying to load album cover. I use SXDE 9/07
[13:49:03] <kaiwai> yeap, thats a known issue
[13:49:20] <kaiwai> known issue since it was merged
[13:49:25] <Vanuatoo> any workaround?
[13:49:33] <kaiwai> reported it against 2.18 and still merged the damn thing
[13:49:39] <kaiwai> nope, apart from disabling the plugin
[13:49:39] <Vanuatoo> do not run rhytmbox?
[13:49:57] *** yess has left #opensolaris
[13:50:08] <kaiwai> yeap, I run it
[13:50:10] <Vanuatoo> kaiwai: thanks
[13:50:12] <kaiwai> hence, I disabled the plugin
[13:50:18] <Vanuatoo> I already disabled Cover art plugin
[13:50:23] *** gaurav__ has quit IRC
[13:50:30] <kaiwai> strange, it shouldn't crash then
[13:50:37] *** gaurav__ has joined #opensolaris
[13:51:08] <Vanuatoo> kaiwai: no it does not crash anymore
[13:51:17] <kaiwai> ok
[13:51:24] <kaiwai> good boy
[13:51:30] *** hile_ has joined #opensolaris
[13:51:31] * kaiwai gives Vanuatoo a biscuit
[13:53:33] * Vanuatoo swings his tail
[13:54:24] <kaiwai> what a good dvd player to use for Solaris?
[13:54:32] <dlg> macos
[13:54:39] * dlg giggle
[13:54:45] <Vanuatoo> kaiwai: I compiled mplayer yesterday
[13:54:52] <Vanuatoo> It can play dvds IIRC
[13:56:01] <cmihai> kaiwai, VLC if you want better menus.
[13:56:12] <kaiwai> eww
[13:56:19] <kaiwai> dependency hell
[13:57:16] <JWheeler> Or, just use the Blastwave package?
[13:57:35] <Vanuatoo> Blastwave would be the fastest way
[13:58:09] * kaiwai wails in pain
[13:58:23] <kaiwai> Blastwave, everything went pear shaped onme
[13:58:26] <JWheeler> feel free to complain about the bloat, but pain?
[14:00:04] <hile_> talk to delewis..
[14:00:16] <hile_> he's an active mplayer developer
[14:00:30] <JWheeler> and a really nice guy, I might add
[14:03:22] <kaiwai> I guess its just annoying having to compile all the wxwidgets crap
[14:04:26] *** cmihai has quit IRC
[14:07:37] <kaiwai> oh well, got GCC 4.2.1 working nicely now
[14:08:42] *** gaurav__ has quit IRC
[14:09:02] *** tazz has joined #opensolaris
[14:10:35] *** cmihai has joined #OpenSolaris
[14:13:49] <kaiwai> oh well, lots of fun :) ffmpeg works perfectly
[14:16:24] *** master_of_master has joined #opensolaris
[14:23:24] *** tazz has quit IRC
[14:23:39] *** tazz has joined #opensolaris
[14:25:07] <kaiwai> hmm, this should be fun, wxwidgets compiling
[14:26:24] *** karrotx has quit IRC
[14:27:08] *** gaurav__ has joined #opensolaris
[14:27:26] *** tazz has quit IRC
[14:32:25] *** master_o1_master has quit IRC
[14:37:04] *** Somethingelse has quit IRC
[14:37:18] *** RainDoctor has joined #opensolaris
[14:56:43] *** yarihm has quit IRC
[14:57:07] *** Gman has quit IRC
[15:05:35] *** yarihm has joined #OpenSolaris
[15:06:18] *** gaurav__ has quit IRC
[15:09:19] *** tinman2k has joined #opensolaris
[15:11:11] *** Gman has joined #opensolaris
[15:23:10] *** yarihm has quit IRC
[15:24:45] *** kFuQ has quit IRC
[15:25:16] <kaiwai> hmm, and again, Adobe proves it sucks
[15:25:21] <kaiwai> and no better than Microsoft
[15:25:35] <Vanuatoo> kaiwai: why?
[15:25:47] <kaiwai> http://blogs.zdnet.com/Gardner/?p=2549
[15:25:54] <kaiwai> Adobe Air
[15:25:59] <Gekz> air?
[15:26:06] <trochej> CO2
[15:26:09] <trochej> STOP
[15:26:23] <trochej> I'm not thinking today
[15:27:49] <kaiwai> same shit "we hate Microsoft and yet we only support windows"
[15:28:03] *** Doc has quit IRC
[15:28:27] <cmihai> I don't get it.
[15:28:35] <cmihai> What's wrong with Adobe AIR?
[15:28:44] <cmihai> Seems to be competition for Microsoft Silverlight.
[15:28:57] <cmihai> And by the looks of it, they both run on the same platforms (Windows, Linux, MAC)
[15:29:24] <kaiwai> which means it won't be available on Solaris :(
[15:29:34] <cmihai> Well, there is JavaFX
[15:29:41] <cmihai> It's virtually the same shit(TM).
[15:30:18] *** gammon has joined #opensolaris
[15:30:19] <kaiwai> JavaFX, nice idea, shit IDE
[15:30:45] <cmihai> IDE? What does the IDE have to do with anything?
[15:30:53] <cmihai> Besides, there is no _default_ JavaFX IDE.
[15:31:11] <kaiwai> there is no JavaFX ide full stop
[15:31:43] <trochej> there is no.
[15:31:45] <cmihai> Huh
[15:31:54] <cmihai> Like I said, what does the IDE have to do with ANYTHING?
[15:32:21] <cmihai> There's plugins for Eclipse and NetBeans and all that for JavaFX, but the support is in the language, NOT the IDE. Shit, you can use ed(1) for all I care.
[15:32:39] <kaiwai> If hthere isn't a good ide, people won't use or develop for it
[15:32:47] <cmihai> Bullshit.
[15:32:50] <Cyrille> o rly?
[15:32:51] <trochej> shit
[15:32:53] <kaiwai> there fore, if people don't use it, we're screwed between adobe and Microsoft
[15:33:10] <cmihai> Most developers just use a good text editor. See EMACS, VIM, JEdit, ed(1),...
[15:33:43] <Cyrille> people wrote in various languages before IDEs even existed.
[15:33:56] <hile_> Even most developers i know at MSFT use fucking vi
[15:33:58] <cmihai> Besides, unless you get into UML, reverse engineering and CASE tools you really don't need more.
[15:34:11] <hile_> nobody uses the Visual Studio editor
[15:34:18] <kaiwai> I do
[15:34:23] <cmihai> Yes, but you suck.
[15:34:27] <trochej> kaiwai: You are gay.
[15:34:31] <kaiwai> I like visual studio *kaiwai goes and hides under a rock*
[15:34:35] <cmihai> Heh
[15:34:38] <hile_> okay let me rephrase that...
[15:34:40] <Berny> .oO(big puffy)
[15:35:02] <hile_> nobody who isn't thieving their oxygen supply uses the VS editor :)
[15:36:08] <cmihai> Meh, IDE's are more popular with object oriented programming and such. Usually the Java and C++ app developers. Not too popular with the C system developer crowd.
[15:36:13] <cmihai> And it's usually for the CASE tools.
[15:36:46] <kaiwai> oh well, apparently the AIR framework is opensource; hopefully that'll make things nice for alternative platforms in the future
[15:36:50] <Cyrille> IDEs are also popular for their debuggers.
[15:36:51] <cmihai> Besides, beats using endless refactoring scripts in EMACS or whatever
[15:36:58] <trochej> IDEs are nice, if you can model your UML and then generate stub code
[15:37:11] <cmihai> Yep, code generation and reverse engineering.
[15:37:28] <cmihai> Cyrille, yeah, but most people can use external / 3rd party / stand alone debuggers :-).
[15:37:36] <Gekz> gdb
[15:38:14] <alanbur> NetBeanz ;-)
[15:38:50] <alanbur> The Tomcat debugger is awesome
[15:39:05] <cmihai> And the Emacs NUTS have elisp plugins for everything from CVS/SVN/HG to debugging and code refactoring. So unless they're doing visual development (which is kind of rare in C and even Java), UML code generation and reverse engineering is the ONLY reason you'd want an IDE.
[15:39:17] <cmihai> I'm more of a Oracle JDeveloper guy ;-)
[15:39:22] <kaiwai> moonlight will be interesting - too bad Sun will never suck up its pride and bundle it with JDS
[15:39:28] <cmihai> NetBeans is just heavy, bloated, slow and ugly :-)
[15:39:34] <flyingparchment> cmihai: or maybe you don't like emacs..
[15:39:47] <cmihai> Nah, I like Emacs :-)
[15:39:48] <alanbur> wrong, wrong and wrong.
[15:39:59] <alanbur> EMACS sukz
[15:40:08] <cmihai> I like ed(1) too. Even vi(m). It's nano/pico that I hate.
[15:40:13] <Cyrille> now that's going to be a productive discussion...
[15:40:16] <Gekz> emacs fails
[15:40:29] <alanbur> I ran it once, about 20 years ago - couldn't figure out how to stop it.
[15:40:34] <Gekz> lol
[15:40:42] <cmihai> Oh boy. Here we go again.
[15:40:48] * PerterB sighs
[15:40:55] <Gekz> Time is linear!
[15:40:56] * cmihai locks alanbur's terminal with ed(1).
[15:40:57] <alanbur> There's probably a machine in a dark room somewhere that is still running that EMACS session...
[15:41:23] <cmihai> You really should keep a log of ed(1) sessions once you set it as the default editor on a public machine. It's so fun :-[
[15:41:46] * alanbur chuckles - editor wars are such pointless fun!
[15:41:47] <trochej> kaiwai: moonlight is a piece of crap, because of strings atteched
[15:42:07] *** Doc has joined #opensolaris
[15:42:19] <alanbur> EMACS users use garlic & wooden crosses to get anything done
[15:42:31] <kaiwai> trochej: Sun has an arrangement with Microsoft
[15:42:36] <cmihai> So?
[15:42:47] <cmihai> Mono sucks :-)
[15:42:58] <Cyrille> I'm sure emacs works fine for anyone with sixteen fingers.
[15:43:15] <kaiwai> cmihai: no it doesn't, if it did suck, you'd think I'd be on here? :P
[15:43:16] <trochej> kaiwai: Mono sucs, more even: .net sucks
[15:43:31] <alanbur> EMACS - the editor of our tentacled overlords!
[15:43:38] <kaiwai> bah, vi all the way
[15:43:52] <trochej> cat | grep | sed
[15:44:02] <alanbur> I bet there's at least one person in here who is IRCing via EMACS... ;-)
[15:44:04] <cmihai> That's pointless use of a cat.
[15:44:12] <trochej> cmihai: True
[15:44:35] <PerterB> almost certainly a pointless use of grep too
[15:44:36] <trochej> cmihai: But when I got my skillz, I didn't know that and old habits die hard :)
[15:44:37] <kaiwai> alanbur: I'm surprised if someone isn't running Emacs as an OS (given its size) :P
[15:44:49] <Cyrille> well I've read once that EMACS is a fine OS which lacks a good editor.
[15:44:57] <alanbur> ;-)
[15:44:59] <cmihai> alanbur, yeah, ERC is nice. It's also by default in Emacs 22 :-).
[15:45:02] *** laca has joined #opensolaris
[15:45:06] <cmihai> It's the only decent IRC client I could get on AIX.
[15:45:10] <trochej> Cyrille: I've read it no less than a dozen times
[15:45:27] <Cyrille> Trident, I sort of stopped registering it once I'd seen it ;-)
[15:45:32] <alanbur> cat - I ONZE UR EDITZ
[15:45:50] *** gammon has quit IRC
[15:46:01] <Cyrille> not Trident sorry, trochej.
[15:46:12] *** gammon has joined #opensolaris
[15:46:18] <trochej> Cyrille: I kind of understood :)
[15:46:21] *** gammon has left #opensolaris
[15:46:24] <kaiwai> hmm, xchat isn't too bad IMHO
[15:46:39] <cmihai> Meh
[15:46:41] * trochej bashes kaiwai with loose packets
[15:46:45] <libkeiser> if it can't be put inside screen, it ain't worth having
[15:46:55] <Gekz> irssi, weechat
[15:47:01] <cmihai> ssh + Emacs ERC / irssi + screen = love
[15:47:03] * kaiwai says to trochej: "you've been a bad boy, come to my room"
[15:47:12] <Gekz> weechat is ncurses xchat
[15:47:13] <Gekz> lol
[15:47:30] <trochej> kaiwai: You a school principal? :)
[15:47:34] <cmihai> Oh noes, run away, run away, it's teh ghey, it's teh ghey
[15:47:42] <trochej> You certainly sound like one.
[15:47:42] <cmihai> trochej, run and don't look back!
[15:47:43] * kaiwai is shocked!
[15:47:46] <kaiwai> I'm a catholic priest :P
[15:48:11] <alanbur> child abuser?
[15:48:15] <cmihai> That's just his way of saying he's a perverted child abuser.
[15:48:18] *** Fish has joined #opensolaris
[15:48:31] * alanbur was brought up by selfsame catholic priests...
[15:48:51] <Fish> hello
[15:49:04] <kaiwai> hello fish
[15:49:49] <trochej> Hmmm
[15:50:01] <trochej> alanbur: Your parents were priests?
[15:50:13] <alanbur> No
[15:50:38] *** shou has joined #opensolaris
[15:51:58] *** millhouse has joined #opensolaris
[15:52:02] <millhouse> hey
[15:52:16] *** cmihai has quit IRC
[15:52:38] *** Dar is now known as Dar_HOME
[15:53:52] <millhouse> what tools come with Solaris/OpenSolaris to monitor UPS's?
[15:54:26] <trochej> millhouse: cat, grep, sed
[15:54:32] <millhouse> haha
[15:54:34] <millhouse> nice
[15:55:16] <Berny> whatever your ups maker gave you :-)
[15:55:30] <Berny> or nut (from blastwave)
[15:55:30] <trochej> Fok, Mac OS X is alien
[15:55:40] <trochej> It couldn't possibly be made by humans
[15:56:26] <Berny> how come?
[15:56:36] <trochej> It's simply different than anything
[15:56:58] <trochej> But mostly its overstuffed
[15:57:33] *** gammon has joined #opensolaris
[15:59:04] <trochej> It does ten things, when you request one, and nine are animations around the core task
[15:59:16] <trochej> It's overshadowed, oversugared and overeverything
[15:59:48] <PerterB> but pretty
[16:00:01] <trochej> Depends on what you call pretty
[16:00:10] <trochej> I prefer clean and simple desings and call them pretty
[16:00:36] *** gammon has left #opensolaris
[16:00:50] <Berny> most find it sexy
[16:00:54] <tinman2k> we use upsilon for our UPS
[16:01:09] * Berny screams and runs when he sees a mac
[16:01:27] <tinman2k> but it is buggy with sol 10
[16:01:34] *** Triskelios has joined #opensolaris
[16:05:48] *** Gekz is now known as Gekz[sleep]
[16:07:34] <trochej> http://www.virtualbox.org/
[16:07:35] <trochej> Fok
[16:07:55] <trochej> Pity that I am not a good coder, I'd start project to write it for Solaris
[16:10:17] <quasi> trochej: why isn't xen enough
[16:11:19] <millhouse> we're running CyberPower for our UPSs
[16:11:20] *** tinman2k has quit IRC
[16:11:49] *** alanbur has left #opensolaris
[16:13:21] <trochej> quasi: I don't have the hardware vm flags
[16:17:44] <trochej> ok
[16:17:45] <trochej> [d]
[16:17:46] *** hile_ has quit IRC
[16:19:09] *** jonkri has joined #opensolaris
[16:25:06] <SunTzuTech> trochej: I ported the kbuild stuff to solaris since that's what they use to build virtualbox
[16:25:37] <SunTzuTech> but then got caught up in qemu so I never got back to the actual porting of virtual box
[16:25:54] *** tinman2k has joined #opensolaris
[16:27:35] *** calumb has joined #opensolaris
[16:28:09] *** Snake007uk has quit IRC
[16:37:28] *** Gman has quit IRC
[16:39:17] *** sommerfeld has joined #opensolaris
[16:39:18] *** ChanServ sets mode: +o sommerfeld
[16:39:50] *** Gman has joined #opensolaris
[16:46:00] *** [R]itchie|away has quit IRC
[16:51:56] *** kaiwai has quit IRC
[16:56:16] *** Gman has quit IRC
[17:17:53] <elektronkind> nom nom
[17:18:32] <axisys> zfs snapshot fs@now complains that snapshot exist.. is there a way to overwrite on sol 10 6/06 ?
[17:18:53] <elektronkind> delete it, then make a new one
[17:19:00] <elektronkind> there is no "snapshot overwrite"
[17:19:22] *** Kush- has left #opensolaris
[17:20:02] *** gekln has joined #opensolaris
[17:24:54] <axisys> elektronkind: i have a cronjob that deletes the old snapshot
[17:27:26] <axisys> elektronkind: i am scared of deleteing it as part of a cronjob
[17:27:27] *** ICU has quit IRC
[17:27:36] <axisys> http://opensolaris.pastebin.ca/723135
[17:27:49] <axisys> if I add a zfs destroy right above the snapshot
[17:28:05] <axisys> there might be chance that it destroys the original fs
[17:28:19] <axisys> too risky!
[17:29:11] *** millhouse has quit IRC
[17:29:12] <elektronkind> you might want to add some sanity checking to that script then
[17:29:33] <elektronkind> or use a different date format other than the day of week
[17:29:44] <axisys> i need only weekly snapshot
[17:30:02] <axisys> so with this i dont have to have a second script to rotate old ones
[17:30:18] *** cmihai has joined #OpenSolaris
[17:30:21] *** pauliukas has joined #opensolaris
[17:30:27] <axisys> s/weekly/weeklong/
[17:30:32] <pauliukas> Heya guys. Would anyone mind sharing with me their amazing super dooper PATH configurations?
[17:30:54] *** cmihai has quit IRC
[17:31:07] <elektronkind> axisys: I still don't see your problem... addition some check logic to that script should be good enough
[17:31:10] *** cmihai has joined #OpenSolaris
[17:31:14] <axisys> i guess i could check if $date has a value.. before calling the destroy
[17:31:23] <elektronkind> that would be my suggestion :)
[17:31:28] <axisys> elektronkind: yep
[17:32:32] <axisys> actually come to think of it.. there is a `@' in the middle.. so even without any value in $date it cannot ignore the `@' and hence I should be safe
[17:32:55] <axisys> discussion helps :-)
[17:33:07] <elektronkind> PATH=/usr/bin:/usr/sbin:/usr/sfw/bin:/usr/gnu/bin:/usr/X11/bin:/usr/ccs/bin:/opt/SUNWspro/bin:/usr/local/bin:/usr/local/sbin
[17:33:12] <elektronkind> axisys: you're right about that
[17:33:26] <elektronkind> always good to test, too
[17:33:42] <elektronkind> create yourself a trash fs and test away
[17:33:43] *** dlynes_laptop has joined #opensolaris
[17:33:49] <pauliukas> elektronkind: Thanks.
[17:35:53] *** BadKarma has quit IRC
[17:36:17] *** tsoome has quit IRC
[17:39:17] *** theRealballchalk has joined #opensolaris
[17:39:53] *** hile_ has joined #opensolaris
[17:39:56] <elektronkind> pauliukas: some here might call me a barbarian for not having /usr/xpg4/bin in my PATH ;)
[17:40:14] <pauliukas> And what is xpg
[17:40:15] <elektronkind> but that's another option for you to put in front of /usr/bin
[17:40:34] <pauliukas> I'm trying to remember where they keep the commonly used ucb binaries.
[17:40:48] <cmihai> In ucb...
[17:40:51] <pauliukas> I'm so used to doing tar -zxf and ps aux
[17:40:55] <pauliukas> Rather than... whatever else.
[17:41:13] * cmihai lodges a pipe in pauliukas's skull.
[17:41:25] <cmihai> This might come in handy...
[17:41:26] <pauliukas> ?
[17:41:42] <elektronkind> pauliukas: BSD ps lives in /usr/ucb
[17:41:46] <pauliukas> ah
[17:41:53] <elektronkind> GNU tar lives in /usr/sfw/bin
[17:42:05] <pauliukas> excellent. I'll add those to the path
[17:42:52] *** tombhad_ has quit IRC
[17:43:20] *** tombhad_ has joined #opensolaris
[17:45:29] <pauliukas> bash: /usr/sfw/bin/tar: No such file or directory
[17:45:36] *** tsoome has joined #opensolaris
[17:45:37] <flyingparchment> pauliukas: gtar
[17:45:52] <flyingparchment> GNU utils are called g* when they conflict, e.g. gtar, gmake, gpatch
[17:45:59] *** dugan has left #opensolaris
[17:46:03] <pauliukas> okay.
[17:46:11] <pauliukas> So how would compiling a software work.
[17:46:12] <flyingparchment> when /usr/gnu/ comes along it'll just be tar, and you can sort out the conflict with your path
[17:46:15] <pauliukas> I'd need to use gmake?
[17:46:23] <flyingparchment> pauliukas: if the software requires GNU make, yes
[17:46:29] <flyingparchment> solaris 'make' is in /usr/ccs/bin
[17:47:24] <pauliukas> How do I know what it requires.
[17:47:32] <pauliukas> If it's, say, Apache or PHP or whatever.
[17:47:39] <flyingparchment> the documentation might say.  or it might give errors when you use solaris make.
[17:47:46] <flyingparchment> apache and php both work fine with solaris make i think
[17:48:21] <pauliukas> Also...
[17:48:23] <cmihai> If it's GPL, it's usually gmake, if it's BSD/MIT, it's usually bmake, if it's something else, it's usually Solaris make :P
[17:48:31] <pauliukas> What kind of compiling/building tools should I install?
[17:48:42] <pauliukas> If I'll just want to compile stuff from time to time like apache and such.
[17:48:46] <flyingparchment> very little software requires bmake/pmake, because no-one ships it :)
[17:48:47] <cmihai> You should already have gcc and make and the likes.. but you probably want Sun Studio too.
[17:48:51] <flyingparchment> pauliukas: just use the bundled gcc and gmake
[17:48:57] <cmihai> flyingparchment, well, true. Unless you use BSD ports :-)
[17:49:12] <pauliukas> I meant... As in libraries and such.
[17:49:21] <flyingparchment> whatever the stuff you're building requires
[17:49:25] <cmihai> Nothing, just the depends listed in the software.
[17:49:26] <pauliukas> I use CentOS for Linux and they have packages which install most of the required libs and stuff.
[17:49:33] <cmihai> Read the documentation, it usually lists depends.
[17:49:37] <pauliukas> I don't want to go by hand and install hundreds of dependancies.
[17:49:42] <cmihai> No, they don't.
[17:50:01] <flyingparchment> btw, sun provides binaries for apache and php (and mysql), look for coolstack
[17:50:01] <cmihai> Well, you can use blastwave..
[17:50:44] <pauliukas> Could you explain what these are?
[17:50:48] <cmihai> Yeah, if you want Apache + PHP + MySQL, you should really use blastwave. Here, read this pauliukas http://unixsadm.blogspot.com/2007/08/installing-coolstack-on-solaris-and.html , it covers SMF integration too.
[17:51:01] <flyingparchment> bleh, don't use blastwave, use coolstack
[17:51:06] <cmihai> Read that article, it explains coolstack. For blastwave aka pkg-get, see www.blastwave.org
[17:51:16] <flyingparchment> blastwave is not compatible with sparse zones
[17:51:17] <cmihai> flyingparchment, Blastwave has other stuff.
[17:51:26] <cmihai> For Apache, you're better off with Coolstack.
[17:51:49] <flyingparchment> for mysql i use the mysql enterprise binaries, but i don't use apache or php, maybe if i did i'd look at coolstack instead
[17:52:09] <cmihai> lighttpd?
[17:52:16] <flyingparchment> sun java web server ;)
[17:52:19] <cmihai> Ugh :-)
[17:52:30] <pauliukas> Is coolstack Apache precompiled and stuff?
[17:52:33] <flyingparchment> that system only runs java apps so it's okay
[17:52:36] <pauliukas> I like installing my own stuff better.
[17:52:36] <flyingparchment> pauliukas: yes
[17:52:46] *** prg3 has joined #opensolaris
[17:52:47] <cmihai> pauliukas, you mean compiling for no good reason?
[17:52:48] *** Erwann has quit IRC
[17:52:53] <flyingparchment> you want to compile your own stuff but you don't want to compile it?  make up your mind :)
[17:52:54] <cmihai> There's sources too, but that's kind of silly.
[17:52:59] <pauliukas> On Linux, the precompiled shit is usually old.
[17:53:03] <cmihai> OMG noes
[17:53:08] <pauliukas> Like... Apache 2.2.2 when 2.2.6 already exists.
[17:53:10] <cmihai> This ain't gentoo mate.
[17:53:21] <pauliukas> I care about fixing security issues.
[17:53:22] <flyingparchment> what feature do you need in 2.2.6 that isn't in 2.2.2?
[17:53:27] <cmihai> Take what you have. For coolstack the source and the package is the same.
[17:53:32] <cmihai> And for Solaris there's patches.
[17:53:33] <pauliukas> That's the primary reason why I keep my stuff up to date.
[17:53:48] <cmihai> For security updates there's binary patches.
[17:54:02] <flyingparchment> i never saw a linux distribution that doesn't keep up with security patches for apache..
[17:54:05] <cmihai> Doing the whole recompile / reinstall / rebreak everything dance is just plain stupid.
[17:54:10] <pauliukas> So if you're patching... why not just update at the latest version.
[17:54:21] *** hile_ has quit IRC
[17:54:28] <flyingparchment> pauliukas: because the version you have works, and the new version is an unknown quantity
[17:54:37] <flyingparchment> when something works, the best thing you can do is not change anything
[17:54:44] <cmihai> Cuz the latest version isn't always stable. Do you really want this to be lie _some_ Linux systems where they change the config places every now and then?
[17:54:51] <flyingparchment> (unless you like being woken up at 3AM because the web server is broken ;)
[17:54:52] <cmihai> And every time you update you break something?
[17:55:21] <cmihai> Trust me, in production you may test even a patch for weeks before you apply it.
[17:55:28] <flyingparchment> i used to patch my solaris systems every week, now i never touch them unless i need a particular patch (e.g. for a security issue)
[17:55:29] *** RaD|Tz has joined #opensolaris
[17:55:30] <cmihai> And I don't mean a major version. A patch.
[17:55:36] <pauliukas> I test a new version days before I switch over in production.
[17:56:01] <pauliukas> No, I don't believe in distros that change a lot like FC.
[17:56:14] <pauliukas> But I believe in keeping the software which the machine primarily uses up to date.
[17:56:44] <cmihai> Most of the stuff in Solaris is pretty up to date mate.
[17:56:47] <flyingparchment> there are still lots of people using solaris 8, because it works for them
[17:56:54] <cmihai> I think people have just been spoiled by distros like Gentoo or Fedora...
[17:57:06] <cmihai> Shit, take Windows XP, it's like 7 years old...
[17:57:08] <flyingparchment> (although it's close to EOSL, so they might have to upgrade soon ;)
[17:57:11] <cmihai> Or Windows 2003.
[17:57:23] <pauliukas> I'm not complaning about Solaris. I just said why I didn't like packages.
[17:57:28] <tsoome> nah, in linux there are no patches - you need to upgrade
[17:57:40] <pauliukas> cmihai: They're releasing updates often. And also new kernels from time to time, if I recall.
[17:57:46] <tsoome> in solaris, you will get patch for broken sw
[17:57:49] <cmihai> Or HP-UX 11i, or Solaris 8 and so on. Regular production systems, the software is usually a few years old. All you really get is drivers and security updates.
[17:57:50] <flyingparchment> sun distributes 'patches' for linux, but they're just zip archives with rpms in ;)
[17:58:30] <cmihai> pauliukas, so does Sun. Take Solaris 10, it's up to update 4 right now.
[17:59:04] <cmihai> The Desktop stuff like Gnome does tend to be a bit old though.
[17:59:08] *** MattMan is now known as MattAFC
[17:59:28] *** blindfish has joined #opensolaris
[17:59:54] <Cyrille> part of the reason for not updating the Gnome from Solaris 10 is precisely the changes in some interfaces it would trigger.
[18:00:17] <CIA-26> daemon@elpaso: Added tag onnv_75 for changeset 4866b8311a88
[18:00:18] <CIA-26> arutz: 6572593 fmd made incorrect component degraded when DMA write UE happened.
[18:00:24] <cmihai> But that's no reason to compile your own stuff. Compile your own software, and you're a test bed of 1.
[18:00:27] <Cyrille> because from Gnome 2.6 (the one of Solaris 10) to Gnome 2.12, no 2.14, no 2.16, no 2.18, no 2.20, a lot of things have changed places.
[18:00:49] <flyingparchment> cmihai: we actually have to compile our own php on ubuntu, because they don't include a couple of modules we need :(
[18:00:59] <flyingparchment> (ours is based on their package though)
[18:01:38] <cmihai> There's enough bugs, issues and incompatibilities in the production stuff, last thing I need is to mix binaries with hand compiled bleeding edge software and make a wonderfull mess of things. Add blastwave, coolstack and pkgsrc to  the mix, and you won't be able to sort through the garbage.
[18:02:11] <cmihai> Especially when you have 8 gcc versions, 12 Apaches and tons of leftover configs from upgrades :-]
[18:02:40] <pauliukas> Okay, but tell me.
[18:02:50] <pauliukas> Who decides when something gets updated in blastware/coolwhatever?
[18:03:03] <cmihai> Who said blastwave was official?
[18:03:06] <pauliukas> When will they update Apache to a newer version?
[18:03:11] <pauliukas> iono
[18:03:15] <pauliukas> I just assumed.
[18:03:20] <cmihai> Blastwave is a community based 3rd party package system.
[18:03:21] <pauliukas> Sounds like a name that Sun would come up with.
[18:03:28] <flyingparchment> blastwave does 'stable' releases so often that never change
[18:03:39] <Cyrille> it has a distinct lack of "Java" and "System" in it.
[18:03:42] <flyingparchment> in the current stuff, it just gets upgraded when the maintainer feels like it (like debian unstable)
[18:04:22] <cmihai> CoolStack is Sun's work, but it's not part of Solaris either.
[18:04:59] *** MattAFC is now known as MattMan
[18:05:00] <pauliukas> Is it free or something paid?
[18:05:03] <flyingparchment> free
[18:05:24] <cmihai> pauliukas, http://unixsadm.blogspot.com/2007/08/installing-coolstack-on-solaris-and.html - just read and follow the links. It's a free optimized AMP stack.
[18:05:29] <pauliukas> But so, how do they chose the stuff?
[18:05:39] <Cyrille> carefully ;-)
[18:05:59] <cmihai> Read the README... it's pretty worked you know.
[18:06:02] <pauliukas> I like that. But what's the update cycle?
[18:06:07] <pauliukas> And how does the patching work?
[18:06:31] *** theRealballchalk has quit IRC
[18:06:50] *** theRealballchalk has joined #opensolaris
[18:07:49] <cmihai> Just like in Linux, you update to a new version. Since it's not part of Solaris, it's not supported.
[18:07:53] <cmihai> No support = no patches.
[18:08:04] <cmihai> You update to a new version when it comes out.
[18:10:47] <pauliukas> Why were people saying how great patching was just a few moments ago, then
[18:11:44] <cmihai> Patches are only for _supported_ products. That's Solaris and other Sun _products_. And you need a support contract for most of them anyway (only security updates without it).
[18:11:52] <flyingparchment> and drivers
[18:11:57] *** KermitTheFragger has quit IRC
[18:12:02] <cmihai> Yeah.
[18:12:18] <flyingparchment> and all the things they depend on.. which meant that 3/05 users got the zfs patch without a contract ;)
[18:13:08] <cmihai> Without one, all you can do is upgrade / liveupgrade between major updates (update to 08/07 for example) and apply security patches.
[18:14:16] <cmihai> And yes, it's the same for similar products. HP-UX - contract based, they send you encrypted cds/dvds and you can download stuff from their site. Oracle has Metalink, Microsoft has "geniune advantage", etc.
[18:15:41] <cmihai> Most of the open source software / OS don't have actual patches. You basically remove the old and install the new binary and try to keep the config...
[18:15:52] <pauliukas> okay...
[18:16:01] <flyingparchment> http://www.sun.com/servers/x64/x4150/ < i wish i could remove the CD-ROM and install 10 disks
[18:16:04] <pauliukas> So then, packages don't work for me because they're usually not kept up to date.
[18:16:31] * pauliukas hands flyingparchment some duct tape
[18:16:34] <cmihai> flyingparchment, heh, yeah, who needs cdrom when you have remote LOM cd :-)
[18:16:37] <pauliukas> LOTS of duct tape
[18:16:50] <the-decider> who needs CDROM when you can has pxe! :)
[18:17:23] <pauliukas> Who needs CD-ROM because it's a production server and you shouldn't be reinstalling the darn thing.
[18:17:44] <cmihai> Who needs CD-ROM when you can boot from USB drives :D
[18:17:55] <the-decider> but that means physically being near the server.
[18:18:21] <cmihai> Yeah, but since we already covered LOM and PXE...
[18:18:55] *** prg3 has quit IRC
[18:19:07] *** prg3 has joined #opensolaris
[18:19:13] <cmihai> Oy, I think you _can_ remove the cdrom
[18:19:32] <flyingparchment> cmihai: really?  in tech specs it says '8-channel SAS interface'.. i assumed that meant it only took 8 drives
[18:19:41] <the-decider> but can you remove the cdrom from the cdrom once you've removed the cdrom?
[18:20:14] <cmihai> flyingparchment, you can't add drives in it's place though :-). You can chose to buy one without the cdrom...
[18:20:38] <cmihai> And trust me, that saves you a good 200$
[18:20:45] <cmihai> DVD+/-RW Drive   [Add $ 195.00 each]
[18:20:53] <pauliukas> Geez.
[18:21:06] <pauliukas> My server vendor charges $30 for a DVD-R server drive.
[18:21:18] <pauliukas> Speaking of which... Why is Sun's hardware so expensive?
[18:21:31] <flyingparchment> it's not
[18:21:43] <pauliukas> $4K starting price for a Dual Xeon...
[18:21:43] <flyingparchment> the new amd64/xeon stuff is pretty competitively priced (comparable to dell/hp/etc)
[18:21:51] <flyingparchment> ignore list price, get a quote
[18:21:56] <pauliukas> ah...
[18:21:56] <flyingparchment> no-one pays list price for sun :)
[18:21:59] <pauliukas> Really? It's lower?
[18:22:06] <pauliukas> Even for small guys?
[18:22:49] <flyingparchment> we got a quote for 10 servers and something like 40% off list price
[18:22:54] <sickness> Even for small penis guys?
[18:23:02] <sickness> (sorry but couldn't help it >:)
[18:23:23] <flyingparchment> i guess if you just want 1 server it might be a bit more
[18:23:24] <pauliukas> Don't worry. We know that you're sexually frustrated.
[18:23:43] *** theRealballchal1 has joined #opensolaris
[18:24:33] *** nostoi has joined #opensolaris
[18:24:53] *** Cyrille has quit IRC
[18:26:26] *** myrkr has joined #opensolaris
[18:26:37] <myrkr> Hi all
[18:26:42] <pauliukas> hi myrkr
[18:27:02] *** Pedro1 has joined #opensolaris
[18:27:20] <myrkr> I was not able to compile an XPG6 source with CC (C++) -- I ran into #error -- is that intentional?
[18:27:44] <myrkr> that is, doing CC -D_XOPEN_SOURCE=600 doesn't work :/
[18:28:02] <flyingparchment> myrkr: yes.  at least, it's meant to happen; but i think that behaviour is stupid.
[18:28:07] *** vmlemon has joined #opensolaris
[18:28:11] <flyingparchment> myrkr: the best you can do for C++ is XPG4
[18:28:54] <myrkr> hmm, ok
[18:29:04] <myrkr> is there any hope of that changing?
[18:29:22] <flyingparchment> i don't know.  you might ask at http://forum.sun.com/, the studio guys read it
[18:29:34] <sickness> "all work and no play makes sickness a dull boy" ;P
[18:29:39] *** sfire||mouse has joined #opensolaris
[18:30:18] <flyingparchment> i wish posix would take c++ seriously.. i want nice raii classes for system interfaces ;)
[18:30:27] <myrkr> ;)
[18:30:32] *** Drone has joined #opensolaris
[18:31:15] *** theRealballchalk has quit IRC
[18:31:19] <myrkr> oh, erm, what is raii?
[18:31:53] <flyingparchment> http://www.efnetcpp.org/wiki/RAII ;)
[18:32:56] *** bnitz has left #opensolaris
[18:34:46] *** mikefut has quit IRC
[18:34:58] <theRealballchal1> so what's the word on WPA auth for intel 2200bg cards?
[18:35:51] <sickness> theRealballchal1: wpa_supplicant ?
[18:36:13] <theRealballchal1> sickness: yea i was just reading on freebsd6 and ubuntu with them
[18:36:31] *** estibi has joined #opensolaris
[18:36:42] *** sioraiocht has joined #opensolaris
[18:37:46] <theRealballchal1> sickness: found it thanks
[18:38:12] <sickness> yw
[18:39:19] <theRealballchal1> hm nah still no intel 2200
[18:41:32] <sickness> :|
[18:41:54] <theRealballchal1> i'll just order the ath
[18:42:07] <theRealballchal1> :)
[18:45:02] *** gekln has quit IRC
[18:50:16] *** theRealballchal1 has quit IRC
[18:50:28] *** timsf has quit IRC
[18:53:51] *** boyd has joined #opensolaris
[18:54:44] *** gdamore has left #opensolaris
[18:55:29] *** migi has quit IRC
[18:56:51] <LuckyLuke> I'm compiling boxbackup (Large files:         yes
[18:56:52] <LuckyLuke> Berkeley DB:         no
[18:56:52] <LuckyLuke> Readline:            no
[18:56:52] <LuckyLuke> Extended attributes: no
[18:57:01] <LuckyLuke> sorry. semi-wrong paste.
[18:57:42] <LuckyLuke> I'm compiling boxbackup on SXCE. At the end of ./configure I get the previous things. I think it doesn't find headers for readling and bdb. any help?
[18:58:56] <flyingparchment> readline and bdb aren't shipped with solaris.  did you installed them yourself?
[18:59:27] <cmihai> LuckyLuke, ./configure --help | grep readline
[18:59:44] <cmihai> You'll see something like --with-readline or whatever. You're supposed to give it the path to your libs or whatever.
[19:00:16] <CIA-26> lclee: 6275934 fdisk: EFI disk partition is allocated with wrong size, 6598208 Installer says there is not enough space to install in 50GB partition, 6608336 fdisk could have better error mechanism when user allocating more cylinders than are available, 6608510 fdisk has many cstyle errors
[19:00:27] <LuckyLuke> flyingparchment: no. thanks for telling, so I won't loose time looking for them :)
[19:00:47] <LuckyLuke> actually, I don't even now what difference makes compiling bb with bdb!
[19:00:52] <LuckyLuke> s/now/know/
[19:01:36] *** FelipeBare has quit IRC
[19:02:23] <salmandr> solaris noob question here - i'm running dansguardian (filtering web proxy) and it keeps complaining that it can't fork more processes
[19:02:29] *** FelipeBare has joined #opensolaris
[19:02:34] *** tehjackal has joined #opensolaris
[19:02:50] <myrkr> Qt doesn't seem to antialias text for me (snv 68, qt 4.3.1) is there something I can do to fix it?
[19:03:00] <cmihai> salmandr, ps -ef | wc -l
[19:03:02] <salmandr> i haven't set any limits with ulimit or rcap, how can i let it just do whatever it wants?
[19:03:03] <cmihai> What does that say?
[19:03:14] *** dmesg has joined #opensolaris
[19:03:17] <dmesg> hi dudes
[19:03:23] <dmesg> i have one question
[19:03:35] <salmandr> cmihai: 100
[19:03:45] <cmihai> That's not that big...
[19:03:47] <dmesg> is posible to install srss on www.gnusolaris.org?
[19:03:48] <cmihai> Check /etc/system
[19:03:54] <salmandr> ps -ef | grep dansguardian | wc -l gives 59
[19:03:55] <cmihai> You want to look at maxurpc
[19:04:16] <cmihai> erm
[19:04:17] <salmandr> however it's not being heavily used right now, it gets hit more in the afternoon when people are slacking and surfing the net
[19:04:24] <cmihai> maxuprc
[19:04:25] <salmandr> say between 4 and 5 pm :)
[19:05:06] <cmihai> Gah
[19:05:12] <salmandr> maxuprc isn't set in /etc/system - doesn't look like anything is actually
[19:05:17] <salmandr> this is a fresh install of solaris 10 U4 btw
[19:06:41] *** boyd_ has quit IRC
[19:06:56] <cmihai> Is there a maxusers?
[19:07:26] <salmandr> * set maxusers=40
[19:07:41] <flyingparchment> people still use maxusers?
[19:07:49] <cmihai> How much ram do you have?
[19:07:57] <cmihai> flyingparchment, no, it doesn't mean max users anymore
[19:07:58] <salmandr> but adb tells me that maxuprc is 16341 and maxusers is 1021
[19:08:15] <salmandr> Memory size: 2048 Megabytes
[19:08:17] <cmihai> It derives the max number of processes from that
[19:08:44] *** BadKarma has joined #OpenSolaris
[19:08:48] <salmandr> just fyi, this is the message that gets syslogged
[19:09:11] <salmandr> proxy dansguardian[2080]: [ID 702911 user.error] Unable to fork() any more.
[19:09:27] *** snuff-away has quit IRC
[19:09:29] <cmihai> Usually, when a process can't forked you're either forkbombed or you're out of processes.
[19:09:33] <cmihai> It hits some hard/softlimit.
[19:09:57] <cmihai> Is there an actual message in the system console?
[19:10:07] <salmandr> yeah, i'm just trying to find where it is :)
[19:10:17] <cmihai> Like "out of processes"? Or other processes that can't fork? Like ls or whatever
[19:10:40] <salmandr> hmm, i don't know about the console, i can check the ilom
[19:10:42] <salmandr> working remotely
[19:10:47] <salmandr> other processes work fine
[19:11:07] <cmihai> Heh, the math is close enough for you there.
[19:11:12] <salmandr> but it runs as the 'proxy' user
[19:11:26] *** bengtf has quit IRC
[19:11:27] <salmandr> so i wouldn't expect running apps as other users to bomb
[19:11:44] <cmihai> max_nprocs == 16 x maxusers + 10, that's 16341... that seems good. Since it's probably commented out in your /etc/system, it's automatic based on your ram.
[19:11:51] <cmihai> I'd try to define it by hand to 1024 in /etc/system.
[19:11:54] <salmandr> i tried running it as root just as a test, but it refuses to start that way, heh
[19:12:33] <cmihai> (maxusers). The others are based on that paramater. Just do a adb -k /dev/ksyms /dev/mem after you've rebooted, see what it gives you then...
[19:13:43] <salmandr> does the process limit count threads as well? or just 'real' processes
[19:14:26] <cmihai> Just processes...
[19:14:59] <cmihai> Here, try running this too:
[19:15:00] <cmihai> http://www.brendangregg.com/DTrace/execsnoop
[19:15:05] <cmihai> Just make sure you > log
[19:15:28] <cmihai> It records every process executed.
[19:15:36] <cmihai> Start it before you ruin dansguard
[19:15:46] <cmihai> and stop both when you have the issue...
[19:15:56] <cmihai> If you see a zillion processes... that's the issue :-)
[19:16:00] <salmandr> excellent
[19:16:17] <salmandr> this is a dedicated box so i sort of want dansguardian to take over
[19:16:38] <salmandr> system load never goes over 0.1 or so, there should be plenty of horsepower
[19:16:58] *** sahafeez has joined #opensolaris
[19:16:59] *** dlynes_laptop has quit IRC
[19:17:02] <salmandr> i'll try the dtrace command this afternoon when people start hitting it hard
[19:17:05] <salmandr> thanks for your help
[19:17:11] <cmihai> Well, you might want to read this too
[19:17:16] <salmandr> k
[19:17:24] <cmihai> Gasp... docs.sun.com is acting up again
[19:17:39] *** deather_ has joined #opensolaris
[19:17:45] <cmihai> http://docs.sun.com/app/docs/doc/819-2724?l=en&q=solaris+tuning - This
[19:18:09] <cmihai> Solaris Tunable Parameters Reference Manual, take a good look over the kernel section (2)
[19:19:18] *** kspath has joined #opensolaris
[19:19:38] <salmandr> i do have a ram upgrade to 4GB which hasn't shipped yet
[19:19:40] <salmandr> it says "After computation of dependent parameters is done, maxusers is never referenced again."
[19:19:49] <cmihai> Yep
[19:19:58] <cmihai> As I've said, it's just used to calculate the other stuff.
[19:20:00] <salmandr> does that mean i should adjust /etc/system when i install that?
[19:20:11] <cmihai> Nah, when you reboot
[19:20:12] <salmandr> or does that mean it calculates it once on boot
[19:20:19] <salmandr> ok, reading comprehension is low :)
[19:20:29] <cmihai> You can change them on the fly too
[19:20:45] *** Murmuria has joined #opensolaris
[19:21:07] *** tinman2k has quit IRC
[19:22:13] *** blindfish is now known as MrChick_
[19:23:16] *** MrChick_ is now known as blindfish
[19:27:47] *** jcea has joined #opensolaris
[19:28:11] *** jcea has left #opensolaris
[19:28:18] *** jcea has joined #opensolaris
[19:31:06] *** MattMan is now known as MattAFC
[19:34:39] *** deather has quit IRC
[19:35:51] *** calumb has quit IRC
[19:36:00] *** bengtf has joined #opensolaris
[19:36:11] *** charlesnw has joined #opensolaris
[19:39:42] *** cydork has quit IRC
[19:40:18] <myrkr> does DRI work on solaris?
[19:42:14] *** WifiJane has joined #opensolaris
[19:43:04] *** alanc-away is now known as alanc
[19:44:06] <WickedWicky> howedy
[19:46:04] <cmihai> myrkr yeah
[19:46:26] <cmihai> DRI needs kernel support (DRM subsystem). That's been ported to Nevada for the Intel drivers...
[19:46:48] <cmihai> myrkr see http://www.opensolaris.org/os/community/on/flag-days/pages/2006102901/
[19:47:03] <cmihai> But if you just want 3D, get a NVIDIA card, Solaris comes with NVIDIA drivers
[19:47:24] <cmihai> (nevada and 10u4, for the rest you need to download and install them yourself from nvidia.com)
[19:48:38] <myrkr> I have snv 68, and I believe I'm using nvidia driver, but glx doesn't seem to be loaded -- that's why I started asking myself
[19:48:41] <cmihai> NVIDIA stuff isn't DRI, but it's basically the same thing. Works quite well too: http://unixsadm.blogspot.com/2007/08/gaming-on-solaris.html (quake 1,2,3)
[19:49:02] <cmihai> myrkr check nvidia-settings tool
[19:49:54] *** blindfish has quit IRC
[19:50:07] <myrkr> hmm, ERROR: NV-CONTROL extension not found on this Display
[19:50:23] <myrkr> how do I load this NV-CONTROL?
[19:50:26] <cmihai> Huh
[19:50:30] <cmihai> It should be loaded.
[19:50:40] <cmihai> You sure you have a NVIDIA card and are running XOrg?
[19:50:48] <cmihai> You could be running XSun for all I know :-)
[19:52:10] <myrkr> cmihai: yes, I believe I'm running Xorg, but I haven't asked it, and I also believe I have nvidia
[19:52:45] *** tehjackal has quit IRC
[19:52:52] <myrkr> at least scanpci tells me I have nvidia card
[19:53:21] <cmihai> Can't you like.. mm.. check?
[19:53:29] <myrkr> GeForce Go 7600
[19:53:35] <cmihai> kdmconfig or ps should give you an idea on what you're running.
[19:53:41] <cmihai> Oh, GeForce Go isn't actually NVIDIA
[19:53:53] <myrkr> oh
[19:53:56] <cmihai> Just the chipset, but it's modded by vendors
[19:54:05] <cmihai> So.. Toshiba, Acer, Dell, etc ship their own drivers
[19:54:13] <cmihai> Yeah, sucks :-)
[19:54:20] <cmihai> Go blame $VENDOR
[19:54:21] <myrkr> yeah :/
[19:54:27] <cmihai> Even on Windows say.. Vista
[19:54:34] <cmihai> You need their driver, nvidia.com ones won't work
[19:55:04] <cmihai> So, what is it? (vendor)
[19:56:29] <WickedWicky> Erm.. is that 100% accurate?
[19:56:47] <cmihai> What now
[19:56:48] <myrkr> acer
[19:56:57] <cmihai> Yeah, for acer it is :-)
[19:57:01] <WickedWicky> My HP laptop has a Geforce Go 5600
[19:57:04] <WickedWicky> and i use the nvidia drivers
[19:57:05] <cmihai> I was asking... since I have an Acer too :-).
[19:57:15] <cmihai> (8600 GT)
[19:57:47] <palowoda> Nice chipset cmihai.
[19:57:48] <cmihai> WickedWicky, not all do it, and some work with modded drivers, but most will modify it enough so it won't work without. I know Dell and Acer do it on most of their models.
[19:58:27] <myrkr> cmihai: so, what driver am I using?
[19:58:42] <cmihai> palowoda, yeah, I got one of those Santa Rosa platform laptops. It's pretty neat... Core 2 Duo 4MB cache, 4GB DDR2-800, 250GB Disk, and all that.
[19:58:56] <cmihai> myrkr check your config or X logs. Probably VESA or nv or something
[19:59:07] * WickedWicky wants cmihai's setup
[19:59:14] <cmihai> palowoda, too bad I've seen a similar Dell but with 2 250GB disks only days later
[19:59:25] *** derchris^ has quit IRC
[19:59:32] <myrkr> nv according to xorg.conf
[19:59:36] *** dmesg has left #opensolaris
[19:59:41] *** derchris^ has joined #opensolaris
[19:59:41] <cmihai> Yeah, no 3D for you :-)
[19:59:49] <cmihai> That's the open sores driver
[20:00:01] <myrkr> ah, ok
[20:00:13] <cmihai> WickedWicky, http://global.acer.com/PRODUCTS/notebook/as5920g.htm, I just changed the disks and RAM. But you'll probably want to check the Dell too
[20:00:18] <myrkr> is that why I don't have glx?
[20:00:26] <WickedWicky> I think you can tell by the nvidia logo popping up when X starts, right?
[20:00:29] <cmihai> They sell a similar Santa Rosa model, but with 2 SATA disks... nice for RAID and all that, since laptop disks SUCK
[20:00:29] <CIA-26> tm144005: 6551509 modunload while the port(nxge) is running, causes system panic
[20:00:30] <CIA-26> sb155480: 6531266 link-aggregation with Nemo on e1000g primary ldom does not work, 6581720 IO-DOMAIN-RESET (Huron/Ontario): Guest domain may lose connection to vsw if primary is rebooted, 6604983 multicast processing after a channel-reset is broken in vnet, 6610044 vsw should mac_register() in attach()
[20:00:55] <cmihai> WickedWicky, you can disable the logo, so it's not a good idea to base your assumptions on that. Just check the config / logs
[20:01:08] <palowoda> You really have to get the 7200rpm 2.5 sata disks in a laptop.
[20:01:21] <palowoda> Anything else isn't worth it.
[20:01:23] <myrkr> hmm, I think I have seen that logo, I think it disappeared when I made the xorg.conf file :/
[20:01:35] <cmihai> palowoda, I friend of mine got that
[20:01:41] <cmihai> It came with 5200rpm disks though
[20:01:42] <alanc> how did you make the xorg.conf?
[20:01:58] <palowoda> I've replaced all the slow sata disks in my laptops.
[20:02:00] <myrkr> I think I used some xconf utility to do it
[20:02:00] <cmihai> myrkr oh, lol, XOrg -configure mate
[20:02:07] *** stevel has joined #opensolaris
[20:02:12] <cmihai> myrkr then make sure you have Driver nvidia
[20:02:14] <alanc> also, the nvidia driver bundled in Solaris is old, doesn't support the latest nvidia chipsets, download newer from nvidia.com if you have never
[20:02:26] <alanc> err, have *newer* chipsets
[20:02:56] <palowoda> alanc: doesn't the nv driver get updated around build 75?
[20:03:00] <myrkr> cmihai: does that dump my current config?
[20:03:02] *** pauliukas has quit IRC
[20:03:09] <cmihai> nah, it saves a auto-generated one in root
[20:03:20] <cmihai> edit that, make sure you have your proper Modes lines and Driver nvidia
[20:03:23] *** tinman2k has joined #opensolaris
[20:03:26] <alanc> palowoda: nv does, nvidia update is still in process, since it EOF's older chipsets
[20:03:28] <myrkr> alanc: ok, will do
[20:04:22] <alanc> the /usr/bin/nvidia-xconfig utility should also generate a correct config file for the nvidia driver
[20:04:24] <palowoda> Yeah I know about the older nv chipsets.  Anything below the 66xx series isn't even worth it.
[20:04:27] <cmihai> palowoda, I've actually seen a dell notebook with dual disks and dual NVIDIA cards in SLI :-)
[20:04:39] *** ludovicp has quit IRC
[20:04:40] <palowoda> also nvidia-settings
[20:04:49] <cmihai> Comes with dual 7200 rpm disks too
[20:04:56] <alanc> palowoda: unfortunately, Sun sold a handful of those in the first generation W1100z/W2100z
[20:04:59] <palowoda> chmihai: Yeah Sager has some interesting models.
[20:05:02] <cmihai> It's also 5000$, but meh :P
[20:05:09] <RElling> cmihai: does it weigh less than 20#?
[20:05:28] <RElling> might as well carry a desktop under your arm :-)
[20:05:29] <palowoda> alanc: Yeah it's cheap to replace them with a decent card.
[20:05:41] *** kFuQ has joined #opensolaris
[20:06:24] <alanc> yeah, which is why we'll probably just stick in a release note and do it, but have to get approval from various people first
[20:06:26] <palowoda> RElling: When you want a poratble desktop what do you expect.  Those LCD and disks start to wieght down.
[20:06:38] *** m0le has quit IRC
[20:06:58] <palowoda> Hey 20inch LCD on a laptop is serious wieght.
[20:08:12] <cmihai> Heh, yeah, and the case is too
[20:08:19] <cmihai> I bought a light titanium / aluminium case for my lappy
[20:08:25] <cmihai> it's 3 kilos empty ;-)
[20:08:28] <palowoda> 20inch LCD on a laptop is amazing to look at too.
[20:09:05] <madhatter> Hi. I used Live Upgrade at the weekend. Now I am stuck while moving an ufs disk to zfs, because it's used in the old build.
[20:09:12] <palowoda> I have two Haliburton aluminium cases.  They are cool.
[20:09:18] <cmihai> You want something you can move around, get one of those 11'' LCD 1.1 kilo sony's or lenevos...
[20:09:39] <palowoda> Oh God, Sony's are crap.
[20:09:52] <cmihai> palowoda, yeah, my case is very much like that
[20:10:18] <madhatter> And I can't delete the old build it seems...
[20:10:20] <RElling> same weight as a Kaypro?
[20:10:25] <cmihai> I really like it... too bad it's so damn expensive..and heavy :-)
[20:10:39] <myrkr> now, for my original dive into X settings and stuff: my Qt 4.3.1 doesn't seem to antialias text -- would anyone here have a clue what I might try?
[20:10:50] <palowoda> Hey I worked for Osborne at one time don't talk about Kaypro.
[20:11:03] <cmihai> RElling, yeah :-)
[20:14:37] *** timsf has joined #opensolaris
[20:15:40] *** Triskelios has quit IRC
[20:16:40] *** timsf has quit IRC
[20:17:14] *** timsf has joined #opensolaris
[20:17:28] <madhatter> Hm. Nobody an idea how I can remove the old BE? It says 'You will need to create a new Live Upgrade boot environment on the boot disk to find a new candidate for the GRUB menu.' But when I do that I have the partition in there again... Looks like a deadlock.
[20:18:29] *** timsf has quit IRC
[20:20:55] *** timsf has joined #opensolaris
[20:30:33] *** slackjr has joined #opensolaris
[20:30:55] <paul> madhatter: ludelete? or can you not boot?
[20:31:02] * paul hasnt scrolled back
[20:31:29] * paul wonders if there's any way around circular patch dependencies
[20:31:48] <madhatter> paul: I am trying to get a disk out of my configuration and make it a zfs pool. But I can't because the disk is automatically mounted in another BE
[20:31:55] <the-decider> paul: a moebeus patch installer?
[20:32:42] <paul> (125548-02 wants 122661-08, which is obsoleted by 120012-14, which depends on 125548-01)
[20:32:53] *** tsoome1 has joined #opensolaris
[20:33:28] <paul> madhatter: mounted as?? (and what fs?)
[20:33:47] <the-decider> Toss them all in a directory and run patchadd -M until fully baked?
[20:33:51] <madhatter> home, ufs
[20:34:06] <palowoda> http://www.news.com/Schwartz-Sun-to-focus-on-storage-business/2100-1015_3-6211162.html?tag=nefd.top  Jonathan is planing to integrate AVS and Terabyte storage systems into project Indiana.
[20:34:28] <paul> presumably UFS, so use lumount to get at the other BE and edit its /etc/vfstab
[20:34:38] <elektronkind> palowoda: they're already in Nevada
[20:34:47] <elektronkind> you can use AVS to replicate today
[20:34:50] <palowoda> But I said Indiana.
[20:35:11] <elektronkind> Indiana... of course it'll be in Indiana. It's already in Nevada
[20:35:33] <the-decider> you could easily fit all of indiana in nevada.  nevada's a very large statae.
[20:35:36] <palowoda> Damit we need terabyte tape in Indiana.
[20:35:55] <elektronkind> terabyte tape?
[20:35:58] <the-decider> oh... forgot, this not #geography
[20:36:02] *** RaD|Tz has quit IRC
[20:36:39] <elektronkind> TB storage?
[20:36:46] <elektronkind> someone's smoking something
[20:36:47] <elektronkind> [daleg@ds2]~$ zpool list
[20:36:48] <elektronkind> NAME                    SIZE    USED   AVAIL    CAP  HEALTH     ALTROOT
[20:36:48] <elektronkind> ds2-store              12.7T   3.72T   8.93T    29%  ONLINE     -
[20:37:00] <elektronkind> 12.7T right there.
[20:37:03] <the-decider> elektronkind says "mine is bigger than yours"
[20:37:13] <paul> the-decider: seems an idea except 'Patch 120012-14 must be installed by itself'
[20:37:29] <the-decider> paul: gah!
[20:37:50] <paul> i seem to be hitting the lucreate bugs though..
[20:40:10] *** jonkri has quit IRC
[20:43:00] <HCoyote> Anyone know if the checksumming on zpools is still single threaded in nv_70?  If so, I think I might have found my bottleneck for my NFS write performance.
[20:43:50] *** tsoome has quit IRC
[20:44:10] <palowoda> Hmm what types of compression are multithreaded?
[20:44:33] <HCoyote> (actually, it's pure conjecture that it's the checksumming at this point ... I just know that with 1 zpool, I get only 60MB/s for streaming NFS, but splitting that same disk set into two zpools, I get an aggregate higher throughput of about 80MB/s)
[20:45:23] <palowoda> Oh checksum sorry.
[20:45:38] <madhatter> paul: I mount the old BE's / and edit the vfstab?
[20:45:56] <paul> madhatter: yes, use lumount
[20:46:20] <paul> (if i understand your problem correctly)
[20:48:03] <palowoda> HCoyote: So what do you think the results should be?
[20:48:10] *** snuff-away has joined #opensolaris
[20:48:19] <HCoyote> palowoda:  ~130MB/s
[20:48:22] <madhatter> paul: We'll see ;)
[20:48:31] <HCoyote> based on my iozone tests run directly on my x4500.
[20:49:22] *** tsoome has joined #opensolaris
[20:49:42] <palowoda> You expect what percentage of nfs throughput to be that of local disk performance?
[20:50:07] *** stevel has quit IRC
[20:50:56] <HCoyote> I expect that NFS write performance should be within 10% or so of direct write performance on the system.
[20:51:31] <palowoda> On a 10 megabit network?
[20:51:58] <HCoyote> Gigabit.
[20:52:06] <HCoyote> (those are MBytes, not Mbits)
[20:52:25] <palowoda> What is your total Gigabit thorughput?
[20:52:49] *** stevel has joined #opensolaris
[20:53:53] <HCoyote> I've tried with a single e1000g, all the way up to having all four in a single aggregate (which has been tested to provide about 300MBytes/s of read throughput)
[20:54:34] <HCoyote> So, I'm able to get just about wirespeed when handling reads.
[20:55:07] <HCoyote> Write performance ... I would expect to get near what the backend has shown to be physically capable of  ... 130MB/s.
[20:55:18] <HCoyote> and I"d be happy if I got within 10% of that for NFS writes.
[20:55:25] <palowoda> Log a bug see what happens.
[20:55:37] <flyingparchment> do all solaris services (e.g. nfs, kerberos, ldap) support ipv6?  is there a list somewhere?
[21:00:35] <CIA-26> edp: 6610046 memory leak for branded processes in elfexec()
[21:03:55] *** peemus has joined #opensolaris
[21:04:57] *** cmihai has quit IRC
[21:05:17] *** stevel has quit IRC
[21:05:51] <madhatter> paul: Yes, that was exactly what I was looking for :)
[21:06:15] *** tsoome1 has quit IRC
[21:06:32] <paul> madhatter: great :)
[21:09:16] *** gdamore has joined #opensolaris
[21:09:19] <gdamore> hi *
[21:10:50] <trygvis> o/
[21:11:42] *** redbeard0x0a has quit IRC
[21:16:10] *** Kush- has joined #opensolaris
[21:17:59] <RElling> HCoyote: NFS write performance depends quite significantly on the type of writes: sync writes are usually the bottlenecks
[21:18:14] <RElling> metadata updates are sync
[21:19:11] <palowoda> Maybe filebench is a better benchmark for nfs over zfs.
[21:20:00] <RElling> ZFS is especially sensitive to sync writes under NFS, you'll definitely want a slog device!
[21:20:33] <palowoda> What ever happen to the ZIL issue with nfs on zfs?
[21:20:52] <elektronkind> that's what slog is for
[21:21:08] <elektronkind> to put the ZIL on a faster (local) device
[21:21:39] <the-decider> how does that effect data consistency; e.g. if the local device goes tits-up?
[21:22:21] <palowoda> But I don't think that would affect a iozone test with nfs on zfs.  But than again is iozone a standard test used for comparison with nfs filesystems.
[21:22:58] <HCoyote> I'd be curious to see what performance others were getting out of an x4500 in various configurations for NFS usage.
[21:23:04] <HCoyote> sadly, I haven't been able to find any. :-(
[21:23:20] <elektronkind> it doesn't. ZIL not being there will not affect consistency, but I dunno what happens if it's in use and then suddenly goes away
[21:23:23] <palowoda> And when one expect something to be within 10 percent of some expected result why?
[21:23:30] <elektronkind> this blog has details
[21:23:31] <elektronkind> [daleg@ds2]~$ zpool list
[21:23:31] <elektronkind> NAME                    SIZE    USED   AVAIL    CAP  HEALTH     ALTROOT
[21:23:31] <elektronkind> ds2-store              12.7T   3.72T   8.93T    29%  ONLINE     -
[21:23:33] <elektronkind> oops
[21:23:36] <elektronkind> http://blogs.sun.com/perrin/entry/slog_blog_or_blogging_on
[21:23:37] *** mikefut has joined #opensolaris
[21:23:41] <elektronkind> stale paste buffer ;)
[21:23:42] <the-decider> elektronkind: put it back in your pants
[21:24:12] *** Murmuria has quit IRC
[21:24:14] <elektronkind> sorry. my 12TB array has a mind of its own.
[21:24:41] *** dlg has quit IRC
[21:24:47] <palowoda> elektronkind: No problem, how much did you pay for that thing anyways?
[21:25:40] *** sioraiocht has quit IRC
[21:25:49] <elektronkind> it's 4 mirrored apple xserve raids. Two with 10TB of disk and the other two with 6TB
[21:26:18] <elektronkind> I guess total cost would be around $25k new
[21:26:35] <elektronkind> the 10TB flavor was 10k each
[21:26:42] <palowoda> Shesh I thought Apple xserves where cheaper than that.
[21:27:16] <elektronkind> they're pretty cheap for the space you get. if I didn't mirror them 1-to-1, that would be 26TB after all
[21:27:22] <salmandr> 10TB for 10k isn't bad :)
[21:27:26] <elektronkind> $1k / TB
[21:27:41] <the-decider> and they look rather nice next to the Sun Galaxy / Coolthreads gear
[21:27:44] <the-decider> all that nice silvernes
[21:28:18] <salmandr> anyone running a 2530 (SAS) in JBOD for ZFS?
[21:28:19] <elektronkind> and the wavelength that their green LEDs emit also match!
[21:28:40] <the-decider> sun needs to use some of the blue ones.
[21:28:49] *** bengtf_ has joined #opensolaris
[21:28:49] <the-decider> can they offer that as a BTO option? ;)
[21:29:12] <elektronkind> blue ones are there. they light up on ok-to-remove components
[21:30:04] <HCoyote> elektronkind:  yeah, I saw that slog blog entry.  It's hard to tell what he was actually testing in that (besides direct performance of slogs on and off disk).  And since there's no testing methodology listed, it's hard to replicate what was going on to compare.
[21:30:07] *** dlg has joined #opensolaris
[21:31:20] *** sfire||mouse has quit IRC
[21:32:07] <elektronkind> HCoyote: a simple filebench run should show results, and in real-world use... if your ZFS pool is currently getting bogged down with an app that fsync()'s a lot, moving the ZIL off the pool to a dedicated device (fast local dedicated SAS drive, for example) you'll see an improvement
[21:32:34] <elektronkind> esepcially since if the pool lives on a RAID array, the array's cache would be better utilized
[21:32:40] <HCoyote> elektronkind:  well, it's NFS ... it's going to fsync() anyway.
[21:32:53] <HCoyote> and it's an x4500. It's not like it has any room in it for anything else :-)
[21:33:37] <elektronkind> HCoyote: that's what slog is aimed at, NFS
[21:34:02] <the-decider> wouldn't be bad for mailstores as well; for some reason, anything mail related loves to fsync()
[21:34:12] <HCoyote> hmm.
[21:34:31] <elektronkind> I mean, there's even 3.5" and 2.5" solid state disk in the near future
[21:34:51] <elektronkind> use that for ZIL. expensive, but hella fast if your app depends on performance
[21:34:52] <HCoyote> I'll be happy whenthe next gen thumpers come out.
[21:35:09] <elektronkind> yeah, they're overdue for a PCIe update
[21:35:46] *** kFuQ has quit IRC
[21:35:52] *** vmlemon_ has joined #opensolaris
[21:36:37] *** kFuQ has joined #opensolaris
[21:36:46] <HCoyote> andy b. came out a few weeks ago and was telling us about some of the things in the pipe.  Definitely lots of good options.
[21:37:36] <palowoda> Doesn't the X4450 have more pci-e slots?
[21:37:46] <HCoyote> We're not allowed to use those.
[21:37:52] <elektronkind> in its current shipping incarnation, it's PCI-X
[21:38:28] <elektronkind> oh you said 4450
[21:38:31] <elektronkind> that's PCIe
[21:38:36] <palowoda> 6 pci-e slots.
[21:38:50] <RElling> Just use some spare space on your X4500 boot disk for an slog... it will work far better than no slog
[21:39:10] *** vmlemon has quit IRC
[21:39:12] <elektronkind> iirc the 4500 is still the only Gallaxy box which is non-PCIe
[21:39:21] <HCoyote> RElling:  really?  hmm.
[21:39:26] * HCoyote goes off to try that.
[21:40:14] <HCoyote> Though ... slogs only come in OpenSolaris, right?
[21:40:39] <palowoda> That is what this channel is using. :)
[21:40:53] <elektronkind> HCoyote: yeh, it's ZFS version 5 that has it
[21:41:01] <elektronkind> s10u4 ships zfs v4
[21:41:20] <elektronkind> so no slog in s10 for now. I imagine it'll be added with s10u5
[21:41:27] <HCoyote> Ok.
[21:41:28] <palowoda> Poor S10.
[21:42:13] *** Fish has quit IRC
[21:42:19] *** sparvu has joined #opensolaris
[21:43:25] <HCoyote> This is kinda sad, really.  We've had this x4500 since they came out and it sat on a shelf for six months doing nothing.
[21:44:00] <elektronkind> what were/are you waiting for?
[21:44:05] <palowoda> Damn a nice machine to just let it sit.
[21:44:30] <the-decider> ship it out here, we'll use it.
[21:44:32] <HCoyote> well, it got thrown in as part of a large deal we made.
[21:44:33] <elektronkind> if you were waiting on a feature, a x4500 would make a nice NTP server
[21:44:45] <elektronkind> ;)
[21:44:50] <HCoyote> and no one's had the time to really do anything with it.
[21:45:04] <elektronkind> HCoyote: learn iscsi
[21:45:08] <the-decider> usenet server!
[21:45:08] <palowoda> Think of the P0rn.
[21:45:15] <elektronkind> load s10u4 on it, learn iscsi
[21:45:19] *** cypromis has joined #opensolaris
[21:45:20] <elektronkind> there's your mission now ;)
[21:45:29] <the-decider> you'd at least be able to keep a *couple* days of a full binaries feed
[21:45:37] <sommerfeld> port mythtv to solaris and use it as a dvr!
[21:45:46] <HCoyote> I've only had my hands on it for a month or so, trying to figure out how to make it perform well enough to replace certain file storage activities in our environment.
[21:45:51] <the-decider> mythtv+zfs would be nice
[21:46:07] <palowoda> Make an embedded toliet controller to save water.
[21:46:28] <elektronkind> sommerfeld: heeey. you're on to something there. it's a 19" format chassis so should fit nicely inside a quarter-height home entertainment system enclosure
[21:46:55] <HCoyote> In the grand scheme of things, it's not even 5% of our available disk space on other storage systems.
[21:46:56] <elektronkind> just remove all its fans and it'll be noiseless
[21:47:01] <sommerfeld> hah
[21:47:02] <the-decider> however, you'd want to put it in an isolated equipment rack ;)
[21:47:33] <elektronkind> hmm
[21:51:04] *** vmlemon_ is now known as vmlemon
[21:51:39] <trochej> Anyone here installed Ubuntu and SXCE on one disk?
[21:52:36] <HCoyote> RElling:  how can I tell how well the slog is performing?
[21:56:00] <RElling> what do your performance tests show?  In general, slog is a write-only workload.
[21:56:24] <RElling> putting it on a different disk gets it out of the queues for the rest of the disks in the zpool
[21:56:32] <elektronkind> trochej: yes. with WinXP too even
[21:56:58] <HCoyote> RElling:  zpool iostat is showing worse performance with a dedicated log mirror.
[21:57:35] <RElling> what does iostat show? (IMHO zpool iostat isn't very useful)
[21:58:05] <sommerfeld> you want to measure the performance from the workload's standpoint
[21:58:16] <elektronkind> trochej: I installed solaris first using fisk to partition out the disk... partition 1 for windows, 2 for Solaris and 3 for Linux. I then installed windows on the first parition, then ubuntu on the 3rd, and lastly Solaris on the 2nd parition. The solaris installer picked up on the Windows and Linux installations and made grub entires for itself and them
[21:58:40] <RElling> sommerfeld is right, from the server's perspective being busy is a good thing
[21:58:49] *** sickness_ has joined #opensolaris
[21:59:18] *** sioraiocht has joined #opensolaris
[22:00:31] <HCoyote> hmm.
[22:02:30] <HCoyote> ah, iozone run just finished.  So, slog-enabled zpool full of mirrors is only able to write at about 30MB/s, where previously, I was getting about 45MB/s.
[22:03:30] <kjetilho> HCoyote: sorry, I missed what you're using for your slog?
[22:04:48] <kjetilho> and what benchmark software are you running?
[22:05:14] <HCoyote> two sata drives on different controllers on my x4500.  benchmarked with iozone working in distributed mode.
[22:05:37] <HCoyote> and right now, I'm only running the initial write/re-write tests.
[22:06:25] *** slackjr has quit IRC
[22:07:54] <kjetilho> HCoyote: just one NFS client?  NFSv4?
[22:08:34] <HCoyote> 30 clients, nfsv3.
[22:08:47] <HCoyote> I can go up to 89 clients right now if necessary.
[22:09:15] <kjetilho> ok.  sounds like a good setup!
[22:09:23] <kjetilho> too bad about the performance :)
[22:09:26] <HCoyote> yep.
[22:10:41] * HCoyote wanders off to take a break.
[22:10:58] *** deather_ is now known as deather
[22:11:45] <kjetilho> HCoyote: are you running the slog disks mirrored or separate?
[22:12:19] <kjetilho> I don't see the motivation for mirroring the ZIL actually, perhaps someone can enlighten me?
[22:12:54] *** yippi_ has joined #opensolaris
[22:12:59] *** tombhad_ has quit IRC
[22:13:02] <HCoyote> mirrored
[22:13:36] <RElling> kjetilho: me either, it doesn't buy you anything on the server side
[22:14:18] <kjetilho> it seems to me mirroring easily introduces more latency, since the disks won't be spinning in sync
[22:15:17] <RElling> it won't be that much more latency, only a few ms (historically ~ 25%)
[22:15:33] <kjetilho> "a few ms" is not much? :)
[22:16:23] <holcomb> nfs is really sensitive to latency, too.  get an nvram device
[22:17:47] <kjetilho> HCoyote: I'd be interested to hear if the two slog disks in round-robin performs better.
[22:17:51] <RElling> for two idle disks, with slogs, it should be ~ 1/2 rev, worst case.  For 7,200 rpm that is 4.16ms
[22:18:43] <kjetilho> not worst case -- on average
[22:19:16] <RElling> average worst case :-)
[22:19:55] *** sickness1 has joined #opensolaris
[22:22:33] <RElling> I'd expect a normal distribution around 2.08ms
[22:22:52] *** sickness has quit IRC
[22:24:38] *** tinman2k has quit IRC
[22:24:51] *** stevel has joined #opensolaris
[22:26:41] *** tinman2k has joined #opensolaris
[22:26:58] *** yarihm has joined #OpenSolaris
[22:29:37] <kjetilho> hmm, "The zil behaves differently for different size of writes that happens. For small writes, the data is stored as a part of the log record. For writes greater than zfs_immediate_write_sz (64KB), the ZIL does not store a copy of the write, but rather syncs the write to disk and only a pointer to the sync-ed data is stored in the log record."
[22:30:14] *** sickness_ has quit IRC
[22:30:17] <kjetilho> I wonder how large writes iozone will do over NFS
[22:30:52] *** sickness1 is now known as sickness
[22:31:33] <sickness> uhm
[22:32:42] *** wnorrix has joined #opensolaris
[22:34:03] *** yippi_ has quit IRC
[22:34:19] <HCoyote> kjetilho:  you can define the recordsize as part of the iozone parameters.
[22:35:19] <RElling> for NFS, isn't 32k the default?  or is there a different definition of "block size"?
[22:35:51] <HCoyote> RElling:  32k is what the OS will write
[22:36:56] <RElling> you can check using iostat on the clients and compare to the server
[22:37:21] <holcomb> nfs4 is 1M by default, and nfs3 will start out at 32K but go up if it can
[22:37:32] *** JBeck has joined #opensolaris
[22:37:32] *** ChanServ sets mode: +o JBeck
[22:39:59] * HCoyote will need to test nfs4 at some point also.
[22:41:05] <RElling> holcomb: yep, there it is, right in the fine man page... I guess I hadn't looked at it in a few years ;-)
[22:41:40] *** Gman has joined #opensolaris
[22:42:01] *** cypromis has quit IRC
[22:43:45] <HCoyote> So ... why would I get a higher aggregate performance using two zpools on one system versus having all disks be in one zpool?
[22:44:16] *** estibi_ has joined #opensolaris
[22:45:34] *** jonkri has joined #opensolaris
[22:46:29] <e^ipi> I would assume the opposite actually
[22:46:44] <e^ipi> ZFS will lay out the files in an efficient manner if it's allowed to IIRC
[22:47:01] <HCoyote> that's what I expected too.
[22:47:36] <RElling> HCoyote: pedantically, there is a different resource as the bottleneck.  kinda hard to look at all of the queues involved, especially since some of them are in the disks
[22:47:39] *** ChanServ sets mode: +o stevel
[22:49:43] <RElling> by default, ZFS will queue 35 iops to each vdev, which is one reason why using slog will improve interactive performance
[22:55:26] <HCoyote> I saw the zfs_vdev_max_pending tunable, but wasn't quite sure what would be appropriate to set to.  Roch's blog about this last year doesn't give me warm fuzzies about touching this param for NFS.
[22:57:29] <RElling> I would only look at zfs_vdev_max_pending as a last resort, especially if NCQ is working (which it should be for X4500 -- iostat -x will show the actv entries > 1
[22:58:34] <RElling> BTW, when I say "iostat" I mean a useful output option, I usually use something like "iostat -xnz 1"
[22:58:56] <HCoyote> cool, thanks.
[22:59:05] <HCoyote> I was using -xC
[22:59:15] <RElling> "-T d" is also useful
[22:59:34] <e^ipi> you can put the ZIL on external devices?
[22:59:46] <RElling> On X4500, "-n" is almost mandatory :-)
[22:59:55] <RElling> ^-n^-z^
[23:00:08] *** Arnald has quit IRC
[23:00:23] <e^ipi> how's it work if the ZIL becomes corrupt, say for example if you were to hypothetically put it on a CF card and the thing cratered because of too much activity ?
[23:00:59] <RElling> ZIL is write-only except during zpool import
[23:01:31] <RElling> if a ZIL write error occurs, ZFS will write them into the main pool, just like with no slog
[23:01:44] <e^ipi> exactly, CF cards can only be written to so many times
[23:02:04] <HCoyote> should I see actv > 1 whenever I see active writes occurring?  all but two or three disks are staying at or below 0.5
[23:02:15] <kjetilho> e^ipi: CF isn't very fast anyway
[23:02:26] *** estibi has quit IRC
[23:02:37] <RElling> yeah, but as long as it has some spare room, the wear leveling should work well
[23:03:10] <elektronkind> NAND flash is the norm now and last longer than NOR flash
[23:03:24] <RElling> flash SSDs can have even more reliable data store, it just costs bits
[23:03:32] *** Triskelios has joined #opensolaris
[23:03:33] <kjetilho> since ZIL is write-only in ordinary operation, the 0ms access time isn't very useful.  so you get a, what?  40 MB/s device?
[23:04:06] <RElling> NB modern rotating rust drives have lots of ECC
[23:04:42] <kjetilho> the SanDisk Extreme III is 20 MB/s
[23:05:27] <holcomb> http://www.umem.com/Umem_NVRAM_Cards.html
[23:05:28] <RElling> one thing we don't have much data on is how busy the ZIL really is.  need to crank up a script to measure it...
[23:05:45] <holcomb> when slog becomes part of regular solaris - the thumpers should just come with those installed
[23:06:02] <kjetilho> ah, the Sandisk Extreme Ducati Edition is 45 MB/s
[23:06:14] <elektronkind> Ducati edition?
[23:06:15] * elektronkind groans
[23:06:29] <kjetilho> it's a lovely red!
[23:06:35] <the-decider> it runs up and down the interstate in packs.
[23:07:02] <kjetilho> how big a ZIL do you actually need?
[23:07:42] *** cypromis has joined #opensolaris
[23:07:55] <kjetilho> RElling: there are two Dtrace snippets at http://blogs.sun.com/realneel/entry/the_zfs_intent_log
[23:08:56] *** victori_ has joined #opensolaris
[23:10:32] <RElling> kjetilho: excellent question!  The answer is a function of the workload (how many sync writes), the txg commit interval (5 s), and the speed at which the txg is actually committed.  I recall lobbying for  a high water mark, but even that is difficult to turn into some reasonable action plan
[23:10:44] <HCoyote> can the txg commit interval be increased?
[23:11:03] <sommerfeld> yah, there isn't good visibility as to what the high water mark is
[23:11:10] <RElling> for now, we're kicking around the thought of 10 seconds worth of sync writes
[23:11:23] <holcomb> Irssi 0.8.11 (20070425) - http://irssi.org/ level
[23:11:27] <holcomb> whoops
[23:11:30] *** cypromis has quit IRC
[23:11:30] <sommerfeld> RElling as in 2 * txg-commit-interval ?
[23:11:34] <RElling> yep
[23:11:41] <kjetilho> okay, so with a throughput of even 500 MB/s, 8 GB of NVRAM is enough
[23:11:42] *** hile_ has joined #opensolaris
[23:13:03] <RElling> something like that.
[23:13:56] <Triskelios> can I use wildcards in Studio mapfiles?
[23:14:10] <Triskelios> I can't find anything that actually describes the format
[23:15:51] *** FelipeBare has quit IRC
[23:18:30] *** theRealballchalk has joined #opensolaris
[23:18:57] <theRealballchalk> hey guys how do i rm -rf  a dir with a space in the end?
[23:19:28] <jmcp> quote the name of the directory
[23:19:29] <sommerfeld> there are many ways.  rm -rf "dir_with_space "
[23:19:32] <theRealballchalk> it looks like i'm trying to rm -rf "blah\ blah\ /"
[23:19:40] <theRealballchalk> but won't go away
[23:20:01] <sommerfeld> and "ls | cat -v " may prove instructive (in case it isn't a space..)
[23:20:05] <theRealballchalk> jmcp: ok ill try
[23:21:42] <theRealballchalk> jcmp:  i did 'rm -rf "blah blah " it doesn't return anything and it's still there
[23:21:43] <theRealballchalk> ?
[23:22:31] <RElling> rename it
[23:23:01] <RElling> mv blah* toastme
[23:23:02] <theRealballchalk> RElling: yeeeeea i can but.......that would defeat my purpose of learning :)
[23:23:37] <theRealballchalk> it thinks the space in the end is not a char
[23:23:41] <RElling> ls blah* | od -c
[23:23:50] <theRealballchalk> ok hold on
[23:23:51] <tomww> try with filename-completion...
[23:24:06] <theRealballchalk> tomww: yea i know
[23:24:17] <tomww> bash ... then ls blah<tab><tab>    should give you the needed escape-chars visible
[23:24:22] <theRealballchalk> despite file name completion it's what's happening
[23:24:35] <theRealballchalk> yea
[23:24:47] <RElling> it used to be fun to put backspace characters in filenames :-)
[23:25:07] <elektronkind> it
[23:25:25] <elektronkind> that's how the warez kiddies used to hide their directories on open FTP sites
[23:25:35] <theRealballchalk> RElling: ls blah* |od -c gives me 0000000
[23:25:41] <kjetilho> I like to put a LF at the end of my directory names.  then I make a subdirectory called etc
[23:25:51] <HCoyote> can a single spare be allocated to different zpools or must I have one spare per zpool?
[23:26:08] * the-decider likes to start all of his directories with "..^H^G"
[23:27:13] <RElling> adding ";rm -rf*" to the end of a filename can find bugs in clean scripts :-)
[23:27:22] *** Pedro1 has quit IRC
[23:27:38] <elektronkind> HCoyote: a spare can be assigned to a specific pool, shared amonst a set of pools, or globally available to all pools
[23:27:39] <sommerfeld> RElling: and that's a classic.  multics had bugs like that
[23:27:42] <tomww> eval is your friend :-)
[23:28:26] *** bubbva has joined #opensolaris
[23:31:24] *** sahafeez has quit IRC
[23:31:36] *** sahafeez has joined #opensolaris
[23:35:08] <HCoyote> cool, thanks.
[23:35:33] <HCoyote> BTW, anyone here going to Usenix LISA in November?
[23:37:41] <the-decider> also on the conference bandwagon -- Internet2 MM & Educause? ;)
[23:42:09] <elektronkind> where is LISA this year?
[23:42:17] <the-decider> Lisa lives in Oregon now.
[23:42:23] <the-decider> oh!  LISA...
[23:42:32] <hile_> who is Lisa? :)
[23:42:35] <hile_> and is she hot?
[23:42:56] <the-decider> which one? ;)
[23:43:03] *** luc_ has joined #opensolaris
[23:43:12] <elektronkind> ugh! in DFW
[23:43:14] <the-decider> Ew.  Dallas
[23:43:17] <the-decider> bleach
[23:43:18] <elektronkind> gack
[23:43:26] <hile_> hey, nothing wrong with being in dallas
[23:43:28] <kjetilho> Dallas is hot
[23:43:42] <hile_> if you're lucky, you'll catch the boys playing on Sunday
[23:43:44] <elektronkind> dallas is the anus of the US
[23:44:00] <the-decider> hmm, Middleware 2007 in Newport Beach
[23:44:18] <the-decider> omg, look who's a sponsor...
[23:44:31] <elektronkind> BBN!
[23:44:39] <the-decider> wow, I haven't looked at the usenix site in eons.
[23:44:58] <the-decider> elektronkind: heh, I'm sure that's a mistake... must be using the wrong logo?  They're still around!?
[23:44:58] *** luc_ has quit IRC
[23:45:16] <elektronkind> yeah, I was wondering about that, and why they'd be interested in middleware
[23:45:21] <the-decider> didn't they become GTE Internetworking, and then .bomb?
[23:45:28] <elektronkind> yeah
[23:45:32] <hile_> BBN?
[23:45:39] <the-decider> integrating with state unemployment agencies/
[23:45:44] <elektronkind> bbnplanet -> gte -> verizon, I think
[23:45:55] <sommerfeld> the-decider: actually ran into a high school classmate last spring.  he works for the BBN research group, which apparently survived the .bomb intact
[23:46:19] <the-decider> ah, so BBN dumped their network operations, but kept their research end?
[23:46:24] <elektronkind> I used to trade news feeds with BBN
[23:46:29] <elektronkind> out of cabridge
[23:46:32] <elektronkind> er cambridge
[23:46:34] <sommerfeld> the-decider: something like that, yes.
[23:46:54] <elektronkind> that's wacky
[23:47:05] <the-decider> hmm -- there's the LSF'08 (Linux Storage & Filesystem Workshop).  Ya'll should show up in mass and gloat ;)
[23:47:10] <elektronkind> Internet2, sponsored by Williams
[23:47:19] <elektronkind> ok, maybe that's too obscure
[23:47:42] *** cypromis has joined #opensolaris
[23:48:20] <theRealballchalk> hey guys this is erally weird
[23:48:37] <the-decider> the last LISA I went to was '99
[23:48:49] <theRealballchalk> i just manipulated that Busi\ Comp\ / folder name to ./test and i can move it to ..
[23:48:54] <elektronkind> LSF seems interesting
[23:49:01] <theRealballchalk> but now i can't even rm -rf test ???? what's the matter?
[23:49:22] <elektronkind> too bad it's linux-specific
[23:50:33] <elektronkind> theRealballchalk: if you think all hope is lost, please report to the clri man page
[23:51:16] <elektronkind> you'll need the output from ls -li to aid you in that
[23:52:06] <theRealballchalk> yea ok
[23:52:07] <theRealballchalk> thanks though
[23:52:22] <theRealballchalk> no hope lost here
[23:52:44] <theRealballchalk> it's just an anomaly haa
[23:52:48] <theRealballchalk> jk
[23:52:59] <elektronkind> basically, it's an inode zapper. it's a "break glass in case of emegency" type of utility
[23:53:41] <sommerfeld> it's more like the "blow hole in wall because door isn't working" type of utility.
[23:53:47] <elektronkind> heh
[23:53:55] <theRealballchalk> brb
[23:53:56] <theRealballchalk> class
[23:53:59] *** theRealballchalk has left #opensolaris
[23:56:39] *** mikefut has quit IRC
[23:58:16] <tomww> nice to see SFEzfs-auto-backup.pkg downloaded at least by 3 people :-)
[23:58:54] <tomww> (build on Tim Fosters's zfs-auto-backup scripts)
[23:59:21] *** jpdrawneek has joined #opensolaris
[23:59:22] <madhatter> tomww: What's the scope of that script?
[23:59:41] *** axisys has quit IRC





top