Switch to DuckDuckGo Search
   March 20, 2012  
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >

Toggle Join/Part | bottom
[00:08:25] *** randw has joined #illumos
[00:09:26] *** fujin has quit IRC
[00:10:28] *** randw_ has quit IRC
[00:13:07] *** randw has quit IRC
[00:14:43] *** elijah-mbp has quit IRC
[00:15:53] *** elijah-mbp has joined #illumos
[00:17:01] *** Italux has quit IRC
[00:19:10] *** denizr has quit IRC
[00:19:17] *** _ams_ has quit IRC
[00:19:53] *** denizr has joined #illumos
[00:24:18] *** denizr has quit IRC
[00:27:10] *** mjr_ has joined #illumos
[00:38:34] *** stevel_ has joined #illumos
[00:40:09] *** gila_ has joined #illumos
[00:40:21] *** hajma_ has quit IRC
[00:40:21] *** gila has quit IRC
[00:40:21] *** stevel_ has quit IRC
[00:40:21] *** gila_ is now known as gila
[00:40:41] *** hajma_ has joined #illumos
[00:41:07] *** randw has joined #illumos
[00:41:42] *** raichoo has quit IRC
[00:42:56] *** spartha has joined #illumos
[01:02:52] *** andy_js has quit IRC
[01:03:04] *** carmark has quit IRC
[01:03:08] *** Bleomycin has joined #illumos
[01:04:14] *** Bleomycin has quit IRC
[01:09:49] *** lewellyn is now known as lart
[01:11:14] *** lart is now known as ihateyou
[01:12:43] *** Ali_G_Indahouse has quit IRC
[01:15:57] *** ihateyou is now known as lewellyn
[01:22:11] *** dimdm has quit IRC
[01:29:16] *** mjr_ has quit IRC
[01:40:09] *** denizr has joined #illumos
[01:40:23] *** denizr has quit IRC
[01:44:37] *** denizr has joined #illumos
[02:03:58] *** sebasp has quit IRC
[02:04:16] *** sebasp has joined #illumos
[02:05:18] *** carmark has joined #illumos
[02:05:38] *** jelmd has quit IRC
[02:26:40] *** splank has joined #illumos
[02:27:01] <splank> any SmartOS fiends around?
[02:40:28] *** darjeeling has joined #illumos
[02:41:05] <konobi> depends what you mean by fiend
[02:44:00] <splank> like somebody interested in it enough to know how to get me in to this VM I made.
[02:44:26] <splank> I imported the joyent ubuntu dataset.
[02:44:33] <splank> I get that it wants me to vnc in.
[02:44:53] <splank> the vm says it's running but my vnc client is just hanging..
[02:46:46] <splank> the port I'm supposed to be connecting to doesn't seem to be listening..
[02:47:18] <splank> oh, no, I take that back
[02:47:34] <splank> if I telnet to it, I get RFB 003.008, so it's at least trying...
[02:49:20] <splank> maybe the macosx vnc is teh sucks
[02:51:49] *** mtzaurus has quit IRC
[02:52:24] *** sross has joined #illumos
[02:52:39] <konobi> the VNC IP is the one set for your host machine, right?
[02:52:47] <splank> correctamundo
[02:53:05] <splank> and the port was some random crazy offset.
[02:53:49] <splank> but my built-in mac vnc client just stood there like an old man standing over a ditch looking at his dead dog.
[02:54:06] <konobi> Chicken works pretty well
[02:54:19] <splank> I decided to brew install tiger-vnc; chicken's next...
[02:54:29] *** deirdres has quit IRC
[02:54:38] <konobi> "chicken" rather than "chicken of the vnc" that is
[02:54:42] <splank> unfortunately now the vnc info has disappeared from the vm's info list
[02:54:53] <splank> so I guess I gotta shut it down and start it up again?
[02:59:11] <splank> wow, took all that time to halt.
[03:00:15] <konobi> yeah, you'll probably want to install acpid
[03:00:55] <splank> wow, sweet, got vnc to connect!
[03:01:16] <splank> but it's just a screen with boot failed:could not read from cdrom
[03:01:19] <splank> no bootable device
[03:01:31] <splank> does that mean I have to grab an ISO?
[03:02:06] <konobi> if it's a joyent image, you need to boot from disk rather than cd
[03:02:10] <splank> I was thinking that the latest smartOS obviated the need for that messing about
[03:02:15] <splank> ah, ok.
[03:02:29] <splank> it tried that further up the boot sequence
[03:02:45]
[03:02:56] <splank> "Boot failed: not a bootable disk"
[03:03:05] <konobi> odd
[03:03:42] <splank> I skipped the part about CURLing the iso from ubuntu because I thought that dsadm import did that for me. Was I wrong?
[03:04:06] <e^ipi> splank: you probably don't want to use our linux datasets, since you have no way of logging in to them
[03:04:41] <splank> zoiks. So importing your datasets is useless? It was so tempting!
[03:04:42] <e^ipi> http://wiki.smartos.org/display/DOC/How+to+create+a+Virtual+Machine+in+SmartOS
[03:05:34] <splank> isn't there a tacky way of dumping a vnc password hash onto the host's filesystem though :)
[03:05:58] <e^ipi> possibly
[03:06:05] <e^ipi> how to do that is left as an exercise to the reader.
[03:06:09] <splank> bummer.
[03:06:34] <splank> it'd be pretty huge for us to be able to keep parity with Joyent in our footings around here...
[03:06:45] <splank> hence using their stock datasets, etc.
[03:11:43] *** proberts has quit IRC
[03:12:40] *** liberal has joined #illumos
[03:13:56] *** enmand has quit IRC
[03:15:48] *** proberts has joined #illumos
[03:20:52] <splank> so is the dsadm import command not equivalent to downloading a system image?
[03:23:18] <jeffpc> did solaris have savecore -L before the parallel dump was added?
[03:24:00] <lewellyn> jeffpc: when was it added?
[03:24:25] <jeffpc> I'm looking to undo the brain damage
[03:24:26] <jeffpc> https://hg.openindiana.org/upstream/illumos/illumos-gate/file/eb166ea6e74e/usr/src/uts/common/os/dumpsubr.c
[03:24:28] <Triskelios> jeffpc: I'd assume savecore -L is ancient
[03:24:43] <jeffpc> DF_LIVE seems to be always set
[03:24:52] <jeffpc> err https://hg.openindiana.org/upstream/illumos/illumos-gate/diff/eb166ea6e74e/usr/src/uts/common/os/dumpsubr.c
[03:25:20] <lewellyn> jeffpc: i can confirm -L is in S10.
[03:25:32] <lewellyn> that's all i can confirm at the moment. food time is imminent.
[03:25:34] <jeffpc> lewellyn: ok, thanks
[03:25:41] <jeffpc> lewellyn: that's good enough
[03:25:45] <lewellyn> figured :)
[03:26:41] <jeffpc> ah, sometimes it's also cleared
[03:30:01] <splank> anyone care about spare?
[03:30:12] <splank> SPARC (dang autocorrect!)
[03:32:52] <LeftWing> e^ipi does
[03:33:13] <e^ipi> hah!
[03:33:18] <LeftWing> :D
[03:33:41] <splank> Yikes.
[03:34:22] <splank> So I guess hanging on to my collection of original SPARCStation 1+'s is kinda futile at this juncture.
[03:34:37] <LeftWing> It's not if you want to run old SunOS releases.
[03:34:49] *** DrLou has quit IRC
[03:35:07] <splank> nah, I want to run brand spanking new Illumos.
[03:35:18] <LeftWing> Then you'll be wanting to get some sort of 64bit x86.
[03:35:18] <e^ipi> if you feel like burning power on useless hardware, hang on to those puppies
[03:35:32] <Triskelios> an SS1 was already merely a historical curiosity a decade ago
[03:35:39] <splank> But! My nice beefy three-headed SunBlade 2000?!?
[03:35:53] <splank> It was really cool a decate ago!
[03:37:32] <splank> I even went to the pains of doing an AI install of opensolaris, cloning the zpool and moving it into the SB2000 disk so I could run osol w/o the correct OBP version,
[03:38:10] <LeftWing> You don't actually need that OBP version if you hack up the AI media a little. :P
[03:38:15] <LeftWing> The wanboot requirement was dumb.
[03:38:20] <splank> oh, man. Now you tell me.
[03:38:46] <LeftWing> That said, it's all kind of pointless these days. If you're looking for a museum piece, run a museum OS.
[03:39:03] <LeftWing> It'll be more authentic.
[03:39:03] <richlowe> more so than the rest of us, that is.
[03:39:57] <jeffpc> dumpsubr.c | 723 ++++---------------------------------------------------------
[03:40:00] <jeffpc> 1 file changed, 48 insertions(+), 675 deletions(-)
[03:40:01] <jeffpc> take that!
[03:40:08] <jeffpc> and there's still so much to do!
[03:41:12] <splank> Ok, so my SPARCStation comment was an exaggeration. I would like to keep my fairly modern and nicely upgraded SunBlade 2000 going on latest stuff though. Is there any effort to resurrect the SPARC code?
[03:41:12] *** kart_ has joined #illumos
[03:41:19] *** marsell has quit IRC
[03:42:39] *** denizr has quit IRC
[03:42:57] *** denizr has joined #illumos
[03:43:14] <alanc> "fairly modern" is another exaggeration for a 10 year old machine that was just a speed bump to a 12 year old design
[03:43:57] <richlowe> The SPARC code is not dead
[03:44:03] <LeftWing> splank: If you're keen to get involved in the SPARC effort, have a look at: https://www.illumos.org/projects/illumos-gate/wiki/SPARC_resources
[03:44:04] <richlowe> It's just that the hardware is rare, and crap
[03:44:08] <richlowe> and it's hard to find people who give a crap.
[03:44:16] <richlowe> and I say this as, thus far, one of two people to have given a crap.
[03:44:25] <richlowe> it builds, it boots.
[03:44:30] <alanc> especially people who give a crap enough to dig in and work on it
[03:44:44] <richlowe> the things that used to work still do.
[03:45:19] <richlowe> If you really want to get involved with SPARC, the most useful thing would be actively caring about it, and for it.
[03:46:02] *** randw has quit IRC
[03:46:06] <splank> I haven't chosen a 20% project yet and I truly thought SPARC + OBP was a nicer bit of kit than PC, so was wanting to work on it. Is it something that one person like me could help with?
[03:46:39] <Alasdairrr> I forsee SPARC support gradually dying, modern SPARC kit is so darn expensive anyone that buys it will want to run Solaris 11 on it
[03:46:56] <Alasdairrr> its going to be the second hand hobbiest market that runs second hand SPARC kit
[03:47:09] <LeftWing> Plus I imagine it will be hard to buy without a Solaris 11 license anyway, so you may as well.
[03:47:29] <Alasdairrr> Second hand SPARC kit costs more than modern fast Intel kit anyway
[03:47:33] <Alasdairrr> so why bother
[03:47:42] <splank> Ok, now I'm really depressed.
[03:47:44] <Alasdairrr> except to go "Look at me, I'm running SPARC!!"
[03:47:49] <Alasdairrr> sorry splank :-)
[03:47:49] <LeftWing> If all of that turns out to be wrong, and SPARC is suddenly fascinating, then people will emerge with a vested interest in doing the work -- as they have for x86.
[03:47:56] <jeffpc> I really wish there was an alternative to x86
[03:48:10] <Alasdairrr> ARM seems to be the next best thing
[03:48:11] <LeftWing> That said, if you're keen to be in on all the SPARC action, go to that wiki page and get involved!
[03:49:09] <Alasdairrr> what with dual core GHz based ARM cpus being present on lots of devices these days
[03:49:15] <alanc> not sure how much longer you'll be able to buy SPARC kit that illumos has support for, since I think that's just T3 from the current price list, since the gate closed before T4 - not sure about the M-series side though
[03:49:16] <splank> Honestly, when I used to program in assembler, I cared about stuff like this. Now I don't as much, except on principle. The big issue for me as a ( mostly ) sysadmin is that I don't want to deal with freaking BIOS. I miss OBP a lot. Why couldn't someone make a PC server with OBP?
[03:49:45] <Alasdairrr> splank: UEFI is the future
[03:50:00] <Alasdairrr> http://en.wikipedia.org/wiki/Unified_Extensible_Firmware_Interface
[03:50:03] <Alasdairrr> on x86 anyway
[03:50:13] <jeffpc> Alasdairrr: UEFI has major issues
[03:50:22] <jeffpc> (implementations blow)
[03:50:25] <Alasdairrr> i bet
[03:50:34] <LeftWing> splank: If you were an x86 server maker, why would you buck the trend? It's going to cost money and be of questionable use.
[03:50:39] <alanc> though from the Linux guys I know, UEFI support requires serious amounts of gin to deal with
[03:50:40] <jeffpc> Alasdairrr: as in, they blatantly violate the spec
[03:50:42] <Alasdairrr> Anything that involves vendors hiring contract programmers will blow
[03:51:08] <LeftWing> BIOS sucks, but it's "well understood."
[03:51:17] <alanc> mjg59 has some wonderful blogs & youtube'd talks on UEFI
[03:51:41] <Alasdairrr> i will have to add that to my investigate list
[03:51:42] <splank> The bios I was using today (piped through a serial line) actually wanted me to use control-S to save. Who comes up with this stuff?
[03:52:21] <alanc> programmers in whatever third-world country got the lowest bid in for the latest updates
[03:52:47] <jeffpc> 1 file changed, 69 insertions(+), 718 deletions(-)
[03:53:40] <Triskelios> OBP still lives, at least on "legacy-free" systems like OLPC
[03:53:47] <Triskelios> UEFI blows in every possible way
[03:54:05] <LeftWing> awesome
[03:59:58] *** proberts has quit IRC
[04:07:53] *** splank has quit IRC
[04:09:05] *** kart_ has quit IRC
[04:10:44] *** kart_ has joined #illumos
[04:12:43] *** descipher has quit IRC
[04:13:11] *** descipher has joined #illumos
[04:16:29] *** kart_ has quit IRC
[04:18:56] *** kart_ has joined #illumos
[04:19:11] *** libkeiser has joined #illumos
[04:22:17] *** libkeise1 has quit IRC
[04:40:39] <jbergstroem> date
[04:40:41] <jbergstroem> sry
[04:45:33] *** carlomagno has quit IRC
[04:49:10] *** mavhc has quit IRC
[04:50:07] *** deirdres has joined #illumos
[04:55:09] *** mavhc has joined #illumos
[04:56:59] *** seanmcg has quit IRC
[04:56:59] *** hajma_ has quit IRC
[04:57:24] *** seanmcg has joined #illumos
[04:57:26] *** hajma_ has joined #illumos
[05:01:21] *** carlomagno has joined #illumos
[05:01:34] *** carmark has quit IRC
[05:02:16] *** alhazred has quit IRC
[05:06:00] *** kart_ has quit IRC
[05:08:36] *** jbergstroem has quit IRC
[05:09:48] *** kart_ has joined #illumos
[05:11:20] *** jbergstroem has joined #illumos
[05:12:23] *** darjeeling has quit IRC
[05:19:30] *** yffbrave has joined #illumos
[05:22:21] *** voidcoder has quit IRC
[05:28:03] *** jelmd has joined #illumos
[05:32:17] *** seanmcg has quit IRC
[05:32:41] *** hajma_ has quit IRC
[05:32:41] *** seanmcg has joined #illumos
[05:33:08] *** hajma_ has joined #illumos
[05:38:39] *** HarD_RocK_PenanG has joined #illumos
[05:38:48] *** HarD_RocK_PenanG has left #illumos
[05:51:56] *** marsell has joined #illumos
[06:07:56] <eye-scuzzy> moin
[06:21:39] *** darjeeling has joined #illumos
[06:25:39] *** darjeeling has quit IRC
[06:25:40] *** darjeeli_ has joined #illumos
[06:28:01] *** POloser has joined #illumos
[06:31:54] *** darjeeling has joined #illumos
[06:31:55] *** darjeeli_ has quit IRC
[06:33:06] *** darjeeling has quit IRC
[06:37:11] *** denizr has quit IRC
[06:54:09] *** relling has quit IRC
[07:00:21] *** sross has quit IRC
[07:02:05] *** gila has quit IRC
[07:02:46] *** marsell has quit IRC
[07:04:27] *** jw_urodoc has joined #illumos
[07:05:57] *** jw_urodoc has quit IRC
[07:27:35] *** hajma_ has quit IRC
[07:30:18] *** tsoome has quit IRC
[07:31:25] *** alanc has quit IRC
[07:42:22] *** alhazred has joined #illumos
[07:45:32] *** mamash has joined #illumos
[07:53:23] *** MerlinDMC has joined #illumos
[08:05:30] *** spartha has quit IRC
[08:15:41] *** tsoome has joined #illumos
[08:21:38] *** marsell has joined #illumos
[08:24:12] *** Botanic has quit IRC
[08:25:32] *** tsoome_ has joined #illumos
[08:25:32] *** tsoome has quit IRC
[08:25:32] *** tsoome_ is now known as tsoome
[08:27:38] *** gwr_ has quit IRC
[08:47:22] *** Worsoe has joined #illumos
[08:55:21] *** voidcoder has joined #illumos
[09:13:15] *** darjeeling has joined #illumos
[09:34:40] *** darjeeling has quit IRC
[09:42:55] *** marsell_ has joined #illumos
[09:44:01] *** marsell has quit IRC
[09:44:15] *** marsell_ is now known as marsell
[09:59:47] *** darjeeling has joined #illumos
[10:04:37] *** kart_ has quit IRC
[10:05:38] *** kart_ has joined #illumos
[10:22:16] *** darjeeling has quit IRC
[10:22:26] *** _ams_ has joined #illumos
[10:33:13] *** |AbsyntH| has joined #illumos
[10:46:17] *** andy_js has joined #illumos
[10:55:14] *** pheleven has quit IRC
[10:58:26] <cypromis> nvidia suxx
[11:02:13] *** Italux has joined #illumos
[11:10:48] <eye-scuzzy> cypromis: why ?
[11:12:13] *** carmark has joined #illumos
[11:13:21] <buffyg> Moin.
[11:16:59] <buffyg> jeffpc: savecore -L goes at least as far back as Solaris 7 IIRC.
[11:17:02] <cypromis> hehe
[11:17:05] <cypromis> MCP73
[11:17:08] <cypromis> I just hate it
[11:17:30] <cypromis> I remember there was a trick to wake it from sleep so that oi can use it wiut the nfo driver
[11:17:34] <cypromis> but I cannot remember what it was
[11:17:41] <cypromis> so now i have a sleeping NIC
[11:20:36] <eye-scuzzy> cypromis: http://www.ynp.net/Solaris86/hiber
[11:21:08] <eye-scuzzy> wake up net,snd after sleep
[11:27:01] <cypromis> its not that kind of sleep
[11:27:20] <cypromis> at least plumb/unplumb and reloading driver is not the point
[11:28:50] <lewellyn> cypromis: i rarely had problems with nfo and mcp7a after about 121
[11:29:05] <lewellyn> is this recent behavior for nfo for you?
[11:29:19] * lewellyn can bring up a box with nfo tomorrow, to verify, i suppose
[11:46:26] <cypromis> yes
[11:46:28] <cypromis> 2.2.5
[11:46:29] <cypromis> sorry
[11:46:30] <cypromis> 2.2.6
[11:46:33] <cypromis> and its oi 151a
[11:46:52] <cypromis> downloaded from the site tho since I have no otehr network connection on that box
[11:46:59] <cypromis> so cannot use the package of coutse
[11:47:01] <cypromis> course even
[11:54:36] <cypromis> maybe the driver in the oi package works better ?
[11:59:47] <lewellyn> oh. that's right. it uses nge by default now.
[11:59:53] <lewellyn> yeah. nfo sucks. use nge :D
[11:59:58] * lewellyn mixed them up.
[12:00:11] <MGSsancho> neon genesis evangelion?
[12:00:12] <lewellyn> (nge also sucks, but differently and overall less so)
[12:00:24] <lewellyn> MGSsancho: less cool than that
[12:00:39] <MGSsancho> hehehe
[12:00:55] <lewellyn> cypromis: basically, though, your guess is correct: the ethernet blows. :)
[12:01:32] *** pheleven has joined #illumos
[12:02:06] <cypromis> hmmm
[12:02:12] <cypromis> nge never recognised the card
[12:02:19] <cypromis> ok will try to force it
[12:03:14] <lewellyn> hm. i thought mcp73 should be supported by newer nge
[12:03:41] *** enmand has joined #illumos
[12:03:44] <lewellyn> i'm not logged into anything with useful-for-this man pages atm though.
[12:05:55] <cypromis> oi151a seems to not know about that#;D
[12:06:29] <lewellyn> if i weren't doing great evil right now, i'd look :)
[12:07:45] <cypromis> :P
[12:10:13] <lewellyn> i'm not even going to share the evil here, because it's way unrelated to illumos and people would offer me booze to ease my pain ;)
[12:24:37] *** voidcoder has quit IRC
[12:27:07] *** nerens has joined #illumos
[12:31:56] *** hajma_ has joined #illumos
[12:32:26] <cypromis> enforcing drivers always helps
[12:32:26] <cypromis> lol
[12:32:31] <cypromis> especialy with nforce
[12:40:01] *** nerens has quit IRC
[12:45:52] *** gila has joined #illumos
[12:47:47] *** kimc has quit IRC
[12:52:53] *** kimc has joined #illumos
[12:52:53] *** kimc has joined #illumos
[12:55:21] *** himsin has quit IRC
[12:56:29] *** POloser has left #illumos
[12:57:44] *** eschrock has joined #illumos
[13:05:56] *** vortexx has quit IRC
[13:25:34] *** zombietch has joined #illumos
[13:28:28] *** man_u_ has joined #illumos
[13:29:16] *** eschrock1 has joined #illumos
[13:30:03] *** andy_js_ has joined #illumos
[13:30:58] *** man_u has quit IRC
[13:30:59] *** man_u_ is now known as man_u
[13:31:08] *** eschrock has quit IRC
[13:31:15] *** arekinath has quit IRC
[13:31:24] *** arekinath has joined #illumos
[13:31:25] *** arekinath has joined #illumos
[13:31:38] *** andy_js has quit IRC
[13:31:38] *** andy_js_ is now known as andy_js
[13:31:42] *** rennj has quit IRC
[13:32:06] *** rennj has joined #illumos
[13:32:36] *** voidcoder has joined #illumos
[13:38:52] *** gwr_ has joined #illumos
[13:39:52] <buffyg> Has Redmine died? I'm having problems loading the tracker.
[13:44:31] <Tenzer> It could seem like it
[13:47:20] *** proberts has joined #illumos
[13:47:55] *** marsell has quit IRC
[13:47:57] <lewellyn> either i no longer get email when redmine dies, or the emailing thing doesn't know
[13:47:58] *** krystal has quit IRC
[13:48:42] <buffyg> lewellyn: The front-end proxies seem to be confused for lack of response. Can you poke it with a stick?
[13:48:50] <lewellyn> i cannot.
[13:49:15] <buffyg> <pleadingly> Anyone from infra around?
[13:49:18] <lewellyn> i've never had permissions to do anything magic
[13:49:26] <lewellyn> buffyg: i've asked elsewhere already :)
[13:49:39] <buffyg> lewellyn: Ta
[13:49:57] <lewellyn> ah. there's the email.
[13:50:17] <lewellyn> guess they come 15 minutes after the failure starts
[13:50:50] <buffyg> Guess it's time for a Simpsons break.
[13:53:40] <Triskelios> lewellyn: since you're around, mind setting the channel onjoin message to mention the channel is logged?
[14:01:00] <lewellyn> hm. i wonder if i have permissions to that
[14:02:12] <lewellyn> hrmph. freenode services suck
[14:02:14] *** lewellyn has left #illumos
[14:02:18] *** lewellyn has joined #illumos
[14:02:18] *** ChanServ sets mode: +o lewellyn
[14:02:55] <lewellyn> IS there an entrymsg right now? i didn't get one :/
[14:03:03] <lewellyn> i don't want to clobber one if there is one.
[14:03:12] *** Triskelios has left #illumos
[14:03:16] *** Triskelios has joined #illumos
[14:03:30] <Triskelios> I don't see one, either
[14:03:43] <lewellyn> well, we'll assume that means that there isn't one
[14:04:16] <lewellyn> didn't get one for #opensolaris either. and i know that channel has one
[14:04:21] * lewellyn grumbles
[14:04:54] *** MerlinDMC has left #illumos
[14:04:58] *** MerlinDMC has joined #illumos
[14:05:44] <MerlinDMC> i don't get one ... but i get one at #node.js ;)
[14:06:22] *** lewellyn^ has joined #illumos
[14:06:22] *** ChanServ sets mode: +o lewellyn^
[14:06:27] *** DreamCatcher has quit IRC
[14:06:53] <lewellyn> maybe osol's is gone now too.
[14:07:15] *** [1]kdavyd has joined #illumos
[14:08:13] <lewellyn> 06:08 [freenode] -ChanServ(ChanServ at services dot )- You are not authorized to execute this command.
[14:08:26] *** kdavyd has quit IRC
[14:08:27] *** [1]kdavyd is now known as kdavyd
[14:08:41] *** lewellyn^ has quit IRC
[14:09:26] *** DreamCatcher has joined #illumos
[14:13:38] <buffyg> That's an episode of the Simpsons down. Hmm, still no redmine...
[14:19:40] *** sross has joined #illumos
[14:19:42] <yffbrave> illumos.org is down?
[14:20:39] *** DrLou_ has joined #illumos
[14:22:05] <lewellyn> quite
[14:30:17] *** [1]kdavyd has joined #illumos
[14:30:24] *** kdavyd has quit IRC
[14:30:24] *** [1]kdavyd is now known as kdavyd
[14:44:01] *** mtzaurus has joined #illumos
[14:44:59] *** neophenix has joined #illumos
[14:45:33] *** kdavyd has quit IRC
[14:46:17] *** kdavyd has joined #illumos
[14:47:14] *** carlomagno has quit IRC
[14:52:12] *** carlomagno has joined #illumos
[14:53:45] <cypromis> hmm now nge0 is up
[14:53:51] <cypromis> and I switched to physical:default
[14:54:03] <cypromis> and ... somehow the routing does not seem to attach to the interface
[14:54:08] * cypromis is lost
[14:54:23] <cypromis> this is the weirdest piece of network equippment I have seen for some years
[14:54:24] <cypromis> lol
[14:54:37] <buffyg> Looks like redmine is back. Thanks to whomeverdidit.
[14:55:12] <tsoome> you need to configure IP on interface before you can set routing
[14:55:25] <cypromis> I know
[14:55:27] *** esproul has joined #illumos
[14:55:40] * cypromis looks at his AGS+ door holder
[14:56:13] <cypromis> the problem is that somewhere in the stack it just does not want to accept that I have a 192.168.181.42/255.255.255.0 interface
[14:56:19] <cypromis> and that its up
[14:56:32] <cypromis> when I ping any other address in my net
[14:56:35] <cypromis> it says network unreachable
[14:56:41] <cypromis> I hate nvidia
[14:56:52] <tsoome> what route are you trying to add?
[14:57:08] <cypromis> it does not even see its own route
[14:57:13] <tsoome> ?
[14:57:41] <cypromis> it says network unreachable when I ping its own net
[14:57:55] <tsoome> netstat -rn does not list 192.168.181.0?
[14:57:55] <cypromis> but the interface is up and unning
[14:58:01] <cypromis> and with snoop all I can see is arp traffic
[14:58:15] <cypromis> something is fishy and its not the switch since I already replugged it a couple of times
[14:58:24] <cypromis> ipfilter is off as well
[14:58:57] <tsoome> whats your netstat -rn output?
[14:59:32] <tsoome> or you can do netstat -rnv to see mask as well
[15:02:15] <cypromis> netstat -rnv shows it correctly
[15:02:27] <cypromis> that all traffic to the 192.168.181.0/255.255.255.0 net should go to nge0
[15:02:39] <cypromis> what is a bit weird is that the default route shows an ip address but no interface
[15:02:42] <cypromis> anyway
[15:02:46] <cypromis> if I ping any address in my network
[15:02:50] <cypromis> it says network unreachable
[15:02:53] <cypromis> which is wrong
[15:03:20] <tsoome> is interface marked UP in ifconfig output?
[15:04:10] <tsoome> you should see flags like UP,BROADCAST,RUNNING,MULTICAST,IPv4
[15:07:18] <elijah-mbp> illumos.org had a hiccup - fixed.
[15:07:36] *** MerlinDMC has quit IRC
[15:07:58] *** MerlinDMC has joined #illumos
[15:08:35] *** kdavyd has quit IRC
[15:09:49] *** agrajag` has joined #illumos
[15:10:29] *** kdavyd has joined #illumos
[15:13:22] *** d_k has quit IRC
[15:16:49] *** polar_ has joined #illumos
[15:22:27] *** krystal has joined #illumos
[15:26:17] *** polar_ has left #illumos
[15:35:22] <cypromis> yep
[15:39:26] <estibi> What does define __GNUC__ actually mean? Is the code compatible with GCC?
[15:39:59] <Triskelios> usually guards gcc-specific code
[15:41:46] *** ssmoot has joined #illumos
[15:42:58] <estibi> Triskelios: so a lot of code in usr/src/uts/intel/asm/ is GCC only :)
[15:43:38] <Triskelios> inline asm tends to be compiler-specific, yes
[15:52:27] <estibi> Triskelios: is there a reason to keep #if defined(__GNUC__) in let say usr/src/uts/intel/asm/mmu.h if the content is __GNUC__ only? We don't use SS version of for example getcr3 on x86.
[15:54:14] <Triskelios> it could gain other inlines in the future
[16:01:35] *** sross has quit IRC
[16:02:23] *** Worsoe has quit IRC
[16:03:54] *** allanblack has quit IRC
[16:03:58] <lewellyn> it could potentially gain other compilers in the future
[16:05:25] *** allanblack has joined #illumos
[16:05:27] *** sross has joined #illumos
[16:07:03] *** ssmoot has left #illumos
[16:07:50] *** denizr has joined #illumos
[16:12:47] *** kdavyd has quit IRC
[16:13:27] *** kdavyd has joined #illumos
[16:15:15] *** elijah-mbp has quit IRC
[16:16:03] *** Bdragon has joined #illumos
[16:18:09] *** tsoome has quit IRC
[16:20:02] *** joffe has joined #illumos
[16:22:55] *** joshie has joined #illumos
[16:24:52] *** voidcoder has quit IRC
[16:30:02] *** voidcoder has joined #illumos
[16:30:22] *** carmark has quit IRC
[16:34:10] *** relling has joined #illumos
[16:37:01] *** IRCMonkey_ has joined #illumos
[16:39:35] *** IRCMonkey_ has quit IRC
[16:39:48] *** sma78746 has joined #illumos
[16:48:16] <ksalman> How does zfs decide if a disk is bad? The autoreplace property on my pool is set to on, shouldn't it have replaced the seemingly bad one with the spare?
[16:48:21] <ksalman> http://pastebin.com/raw.php?i=C9p5M8YS
[16:49:14] *** d_k has joined #illumos
[16:49:27] <ksalman> It didn't, and the write performance went through the floor, because, I assume, that it was trying to read/write to the bad one
[16:50:04] <ksalman> so i manulally replaced it with the spare and it fixed the performance
[16:51:47] *** ryancnelson has joined #illumos
[16:52:25] <Tenzer> RoyK: You ought to know something about that ^^ :)
[16:52:33] <Alasdairrr> ksalman: Unfortunately there is no logic to do this
[16:52:54] <Alasdairrr> ksalman: ZFS is very bad at offlining dodgy disks, even when there is a lot of evidence available for it to work on
[16:53:01] <ksalman> oh i see
[16:53:12] <Alasdairrr> The main problem isn't ZFS so much as the SATA/SAS HBA drivers
[16:53:20] <Alasdairrr> and the scsi driver layer (sd)
[16:53:35] <Alasdairrr> What HBA are you using?
[16:54:44] <ksalman> I am using AOC-SAT2-MV8 SATA controllers
[16:54:45] *** yffbrave has quit IRC
[16:54:55] <ksalman> supermicro?
[16:54:59] <Alasdairrr> Yeah
[16:55:10] <Alasdairrr> I am familiar with them, PCI-x 133MHz
[16:55:11] <eclark> hey, that's the controller i have
[16:55:25] <ksalman> i think a lot of people are using those
[16:55:39] <eclark> they're fairly inexpensive and can do 8 drives
[16:55:40] <Alasdairrr> They work, but they aren't great - quite long in the tooth now
[16:55:49] <eclark> well, back when i bought it
[16:55:50] *** MerlinDMC has quit IRC
[16:55:57] <eclark> yeah, it's old now
[16:56:06] <ksalman> so i should stop buying them? :P
[16:56:14] <ksalman> what are good ones to buy now?
[16:56:20] <eclark> been looking at the M1015 recently
[16:56:30] <eclark> had that recommended by s0liddi
[16:56:33] <ksalman> everytime we need more ZFS storage at work i just buy those
[16:56:40] <ksalman> so i should probably upgrade
[16:56:56] *** ryancnelson has left #illumos
[16:57:03] <eclark> Alasdairrr: any input on the subject?
[16:58:11] <Alasdairrr> LSI
[16:58:19] <Alasdairrr> Get LSI
[16:58:28] <Alasdairrr> They are the best supported cards in Illumos
[16:58:33] *** ssmoot has joined #illumos
[16:58:36] <ksalman> any specific model?
[16:58:57] <Alasdairrr> Any that use the newer mpt_sas open source driver, claiming to be SAS2008 or newer
[16:59:01] <cypromis> ok I gave up on the nge and ordered 2 new intel cards
[16:59:05] <Alasdairrr> Avoid the older 3081 chipset
[16:59:20] <Alasdairrr> One sec and i'll find you some production model numbers
[16:59:27] <s0liddi> those m1015's use LSI2008 and can be flashed to non-raid mode :)
[16:59:29] <eclark> i think that's what i was looking at
[16:59:46] <eclark> yeah, m1015 is the same?
[17:00:16] *** ssmoot has left #illumos
[17:00:52] <Alasdairrr> Any of these 6Gb/s ones: http://www.lsi.com/products/storagecomponents/Pages/HBAs.aspx
[17:01:03] <ksalman> tyanks
[17:01:10] <ksalman> thanks even
[17:01:44] <Alasdairrr> np! :-)
[17:03:08] <s0liddi> yeah the m1015 rebranded 9210-8i
[17:07:58] <ksalman> how would i go about finding if i am bottlenecked by the 3Gb/s?
[17:10:19] <eclark> are your drives 6Gb/s? I don't think most single drives can saturate a 3Gb/s link…
[17:10:23] <eclark> save SSDs
[17:12:02] <Alasdairrr> the 6Gb/s ones guarantee you get the mpt_sas driver
[17:12:09] <Alasdairrr> the 3Gb/s ones get you the older closed source mpt driver
[17:12:20] <Alasdairrr> The mpt driver has known breakage
[17:12:50] <ksalman> oh i see
[17:14:52] *** vortexx has joined #illumos
[17:15:17] *** tsoome has joined #illumos
[17:17:09] *** Bdragon has quit IRC
[17:17:27] *** maheshwar_ has joined #illumos
[17:21:04] *** himsin has joined #illumos
[17:24:53] *** tomww has quit IRC
[17:26:35] *** tomww has joined #illumos
[17:27:13] *** raichoo has joined #illumos
[17:28:23] *** hajma_ has quit IRC
[17:30:28] *** Bdragon has joined #illumos
[17:33:19] *** enricop has joined #illumos
[17:33:38] <enricop> hi all. i have a ralink rt2860 pci wifi card
[17:33:49] *** eschrock1 has quit IRC
[17:33:55] <enricop> i have installed the rwn driver and reboot the machine
[17:34:16] <enricop> however it is not listed with dladm show-phys
[17:34:34] *** eschrock has joined #illumos
[17:35:06] <enricop> i am using the 151a2 version
[17:35:21] <enricop> i see in the rwm pkg manifest http://pkg.openindiana.org/prestable/info/0/driver%2Fnetwork%2Frwn%400.5.11%2C5.11-0.151.1.2%3A20120209T214719Z
[17:35:28] <Triskelios> "installed the rwn driver"? doesn't it come with rwn?
[17:35:57] <enricop> that it requires a firmware rt2860.ucode
[17:36:44] *** avinash has joined #illumos
[17:37:17] <enricop> Triskelios: yes, i have downloaded it from the prestable repository, but it was not in the already installed packages
[17:37:46] <Triskelios> enricop: strange. you did a pkg install rwn?
[17:38:07] <avinash> hi, i am interested in illumos project. i have downloaded a test tool from http://dlc.sun.com/osol/test/downloads/current/SUNWstc-stf-20090909.i386.tar.bz2
[17:38:08] <enricop> yes
[17:39:08] <avinash> i am having difficulty in using a test tool for solari
[17:39:28] <avinash> anybuddy help me on solari test tool
[17:39:30] *** Teknix has quit IRC
[17:39:36] <enricop> anyway, should this card be supported, right? it is the first time i try it on OI. if it is, i can check what's wrong by myself later
[17:40:45] <Triskelios> enricop: prtconf -dD shows no driver attached?
[17:40:53] *** Teknix has joined #illumos
[17:41:54] *** darjeeling has joined #illumos
[17:42:11] *** darjeeling has quit IRC
[17:42:37] *** darjeeling has joined #illumos
[17:43:49] <enricop> exactly, no driver is attached to that device
[17:44:54] <enricop> booting a live image of openindiana 151 it finds correctly the device type and bus
[17:45:04] <patdk-wk> ah, you need to update driver-aliases
[17:45:20] <patdk-wk> or maybe openindiana hasn't updated that file yet
[17:45:29] <Triskelios> enricop: update_drv -a -i '"pciexwhatever,whatever"' rwn
[17:45:47] <avinash> anybuddy help me on test framework for solaris. i have downloaded this http://dlc.sun.com/osol/test/downloads/current/SUNWstc-stf-20090909.i386.tar.bz2 will this do
[17:46:40] <Triskelios> avinash: http://hub.opensolaris.org/bin/view/Community+Group+testing/testsuites
[17:46:56] <Triskelios> the latest available source is presumably in the Mercurial repo
[17:48:12] <avinash> thanx sir, One more thing for doing project i have installed opensolaris on virtualBox. Is that fine ??
[17:49:12] <Triskelios> use an illumos distro
[17:49:34] <Triskelios> vbox will probably be too slow to do builds, though
[17:50:54] <avinash> ya but its running fine. at initial phase there should not be any problem. isn't it
[17:50:56] <enricop> update_drv did the trick. thanks
[17:51:02] <buffyg> Also vbox is unreliable with the debug kernel. I've updated the "Building from Source" page to reflect that, as that's been a frequent question.
[17:51:16] <buffyg> (lately)
[17:51:35] <e^ipi> s/with the debug kernel//
[17:51:39] <richlowe> I have never seen it be unreliably, just breathtakingly slow.
[17:51:43] <richlowe> it'll run forever.
[17:51:48] <richlowe> it will just also _Feel like_ it'll run forever.
[17:52:15] *** randw has joined #illumos
[17:52:23] <e^ipi> i've had guests seize up a bunch of times with it
[17:52:34] <Triskelios> e^ipi: just the person I was looking for... do you remember how you sorted out the hg-git bullshit with the malformed email address?
[17:52:47] <buffyg> I've seen system degrade into complete unresponsiveness with the debug kernel. With everything else, I run it just long enough to get a test done and then move along.
[17:53:06] <e^ipi> Triskelios: no. i ended up doing some juggling and one time it just worked
[17:53:56] *** |AbsyntH| has quit IRC
[17:54:01] <richlowe> Triskelios: give me 20 minutes, and I can help?
[17:55:10] <richlowe> Triskelios: the trick as I recall it, based partially on wondering why it worked for bdha and myself, is that git only seemed to bitch when it was the ancestor of a merge, and was also a commit actually in transit, so if you push that commit as a head, first, then push everything after it, it'd loosen up
[17:55:13] <enricop> i've successfully built illumos-gate following the updated documentation on the wiki.
[17:55:16] <enricop> i was wondering it illumos.sh supports an option to locate all output files in a different directory from the workspace.
[17:55:36] <enricop> like the -O option you can find in the linux kernel
[17:56:06] <richlowe> No, no support for obj dirs (outside of the proto area, which does work that way)
[17:56:24] <richlowe> everyone agrees it would be cool, but the makefile structure makes it a pain in the ass to get done.
[17:56:40] <Triskelios> I think enricop does just want to change the proto ("all output files")
[17:57:35] <Triskelios> enricop: planning to add the missing PCI IDs to rwn? (might be worth checking the public dbs for any others that might be missing)
[17:57:43] *** eschrock has quit IRC
[17:58:28] *** eschrock has joined #illumos
[17:58:30] <enricop> it is for two main reasons, but i don't know it makes sense: to leave the mercurial repository clean, and to understand better the build process
[17:59:13] <enricop> ^if it
[17:59:36] <avinash> Triskelios: the link u posted contains too many suit. I am working on "Automated test support" project. which one should i download
[17:59:53] <enricop> Triskelios: yes, i'll check the db and see what's is missing that i can update
[18:01:08] *** voidcoder has quit IRC
[18:01:15] *** voidcoder has joined #illumos
[18:08:29] *** s0liddi has quit IRC
[18:09:30] *** s0liddi has joined #illumos
[18:11:50] <avinash> Triskelios: Sir tell me which one should i download for Automates test project
[18:12:05] <avinash> I'll wait for your reply
[18:13:12] *** deirdres has quit IRC
[18:13:36] <Triskelios> avinash: as the page says, all of the tests use the same framework, so start with the framework?
[18:14:17] *** bradend has quit IRC
[18:17:24] *** maheshwar_ has quit IRC
[18:18:30] <avinash> Triskelios: sir i got this http://hub.opensolaris.org/bin/view/Community+Group+testing/stf can i start with this
[18:19:39] *** darjeeling has quit IRC
[18:20:27] <enricop> is it possible to find the 151a2 prestable iso/usb image somewhere? it would be very useful since i have a 82579LM nic...
[18:20:42] <e^ipi> avinash: look at what delphix has done with the test suite for zfs. after that the other suites will run just fine
[18:20:55] <e^ipi> https://github.com/delphix/zfstest
[18:21:05] <e^ipi> protip: it runs better in the bldenv(1) environment.
[18:21:15] <e^ipi> which i discovered by accident.
[18:22:30] <avinash> e^ipi: k thanx
[18:24:55] <avinash> e^ipi: i jsut wanna know i am doing all these will help me for project or these are not necessary
[18:25:26] <e^ipi> STF is a good place to start, i'd say
[18:25:29] <e^ipi> but it's your proposal
[18:25:36] <e^ipi> show creativity in planning
[18:28:38] *** ryancnelson has joined #illumos
[18:37:57] <enricop> this article has been published recently http://www.computer.org/portal/web/csdl/doi/10.1109/MSP.2012.30 (requires an ieee account)
[18:38:19] <enricop> for sure 'parfait' was of great help for the community back in osol days...
[18:43:35] <e^ipi> no it wasn't
[18:43:43] <e^ipi> it was a couple sun labs interns playing with it
[18:43:51] *** raichoo has quit IRC
[18:43:54] <e^ipi> nobody else ever got to look at it
[18:44:46] <enricop> wasn't is used for checking all the nightly builds?
[18:44:51] <e^ipi> nope
[18:45:11] <e^ipi> there were test suites, parfait wasn't one of them
[18:45:17] <e^ipi> it was STF and some others... VST was in there
[18:46:04] <e^ipi> it might be now, but it wasn't while the gate was open
[18:46:40] <enricop> ah ok, i see...
[18:47:07] <e^ipi> VST isn't open source, but nobody really cares about SUS compliance anyways
[18:47:16] <e^ipi> a couple of us have VST kicking around
[18:47:32] *** mamash has left #illumos
[18:47:46] *** mamash has joined #illumos
[18:48:39] <avinash> e^ipi: sir can u guide me about automated test project. I mean what path should i follow so as to have understanding of the project. I am very much interested but lacking in guidance
[18:49:44] <e^ipi> avinash: i think what whoever put that on there is looking for is some kind of automated test harness such that nightly builds are run, and then a battery of tests are run against it
[18:50:00] *** mjmac has joined #illumos
[18:50:09] <e^ipi> STF is a place to start, but it only tests some subset of functionality, and is annoying
[18:50:50] <e^ipi> I might do something like spawn a VM, onu ( update to the latest bits ), install the test framework, run all of them, and generate some kind of output
[18:51:10] <avinash> ok i'll do that
[18:51:24] <e^ipi> once again, it's *your* proposal. Those ideas are just starting points
[18:51:34] <avinash> ya ofcourse
[18:51:35] <e^ipi> points are awarded for understanding the problem in the first place
[18:51:45] <e^ipi> & creativity in coming up with solutions
[18:51:56] <avinash> thats why i need guidance
[18:52:15] <avinash> so i think i can start from here https://github.com/delphix/zfstest
[18:52:40] <e^ipi> i think you misunderstand what i mean when i say 'creativity'... it precludes 'guidance'
[18:52:57] <e^ipi> but yeah, zfstest is a good place to start
[18:53:11] <e^ipi> get it running first, then look at what else STF has to offer
[18:53:13] <avinash> ok thanx
[18:54:13] *** eschrock has quit IRC
[18:55:31] *** eschrock has joined #illumos
[18:57:26] <enricop> i would like to backup my rpool on a different hdd without using zfs send/receive. should i use zpool replace?
[18:58:20] <e^ipi> that sounds like an incredibly fragile way to do it
[18:59:36] *** mamash has left #illumos
[18:59:53] *** mamash has joined #illumos
[19:04:18] <enricop> ok i'll stick to zfs send -R, if there are not alternatives
[19:05:13] *** deirdres has joined #illumos
[19:07:00] <Triskelios> avinash: there's some build automation using Jenkins right now, so that would probably be a good approach for adding test automation
[19:08:11] *** raichoo has joined #illumos
[19:08:34] <avinash> Triskelios: Jenkins ??
[19:08:49] <raichoo> good eve
[19:10:50] <e^ipi> avinash: probably http://richlowe.openindiana.org/builds/
[19:11:30] <e^ipi> you can set jenkins up on your home machine too
[19:11:32] <e^ipi> it's trivial
[19:12:02] *** shank_ has joined #illumos
[19:13:01] *** mjmac has left #illumos
[19:13:17] *** rmustacc has joined #illumos
[19:14:50] *** Norrin has quit IRC
[19:16:07] *** eschrock has quit IRC
[19:16:23] <avinash> e^ipi: i got packages from that link http://richlowe.openindiana.org/pkgs/nightly-debug/en/catalog.shtml
[19:16:53] <e^ipi> those would be the nightly packages yeah
[19:17:24] <e^ipi> there's some way of wrangling onu(1) in to using them that i don't know offhand because i avoid pkg(5) like the plague it is
[19:17:46] <avinash> e^ipi: so i can install any of those packages
[19:17:48] *** randw_ has joined #illumos
[19:18:06] <e^ipi> avinash: no, use onu.
[19:18:18] <e^ipi> those packages are all of O/N ( illumos )
[19:18:25] <e^ipi> you'll want to update them atomically
[19:19:12] *** randw has quit IRC
[19:21:14] <avinash> e^ipi: onu means ??
[19:21:20] <e^ipi> O/N update
[19:21:28] <avinash> ok
[19:21:30] <e^ipi> or upgrade
[19:21:55] <e^ipi> it's a standard build tool in illumos, it installs the kernel, userland & libraries that you just built
[19:22:05] *** Norrin has joined #illumos
[19:22:47] *** randw has joined #illumos
[19:23:14] <e^ipi> illumos is not an installable thing ( like freebsd, or debian ), but it's not just a kernel ( like linux ) either
[19:23:29] <e^ipi> it's in between... the kernel<->library barriers aren't very well defined
[19:23:40] <e^ipi> so you need to update a big chunk of stuff, not just a kernel
[19:24:11] <avinash> e^ipi: hmm, so as of now i should install some onu packages
[19:24:16] <avinash> ??
[19:24:32] <e^ipi> i would suggest to try building illumos first, then updating to those built bits with onu(1)
[19:24:36] *** randw_ has quit IRC
[19:24:37] <e^ipi> it's a good thing to get used to doing
[19:25:31] <avinash> how would i know that among the list of packages listed here http://richlowe.openindiana.org/pkgs/nightly-debug/en/catalog.shtml which one is onu
[19:26:02] <e^ipi> start here: https://www.illumos.org/projects/5/wiki/How_To_Build_illumos
[19:26:04] <e^ipi> forget about those packages
[19:26:08] <buffyg> avinash: You really need to spend a bit of time with the build docs learning to self-serve on this stuff.
[19:26:08] <e^ipi> those are just nightly packages
[19:26:16] *** shank_ has quit IRC
[19:28:00] <avinash> buffyg: i love to do that. can u tell which is the good place to start
[19:28:13] <e^ipi> avinash: you've already been pointed at a starting place.
[19:29:17] <avinash> e^ipi: k i'll start from there only.
[19:29:26] *** deirdres has quit IRC
[19:29:40] *** deirdres has joined #illumos
[19:35:01] *** Akshat has joined #illumos
[19:35:58] <Akshat> hello
[19:36:29] <e^ipi> werd
[19:37:32] *** Akshat has quit IRC
[19:42:37] *** enricop has quit IRC
[19:46:31] *** _ams_ has quit IRC
[19:47:25] *** _ams_ has joined #illumos
[19:47:49] *** kart_ has quit IRC
[19:55:03] *** alanc has joined #illumos
[19:57:38] *** FACEFOX has joined #illumos
[20:02:39] *** krystal has quit IRC
[20:12:41] *** denizr has quit IRC
[20:15:41] *** eschrock has joined #illumos
[20:22:35] *** bradend has joined #illumos
[20:28:06] *** darjeeling has joined #illumos
[20:29:02] *** eschrock has quit IRC
[20:29:43] *** darjeeling has quit IRC
[20:30:52] *** eschrock has joined #illumos
[20:33:00] *** randw has quit IRC
[20:36:54] *** voidcoder has quit IRC
[20:41:48] *** voidcoder has joined #illumos
[20:42:08] *** krystal has joined #illumos
[20:44:19] *** denizr has joined #illumos
[20:46:21] *** denizr1 has joined #illumos
[20:47:20] *** marsell has joined #illumos
[20:50:15] *** denizr has quit IRC
[20:51:45] *** leowmjw has joined #illumos
[20:53:41] *** ripratm has joined #illumos
[20:53:59] *** ripratm has left #illumos
[20:57:06] *** avinash has quit IRC
[20:57:15] *** joffe has quit IRC
[21:00:06] <d_k> Hi. Has anyone worked with ZFS bookmarks (zbookmark_t) here? I need some help
[21:07:35] <richlowe> you'd have more luck (maybe) just asking the questions
[21:07:38] <richlowe> or emailing them to developer@
[21:09:00] <d_k> richlowe: thanks
[21:10:05] *** eclark has quit IRC
[21:10:40] *** eclark has joined #illumos
[21:11:04] *** eye-scuzzy has quit IRC
[21:11:21] *** chrisBF has joined #illumos
[21:16:56] *** enmand has quit IRC
[21:22:16] *** ekix has quit IRC
[21:29:00] *** pheleven has quit IRC
[21:30:47] *** [1]kdavyd has joined #illumos
[21:31:56] <RoyK> Tenzer: me? what? why? where?
[21:33:38] *** kdavyd has quit IRC
[21:33:39] *** [1]kdavyd is now known as kdavyd
[21:38:08] <Tenzer> RoyK: It was some guy who was wondering about what it would take for a spare drive to be used in his pool, since one of the drives was gone bad
[21:38:31] <Tenzer> I presumed you knew something about that, with all the storage experience you have got over time :)
[21:56:03] *** saablover has joined #illumos
[21:56:07] <saablover> hi all
[21:57:10] <RoyK> Tenzer: heh - I have no idea, really, but mostly a spare will jump in if a drive is taken out during normal run, but not if manually "offlined" or similar
[21:57:21] <RoyK> Tenzer: but don't ask me about whatever happens under the hood
[21:58:10] *** ssmoot has joined #illumos
[21:58:21] <RoyK> Tenzer: also, it seems zfs always *restarts* resilver when a drive dies, meaning if resilver was 95% finished when drive 1 died and drive 2 dies, it'll restart the whole thing to resilver both at once, which isn't really a good idea imho
[21:58:38] *** ssmoot has left #illumos
[22:00:41] <saablover> I'm using brendangs example with the grep bug to demonstrate dtrace, http://dtrace.org/blogs/brendan/2011/12/08/2000x-performance-win/, now the dtrace output is correct while the grep is running and I break the trace, when the find finishes and I use -o /tmp/output and look inside the outputfile, the output is mangled
[22:00:56] <Tenzer> saablover: Hello
[22:01:16] <saablover> thx brendan by the way !
[22:01:21] <saablover> hi Tenzer
[22:01:34] <Tenzer> RoyK: Hmm, okay. Fortunately I don't have such big storage systems to manage, where there is a big risk of multiple drives failing at the same time
[22:01:49] *** eye-scuzzy has joined #illumos
[22:02:52] *** joshie has quit IRC
[22:06:34] *** Botanic has joined #illumos
[22:06:34] *** Botanic has quit IRC
[22:06:34] *** Botanic has joined #illumos
[22:07:10] *** esproul has quit IRC
[22:18:16] <vsrinivas> brendangs's blog on dtrace.org is fascinating. :) , particularly recent note on flame graphs.
[22:18:51] <vsrinivas> going to see if we can convince our kernel (BSD) to be profiled, so we can grab stacks for it
[22:19:28] <rmustacc> If you enable DTrace support there, you should be able to do the same thing.
[22:20:08] <e^ipi> what needs convincing, i thought freebsd supported dtrace out of the box
[22:20:52] <vsrinivas> FreeBSD does. DragonFly doesn't :P
[22:21:35] <vsrinivas> and the FlameGraph pl tool doesn't actually require dtrace-specific output, just stack lists and # of samples.
[22:23:05] <vsrinivas> do you folks know if there is any tech note or paper describing dtrace's implementation? i've seen the usenix '04 one, but it wasn't quite what i was looking for...
[22:23:59] <vsrinivas> was interested in how trace data was stored, how cross-CPU logs were correlated (events are time indexed presumably? what time source?)
[22:24:29] <rmustacc> The best source is dtrace_impl.h.
[22:25:01] <e^ipi> vsrinivas: http://src.illumos.org/source/xref/freebsd-head/cddl/lib/libdtrace/ ?
[22:25:39] <rmustacc> vsrinivas: Right now events actually aren't time indexed.
[22:25:46] <rmustacc> At least, not across multiple CPUs.
[22:25:54] <rmustacc> Though there is some work that folks started to do that.
[22:26:33] <vsrinivas> rmustacc: oh? interesting.... so how do you combine lists of events from different CPUs?
[22:26:54] <vsrinivas> e^ipi: yep! FreeBSD has done a great job importing dtrace, but i work on another BSD (DragonFly BSD)...
[22:27:04] <e^ipi> vsrinivas: right, i got that
[22:27:09] <rmustacc> I don't really remember the specifics, I'd look at the implementation.
[22:27:19] <rmustacc> But basically you're pulling data out of per-CPU buffers.
[22:27:22] <vsrinivas> e^ipi: oh, then what should i look at in libdtrace?
[22:27:25] <rmustacc> So, you probably consume them.,
[22:27:41] <rmustacc> e.g. you order it based on how you consume those per-cpu buffers.
[22:27:42] <e^ipi> vsrinivas: i misunderstood; i thought you were looking to pull freebsd's dtrace impl in to dfly
[22:27:50] <e^ipi> which would be awesome, by the way
[22:27:58] <vsrinivas> oh, no. i was just curious how it worked. :
[22:28:06] <vsrinivas> (would be great to have more ports, but...)
[22:28:58] *** enmand has joined #illumos
[22:33:52] *** eschrock has quit IRC
[22:52:24] *** ex9t_ has quit IRC
[22:54:35] *** ekix has joined #illumos
[22:56:45] *** sross has quit IRC
[22:57:56] *** voidcoder has quit IRC
[22:58:16] *** voidcoder has joined #illumos
[22:59:50] *** merzo has joined #illumos
[23:05:24] *** s0liddi has quit IRC
[23:09:31] *** eschrock has joined #illumos
[23:09:52] *** s0liddi has joined #illumos
[23:10:32] *** Ali_G_Indahouse has joined #illumos
[23:10:32] *** eschrock has quit IRC
[23:10:59] <Ali_G_Indahouse> hello
[23:11:05] <Ali_G_Indahouse> i want to talk anout gsoc
[23:11:43] *** saablover has left #illumos
[23:12:49] <e^ipi> what of it?
[23:13:55] <Ali_G_Indahouse> about implamenting the ND RDNSS support
[23:14:20] <Ali_G_Indahouse> it is intended to build a server and client
[23:14:41] <e^ipi> okay
[23:15:02] <Ali_G_Indahouse> and if some clients want to share some information, then the nearest clients share that info?
[23:15:31] <e^ipi> i guess ? It's all in the RFC.
[23:15:47] <Ali_G_Indahouse> what is rfc?
[23:15:53] <e^ipi> request for comments
[23:16:06] <e^ipi> IETF standards, essentially
[23:17:48] <buffyg> I'm looking at the title of RFC 5006 "IPv6 Router Advertisement Option for DNS Configuration".
[23:19:14] *** neophenix has quit IRC
[23:19:22] *** FACEFOX has quit IRC
[23:19:33] *** raichoo has quit IRC
[23:19:43] <e^ipi> they've even got a handy RFC search engine at ietf.org
[23:21:19] *** timclassic has quit IRC
[23:21:25] <Ali_G_Indahouse> so what i can find in the rfc?
[23:21:31] <Ali_G_Indahouse> what is the purpose of the rfc?
[23:21:38] <e^ipi> everything required to implement the protocol
[23:21:45] <alanc> to document the requirements and specification
[23:22:06] <Ali_G_Indahouse> where is that?
[23:22:11] <Ali_G_Indahouse> just in some iste
[23:22:12] <Ali_G_Indahouse> site
[23:22:16] <Ali_G_Indahouse> or various?
[23:22:31] <alanc> put "RFC 5006" into google and it will give it to you right away
[23:22:54] <buffyg> Ali_G_Indahouse: Are you serious, or are you Ali Ging us?
[23:25:23] <Ali_G_Indahouse> ?
[23:25:34] *** mamash has left #illumos
[23:25:41] *** FACEFOX has joined #illumos
[23:26:23] <buffyg> Ali_G_Indahouse: You're asking questions that are basic knowledge for being able to apply to this kind of project.
[23:27:08] <Ali_G_Indahouse> well, at school i have wrote some code to pass messages trought server and client
[23:27:15] <buffyg> If you're interested, you need to undertake sufficient research to answer these questions for yourself.
[23:27:16] <Ali_G_Indahouse> various clients
[23:27:20] <Ali_G_Indahouse> with threads
[23:27:33] <buffyg> If you don't know what an RFC is, Google is your friend.
[23:27:38] *** FACEFOX has quit IRC
[23:27:45] <buffyg> If you don't know what the RFC says, find it with google.
[23:27:52] <Ali_G_Indahouse> i had also this course:Computer Networks and the Internet
[23:28:02] <Ali_G_Indahouse> but never heards this terms
[23:28:11] <Ali_G_Indahouse> this terms are very specific
[23:28:51] <Ali_G_Indahouse> i dont know why meny of you expect here to come student with very profound knowledge about everything
[23:29:00] <Triskelios> if you've ever learned about internet protocols you would know this
[23:29:09] <Ali_G_Indahouse> i will learn it as i will do this project
[23:29:10] <Ali_G_Indahouse> ....
[23:29:14] <Triskelios> this is basic networking knowledge
[23:29:15] <e^ipi> because they're connected to the world's largest repository of all of human knowledge, and looking this up takes under 100ms ?
[23:29:38] *** himsin has quit IRC
[23:29:58] <buffyg> If you don't know what an RFC is, you'll get the Wikipedia article with the answer as your first hit on google.
[23:30:27] <buffyg> You are digging yourself a hole if you're asking these questions here rather than getting these answers for yourself.
[23:31:03] <Alasdairrr> Ali_G_Indahouse: The short version is that to be able to take part in GSOC it's expected that you're already a nerd who spends most of his evenings hacking on his computer, and most nerds who do that will know what an RFC is and how to google answers to questions to avoid looking stupid on IRC
[23:31:41] <Alasdairrr> GSOC isn't a training course for inexperienced people coming into this with little prior knowledge, I'm afraid
[23:31:57] <Alasdairrr> Operating systems are very complex and there isn't enough time to train people from first principles
[23:32:37] <Ali_G_Indahouse> nobody knows every detail
[23:32:56] <e^ipi> Ali_G_Indahouse: no, but nobody minds explaining how our network stack works, but you need to understand what a network stack *is* first
[23:34:16] <e^ipi> but things like knowing what an RFC is is such a basic piece of knowledge for anyone that wants to do anything related to networks that it's expected that you know this, because your peers who you are competing with do. Not just for SoC, but for employment in the computing field in general
[23:35:05] <e^ipi> GSoC isn't free money, students are selected for their nascent expertise, like any other job
[23:35:27] <e^ipi> mentors are aware of the limitations of students, as most of us have been CS undergrads
[23:37:03] *** elijah-mbp has joined #illumos
[23:38:08] *** denizr1 has quit IRC
[23:39:21] <Triskelios> Alasdairrr: not even a matter of "training", just a willingness to do basic research
[23:39:58] *** raichoo has joined #illumos
[23:40:02] <e^ipi> "what do i have to do in this class to get a C-?"
[23:40:14] <CIA-56> illumos Vitaliy Gusev <gusev.vitaliy at nexenta dot com>: 1844 xen: "disk_link: invalid disk device number" Reviewed by: Dmitry.Yusupov at nexenta dot com Reviewed by: Albert Lee <trisk at nexenta dot com> Reviewed by: Garrett D'Amore <garrett at damore dot org> Approved by: Garrett D'Amore <garrett at damore dot org>
[23:40:45] *** ex9t has joined #illumos
[23:40:47] <buffyg> Ali_G_Indahouse: This isn't an argument. It's a basic expectation, which e^ipi is telling you're just as likely to encounter in the job market or any other GSoC project. Either you're willing to meet it, or you should look for somewhere with different expectations. e^ipi is rightly telling you that's not very likely in the technical side of IT.
[23:44:03] *** Tempt has quit IRC
[23:45:18] *** Tempt has joined #illumos
[23:45:18] *** ChanServ sets mode: +o Tempt
[23:51:42] *** FACEFOX has joined #illumos
[23:51:43] *** denizr has joined #illumos
[23:52:07] *** denizr has quit IRC
[23:58:06] <Ali_G_Indahouse> Request for Comments (RFC) is a memorandum ...
[23:58:19] <Ali_G_Indahouse> at my course i dont learn who is who
[23:59:18] <Ali_G_Indahouse> but yes, i am recognizing that that is a basic thing that i need to know
[23:59:24] <Ali_G_Indahouse> or needed to know
[23:59:35] <buffyg> Ali_G_Indahouse: This is not an argument. If you're still interested, please do the research (learn about the IETF process, try to read the RFCs) and come back with informed questioons.
top

   March 20, 2012  
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >