Switch to DuckDuckGo Search
   February 11, 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 | >

Toggle Join/Part | bottom
[00:00:09] *** topping has joined #openindiana
[00:01:58] *** Kurlon has joined #openindiana
[00:06:13] *** timclassic2 has quit IRC
[00:22:04] *** scarcry has quit IRC
[00:41:20] <viridari> RoyK: if you're being serious, I think the burden is on Nexenta to reach out to the community and explain.
[00:56:15] <cypromis> two many problems with WD drives
[00:56:31] <cypromis> But I do not see why Nexenta would have to explain anything
[00:56:33] <cypromis> :)
[00:56:50] <RoyK> viridari: I'm perfectly serious
[00:58:56] <cypromis> Did we ever supported WD Drives ?
[00:59:41] * cypromis checks the HSL and does not see any WD Drives on there
[00:59:57] * tomww can't see what should be specific to a WD drive
[01:00:21]
[01:00:35]
[01:01:03] <cypromis> yah and we are not really fond of SATA
[01:07:59] <RoyK> aren't we?
[01:09:36] <Triskelios> we have? --> <RoyK> altker128: but, incidentially, nexenta has dropped ALL support for WD drives
[01:10:07] * RoyK is confused
[01:10:09] <richlowe> That would be a bloody moronic thing to do.
[01:10:14] <richlowe> not that that affects the chances much.
[01:10:28] <RoyK> moronic? by nexenta?
[01:10:30] <alanc> never added isn't quite the same as dropped
[01:10:33] <kdavyd> Dropped support != never had support.
[01:10:50] <RoyK> nexenta had wd support, but apparently, that was dropped
[01:12:11] <cypromis> RoyK: you seem to know more about this than nexenta
[01:12:12] <cypromis> .9
[01:12:14] <cypromis> :)
[01:12:22] <richlowe> which sure does suggest "wrong".
[01:12:40] <RoyK> cypromis: heh - I was just told this by my supermicro vendor, which is also nexenta certified
[01:12:58] <cypromis> possible
[01:13:01] <cypromis> I am FE at nexenta
[01:13:09] <cypromis> kdavyd is support
[01:13:13] <cypromis> trisk is dev
[01:13:24] <cypromis> of course supermicro might know more about this than we do ..
[01:14:28] <cypromis> and if nexenta did support wd drives it must have been before any of us joined
[01:15:01] <RoyK> they're not supermicro, but they sell that and nexenta....
[01:15:05] <cypromis> maybe the support was not as in 'supported DeviceÄ but as in 'Device provided by hardware vendor of system so not really a way to talk yourself out of it'
[01:15:08] <cypromis> :D
[01:15:16] <RoyK> and, of course, I might have misunderstood what he told me
[01:15:26] <RoyK> once every four years or so, I make a mistake
[01:15:34] * RoyK runs
[01:15:34] <cypromis> hehehe
[01:15:41] <kdavyd> Note - I was a Nexenta customer for 1.5 years, and looked at the product since early 2009. WD drives were never on the HSL; in fact I used WD drives on my first POC, which was officially discouraged because of known issues with them.
[01:16:03] <RoyK> ah
[01:16:11] <RoyK> probably my leap year, then
[01:16:25] <kdavyd> Sure enough, I quickly learned *why* they were discouraged. Those were WD Enterprise 10k Raptors (although sata)
[01:16:39] <RoyK> what happened?
[01:16:43] <kdavyd> Ended up replacing every single one of them with Savvios, never had a problem since
[01:17:12] <RoyK> savvios?
[01:17:38] <kdavyd> Once every couple weeks, one would drop out of the pool - increased latency, eventually marked degraded. Once cleared and readded to pool, would work just fine until the next event.
[01:17:47]
[01:17:56] <RoyK> k
[01:18:08]
[01:18:43] <RoyK> hrmf
[01:18:49] <RoyK> this resilvering takes forever
[01:19:09] <RoyK> should I detach the spare (second dead drive in same vdev is resilvered to the spare)
[01:19:13] <RoyK> ?
[01:19:25] <RoyK> it seems it didn't take that long to resilver a single device
[01:20:04] <RoyK> but that was interrupted by the second disk dying and zpool using its magick to restart the resilver just when it was about to finish
[01:21:27] *** master_o1_master has quit IRC
[01:23:21] *** master_of_master has joined #openindiana
[01:24:38] *** jwk404 has quit IRC
[01:24:57] <tomww> i
[01:25:05] <tomww> <ESC>
[01:26:38] <Dr_Feelgood> any way to currently get multimedia codecs on oi151 ?
[01:26:47] <Dr_Feelgood> ...other than flash
[01:29:08] <Triskelios> they're in SFE
[01:29:21] <Dr_Feelgood> Triskelios SFE?
[01:29:43] <Triskelios> smrt: explain spec-files-extra
[01:29:44] <smrt> ⌂ A collection of popular software which is not in OpenSolaris. The SFE project ports software using RPM-style spec files. See also: bootstrap-sfe-latest-, pkgbuild. http://wong.to/sfewiki for more info, including installation instructions.
[01:29:53] <Triskelios> there's a link to the binary repo in the release notes
[01:30:08] <Dr_Feelgood> ty Triskelios !
[01:31:02] *** enmand has quit IRC
[01:36:38] *** hajma has quit IRC
[01:49:10] *** nvez has quit IRC
[01:49:50] *** raichoo has quit IRC
[01:56:48] *** nvez_ has joined #openindiana
[01:57:19] *** ctable has quit IRC
[02:02:34] <topping> hi all, is the desktop version of OI basically the server version plus gnome et al?
[02:03:42] <topping> can i upgrade the server to a desktop with some fancy keyboard tricks?
[02:04:14] <tsoome> pkg install slim_install
[02:05:03] <topping> is there also a "fat_install"? :)
[02:05:30] <topping> guess not
[02:07:47] <topping> ah http://hub.opensolaris.org/bin/view/Project+caiman/Slim_Install
[02:08:19] <topping> just figured it was a euphemism token, not the name of a project
[02:08:51] <topping> the last time i did much with sun stuff was sunos 4.x in college
[02:08:58] <alanc> it was the project to get the install onto a livecd, down from dvd sized
[02:09:51] <topping> thanks, i guess i'd be open to installing a heavyweight desktop if there's a group name for one
[02:11:19] <alanc> slim install brings in a fair chunk of gnome, though not some of the bigger bits like gimp if I remember correctly
[02:11:40] <topping> i just found a reference to something called a babel_install
[02:11:47] <topping> sounds like everything! :-)
[02:11:57] <alanc> that was slim_install plus all the localizations to other languages
[02:12:02] <topping> ah
[02:12:25] <topping> can do without languages
[02:12:36] *** nvez_ has quit IRC
[02:12:53] <alanc> I don't think there's a "everything desktop" group - there's "redistributable" which is more like an "almost the entire repo" group that brings in all the databases, web servers, etc. too
[02:14:15] <topping> ok thx. i'll stick with slim_install
[02:14:33] <topping> sorry for stupid n00b questions
[02:20:20] *** vados1 has joined #openindiana
[02:20:56] *** vados1 has left #openindiana
[02:21:45] *** vados has quit IRC
[02:32:34] *** vados1 has joined #openindiana
[02:33:06] *** vados1 is now known as vados
[02:42:08] *** Dr_Feelgood has quit IRC
[02:48:36] *** voidcoder has joined #openindiana
[02:51:32] *** H4X0R-1337 has joined #openindiana
[03:01:22] *** TomJ has joined #openindiana
[03:04:47] *** randw has quit IRC
[03:06:41] *** Shep_Book has quit IRC
[03:10:55] *** patdk-lap has joined #openindiana
[03:11:26] *** PoolShark_ has joined #openindiana
[03:17:56] *** PoolShark_ has quit IRC
[03:20:07] *** Botanic has quit IRC
[03:24:48] *** hjf has quit IRC
[03:40:05] *** Licenser_ has joined #openindiana
[03:41:31] *** Botanic has joined #openindiana
[03:41:31] *** Botanic has joined #openindiana
[03:41:33] *** Licenser has quit IRC
[03:41:33] *** Licenser_ is now known as Licenser
[03:42:10] *** axisys has quit IRC
[03:49:30] *** gvm1 has quit IRC
[04:15:45] *** timclassic has joined #openindiana
[04:17:21] *** kart_ has joined #openindiana
[04:35:57] *** gvm1 has joined #openindiana
[04:43:18] *** gwr_ has quit IRC
[04:47:20] *** nvez has joined #openindiana
[04:47:20] *** nvez has joined #openindiana
[04:49:33] *** enmand has joined #openindiana
[04:56:22] *** enmand has quit IRC
[05:15:39] *** kisisten is now known as magyar
[05:15:55] *** magyar has joined #openindiana
[05:19:28] *** Seony has quit IRC
[05:43:20] *** Seony has joined #openindiana
[06:10:39] *** axisys has joined #openindiana
[06:40:10] *** nvez has quit IRC
[06:50:31] *** alanc has quit IRC
[07:03:27] *** bradend has quit IRC
[07:03:43] *** bradend has joined #openindiana
[07:04:56] *** TomJ has quit IRC
[07:05:41] *** alanc has joined #openindiana
[07:05:41] *** ChanServ sets mode: +o alanc
[07:07:57] *** n2deep has quit IRC
[07:10:42] *** TomJ has joined #openindiana
[07:11:35] *** TomJ has quit IRC
[07:24:47] *** b-tag has joined #openindiana
[07:25:56] *** b-tag has left #openindiana
[07:27:50] *** russiane39 has quit IRC
[08:27:25] *** voidcoder has quit IRC
[08:34:40] *** n2deep has joined #openindiana
[08:36:45] *** voidcoder has joined #openindiana
[08:42:33] *** Tenzer has quit IRC
[08:46:06] *** Botanic has quit IRC
[08:47:19] *** Botanic has joined #openindiana
[08:47:19] *** Botanic has joined #openindiana
[08:55:03] *** [1]kdavyd has joined #openindiana
[08:55:30] *** jswlinux has joined #openindiana
[08:55:32] *** Seony has quit IRC
[08:55:40] *** jswlinux is now known as Seony
[08:56:09] *** H4X0R-1337 has quit IRC
[08:57:08] *** kdavyd has quit IRC
[08:57:08] *** [1]kdavyd is now known as kdavyd
[08:58:31] *** vados has left #openindiana
[08:59:08] *** Licenser has quit IRC
[08:59:48] *** Licenser has joined #openindiana
[09:01:01] *** catmistake has quit IRC
[09:09:15] *** H4X0R-1337 has joined #openindiana
[09:20:04] *** deraps has quit IRC
[09:21:15] *** russiane39 has joined #openindiana
[09:21:15] *** InTheWings has joined #openindiana
[09:21:15] *** deraps_ has joined #openindiana
[09:21:15] *** |woody_| has joined #openindiana
[09:21:25] *** PiotrSikora has quit IRC
[09:21:25] *** |woody| has quit IRC
[09:22:01] *** Tenzer has joined #openindiana
[09:22:24] *** movement has quit IRC
[09:22:40] *** PiotrSikora has joined #openindiana
[09:22:54] *** movement has joined #openindiana
[09:27:28] *** vados has joined #openindiana
[09:28:15] *** n2deep has quit IRC
[09:28:15] *** mixis_ has quit IRC
[09:28:28] *** n2deep has joined #openindiana
[09:33:22] *** mixis has joined #openindiana
[09:37:12] *** topping1 has joined #openindiana
[09:37:19] *** kart_ has quit IRC
[09:37:53] *** idnaria has joined #openindiana
[09:38:31] *** |woody| has joined #openindiana
[09:39:33] *** topping has quit IRC
[09:40:01] *** master_o1_master has joined #openindiana
[09:40:49] *** master_of_master has quit IRC
[09:40:49] *** idnar has quit IRC
[09:40:49] *** |woody_| has quit IRC
[09:40:49] *** reddi has quit IRC
[09:40:49] *** tomocha66 has quit IRC
[09:40:59] *** tomocha66 has joined #openindiana
[09:41:01] *** reddi has joined #openindiana
[09:43:28] *** InTheWings has quit IRC
[09:49:29] *** scarcry has joined #openindiana
[09:51:19] *** DwD__ has joined #openindiana
[10:04:39] *** vados has left #openindiana
[10:16:29] *** girafe has joined #openindiana
[10:21:06] *** sivanov_ has joined #openindiana
[10:21:19] <sivanov_> hi
[10:21:32] <sivanov_> after booting into single user mode, and importing-exporting data pools, fmd seems to got broken
[10:21:51] <sivanov_> as i had one faulty dimm i;ve replaced and was going to run fmadm repair
[10:23:08] <sivanov_> at first i had to run fmadm faulty to identify the module
[10:23:13] <sivanov_> then i got a message that fmd faulted and i need to remove /var/fm/fmd/ckpt/eft/eft and restart the service
[10:23:24] <sivanov_> which i just did
[10:23:39] <sivanov_> but now fmd uses about 9.5% cpu tim
[10:23:51] <sivanov_> how to fix it?
[10:28:55] * RoyK has no idea
[10:34:38] <sivanov_> hmm enabling/disabling fmd again made it worse now fmd is 25% (one core) usage
[10:35:17] <richlowe> what do fmdump and fmdump -e say?
[10:35:28] <richlowe> normally fmd becomes busy because it's being pounded with error reports.
[10:35:59] * crsd eyes richlowe
[10:36:22] <richlowe> and yes, crsd, I'm back for a bit.
[10:36:24] <richlowe> what's up?
[10:36:51] <sivanov_> it says ereport.cpu.intel.nb.mem_ce
[10:37:23] <sivanov_> but fmadm falty does not say enything
[10:37:59] <richlowe> depending on when the timestamps are, that's probably just because it's not seen enough of them yet.
[10:39:37] <myrkraverk> Among other things Oracle has shut down is the Open Web Server page. Is that project available somewhere?
[10:40:29] <richlowe> don't remember what that is/was
[10:41:08] *** Botanic has quit IRC
[10:42:01] <sivanov_> how i zap whole disk? dd if=/dev/zero of=/dev/rdsk/c0t1d0p0 ?
[10:49:17] <Tenzer> myrkraverk: Was that an open source version of their Sun Java System Web Server? (Longest name, ever!)
[10:49:57] <myrkraverk> Tenzer, Yes. http://www.c0t0d0s0.org/archives/5203-Open-Web-Server.html
[10:50:55] <tsoome> sivanov_: think that should do it, yes
[10:51:16] <richlowe> wow, on a combination of opensolaris.org _and_ wikis
[10:51:19] <richlowe> so screwed both ways.
[10:51:21] <richlowe> Not sure how you'd find it.
[10:51:26] <Tenzer> sivanov_: You should probably remove "p0" from that, otherwise you are only targeting the first partition on the drive
[10:57:21] <myrkraverk> http://web.archive.org/web/20090221144140/http://wikis.sun.com/display/wsFOSS/Build+Instructions
[10:57:26] <myrkraverk> Checkout instructions work.
[10:57:37] <myrkraverk> At least, something is happening on the hg command.
[10:59:05] <myrkraverk> That means it is possible to resurrect the project - if anyone has interest in that.
[10:59:08] <altker128> Anyone here who uses Macs mess with 'BlueHarvest', the program to quell OSX resource fork data?
[10:59:17] <Tenzer> myrkraverk: Have you seen comment #3 on that c0t0d0s0 blog post? It seems like all the features which made SJSWS somewhat attractive, have been removed from the open source version
[10:59:23] <altker128> I'm curious how well that works with AFP/NFS.
[10:59:44] <myrkraverk> Tenzer, No, I didn't bother to read anything until I knew I could download it somewhere.
[11:00:15] <Tenzer> myrkraverk: After you have read that list, you might not even want to bother downloading it ;)
[11:00:50] <myrkraverk> Tenzer, Well, better to have downloaded it before hg.opensolaris.org disappears.
[11:01:37] <Tenzer> TBH I think resources are better spend on using Glass Fish, Tomcat, nginx, Apache or another more widely used web server - depending on your needs of course
[11:04:24] <myrkraverk> Tenzer, Well, here I am more interested in an archive of the project - for historical reasons more than anything else. I hate when projects just disappear. It's extremely common that commercial software is "lost in space" with source code and everything. With open source it's painful when it happens.
[11:04:26] *** sivanov_ has quit IRC
[11:04:40] <myrkraverk> http://web.archive.org/web/20090227133945/http://wikis.sun.com/display/wsFOSS/Features+Comparison
[11:05:18] <Tenzer> Okay, I just thought run it, without a clear overview of what it was able to do :)
[11:06:57] <myrkraverk> Well, I'd like to test drive it locally. Not sure if I want to use it in production since I have only myself for support :P
[11:10:41] *** hajma has joined #openindiana
[11:15:52] <myrkraverk> If I'm feeling generous someday I'll try to create an archive website somewhere.
[11:17:35] <Tenzer> Perhaps http://wesunsolve.net/ would like to host the project? I know that they have gathered a lot of old archives and such from Sun
[11:27:24] *** TomJ has joined #openindiana
[11:50:44] *** raichoo has joined #openindiana
[11:55:29] *** voidcoder has quit IRC
[12:10:13] *** sivanov_ has joined #openindiana
[12:12:33] <sivanov_> what is the best way to format an SSD for use both as L2ARC and ZIL? will the solaris slices be aligned?
[12:21:25] <russiane39> Triskelios, you around ?
[12:21:30] <russiane39> I've tried your fix
[12:21:34] <russiane39> with no success
[12:23:37] <russiane39> seems we need ISO with latest CPQary3 also
[12:25:02] *** [gorgo] has left #openindiana
[12:26:56] *** init3 has quit IRC
[12:29:02] *** init3 has joined #openindiana
[12:56:13] *** raichoo has quit IRC
[13:43:35] *** lennard has quit IRC
[13:45:42] *** lennard has joined #openindiana
[14:02:24] *** tsukasa` is now known as tsukasa
[14:04:32] *** SH0x has quit IRC
[14:11:24] <tsoome> sivanov_: VTOC slices are aligned at cylinder boundaries
[14:23:08] *** enmand has joined #openindiana
[14:24:34] *** enmand has quit IRC
[14:25:48] <sivanov_> omg, ebay has lto4 libraries for about $1000
[14:29:25] *** Seony has quit IRC
[14:33:43] *** H4X0R-1337 has quit IRC
[14:34:05] *** Monte has joined #openindiana
[14:34:05] *** Monte has joined #openindiana
[14:37:49] *** xenol has quit IRC
[14:49:26] *** jpg has joined #openindiana
[14:56:05] *** _404_Error_ has quit IRC
[15:08:30] *** _404_Error_ has joined #openindiana
[15:21:44] *** raichoo has joined #openindiana
[15:32:59] *** finnland_ has joined #openindiana
[15:33:05] <finnland_> hello
[15:33:40] <finnland_> is it possible to make usb available for a zone?
[15:40:08] <Tenzer> finnland_: Sure, you just need to add the USB device's path in the zones configuration
[15:41:11] <finnland_> Tenzer: hmm, ok.
[15:41:26] *** xenol has joined #openindiana
[15:41:33] <finnland_> Tenzer: Im trying to give Virtualbox inside a zone acces to usb
[15:42:26] <Tenzer> That meaning, if you want to give a zone access to a device connected to the machine, you would enter the following when running the "zonecfg" command: add device; set match=/dev/<device>; end
[15:42:41] <Tenzer> I don't think you can give access to USB, just in general
[15:42:55] <finnland_> yes, ok. no hotplugging then
[15:43:06] <Tenzer> You will have to find the device name which the USB device presents itself as, and that depends on what type of USB device you are working with
[15:45:08] <finnland_> thanks, no touching of USB devices on the server then. :)
[15:53:20] *** DwD__ has quit IRC
[15:54:34] *** laogrymt has joined #openindiana
[16:31:04] *** enmand has joined #openindiana
[16:32:40] *** jw_urodoc has joined #openindiana
[16:40:40] *** LizardKing has quit IRC
[16:41:25] *** nvez has joined #openindiana
[16:41:26] *** nvez has joined #openindiana
[16:55:51] *** LizardKing has joined #openindiana
[16:59:30] *** sivanov_ has quit IRC
[17:04:07] *** ChanServ has quit IRC
[17:07:48] *** ChanServ has joined #openindiana
[17:07:48] *** gibson.freenode.net sets mode: +o ChanServ
[17:10:01] *** michal_ has left #openindiana
[17:16:39] *** troydm has joined #openindiana
[17:17:28] *** troydm has quit IRC
[17:19:35] *** raichoo has quit IRC
[17:45:59] *** troydm has joined #openindiana
[17:46:13] <troydm> hello ppl
[17:46:51] <troydm> so i have a problem with my old opensolaris home server (grub failed) so i was wondering if i could install OpenIndiana 151 on it
[17:47:03] <troydm> so my conserns are about stability
[17:47:08] <troydm> how stable is 151a
[17:47:21] <troydm> and if it supports VirtualBox 4.1.8
[17:47:34] <troydm> those two are my main conserns
[17:53:15] <Tenzer> troydm: Stability wise I wouldn't say OI is anymore unstable than OpenSolaris. Both of them are based on the same source, which has been running Solaris in enterprise environments for many years
[17:54:00] <troydm> I've used last build of OpenSolaris and it worked for me until power failure occured today
[17:54:09] <troydm> now grub load it says
[17:54:14] <troydm> Loading Stage2...
[17:54:17] <Tenzer> I haven't been running VirtualBox on OpenIndiana for a long while, so I can't assure you that it will work, but I think it will
[17:54:19] <troydm> zio_read_failed
[18:08:28] *** TomJ has quit IRC
[18:15:48] *** jw_urodoc has quit IRC
[18:19:58] *** patdk-lap has quit IRC
[18:23:57] *** enmand has quit IRC
[18:27:56] *** easy-iPad has joined #openindiana
[18:32:46] *** laogrymt has quit IRC
[18:43:31] *** easy-iPad has quit IRC
[18:43:35] *** bdrewery_ is now known as bdrewery
[18:49:01] *** easy-iPad has joined #openindiana
[18:53:14] *** hajma has quit IRC
[18:53:37] *** hajma has joined #openindiana
[18:56:41] *** hajma has quit IRC
[18:56:54] *** topping1 has quit IRC
[18:57:10] *** hajma has joined #openindiana
[19:12:07] *** richlowe has quit IRC
[19:13:10] *** flyz_ has quit IRC
[19:13:25] *** richlowe has joined #openindiana
[19:20:30] *** flyz has joined #openindiana
[19:24:06] *** Botanic has joined #openindiana
[19:24:07] *** Botanic has joined #openindiana
[19:27:03] *** easy-iPad has quit IRC
[19:27:31] *** ozquera has joined #openindiana
[19:50:57] *** Shep_Book has joined #openindiana
[19:53:55] *** wcj has joined #openindiana
[19:55:48] *** cis has quit IRC
[20:00:08] *** Shep_Book has quit IRC
[20:02:43] *** haj has joined #openindiana
[20:02:59] <haj> hey
[20:05:44] <haj> I'm having this weird nfsd problem... I've seen it on a couple of nexenta-systems we have... and now I'm setting up a couple of new boxes with SAS Expanders and I'm seeing it again.... nfs performance drops to 0, no activity for a few sec, mostly during high load, but it happens during normal load as well.. with around 30Mbit traffic.
[20:05:49] <haj> any hints on this?
[20:06:20] <haj> could it be the number of nfs threads perhaps?
[20:07:30] <troydm> haj: just curios what version OpenIndiana are you running?
[20:15:50] <mui> how does disks perform?
[20:15:56] <mui> you can watch iostat -xen 1
[20:16:19] <mui> but yes, having too few nfs servers can cause slowness
[20:17:20] *** InTheWings has joined #openindiana
[20:17:27] *** InTheWings has joined #openindiana
[20:18:23] <haj> troydm: 151a
[20:18:44] <viridari> haj: are you using dedup by chance?
[20:22:06] <haj> mui: I get busy rates in the 50's .. they drop to zero when NFS hangs.. thats why I think it looks like an issue with the daemon
[20:22:11] <haj> viridari: nope
[20:22:11] <troydm> haj: this might be helpful http://www.mobygeek.net/blog/2011/08/08/host-refreshed-with-openindiana-151/
[20:22:36] <mui> haj: make sure you've increased amount of nfsd servers
[20:23:03] *** troydm has quit IRC
[20:23:20] <haj> mui: I didn't.. I'll check right away :)
[20:24:24] <haj> mui: could it be dangerous setting it too high?
[20:24:35] <haj> like.. 128 ?
[20:24:40] <mui> well on modern systems, no
[20:24:44] <mui> I use like 1024 or 2048
[20:27:36] <haj> ah.. I see nexenta sets it to 1024
[20:27:58] <haj> but the funny thing is the nexenta systems do it too... could be for other reasons though, I guess
[20:28:25] <haj> on this new one I'm testing on a locally mounted NFS-share, so I'm pretty sure the network isn't to blame.. :)
[20:28:47] <haj> just for testing.. :)
[20:28:51] <mui> then test how disabling sync affects performance
[20:28:56] <mui> zfs set sync=disabled foo/bar
[20:30:15] *** topping has joined #openindiana
[20:38:48] *** wcj is now known as cis
[20:39:55] <haj> btw. I'm using SSD's as log devices
[20:40:35] <viridari> disabling sync... if you're going to recommend that, make sure you tell the guy what he risks by doing that
[20:46:04] <mui> well, fundamentally if someone says to do something in irc
[20:46:11] <mui> its pretty good to figure out what is happening before typing
[20:46:46] <mui> but yes, if you fear loosing few secs of writes in case of electric failure during your performance tests then be aware of what sync setting does (:
[20:47:55] <mui> haj: yes, but if you don't experience performance impact after disabling sync, then it suggest problem might be in zil/ssd etc
[20:50:07] <kdavyd> haj, you could also be seeing an occurrence of https://www.illumos.org/issues/417
[20:50:26]
[20:50:45] <Patrickdk> I'm using 2k for nfs
[20:50:46] *** enmand has joined #openindiana
[20:51:06] <haj> mui: it still happens... I'm just noticing that I get a lot of errors on one of the controllers though...
[20:51:18] <mui> yeah
[20:51:22] <mui> well that might be to blame
[20:51:29] <haj> wonder if I can disable mpt_sas1 remote?
[20:51:41] <Patrickdk> hmm, I haven't had issues, with normal loads of 80MB/sec and spikes of 400MB/sec
[20:51:48] <mui> offline disks connected to it and then unconfigure with cfgadm
[20:52:08] *** girafe has quit IRC
[20:52:20] <haj> mui: I'm running multipathing.. ?
[20:52:20] <mui> depending on driver, you might be possible
[20:52:27] <mui> can't stay then
[20:52:45] <Patrickdk> oi 151a?
[20:52:56] <mui> also try disabling multipath if you experience perf issues and errors on another controller
[20:53:01] <mui> might be some issues
[20:55:10] *** spanglywires has joined #openindiana
[20:56:59] <haj> kdavyd: I've checked... I'm running v3 so I don't think it can be that problem... but thanks :)
[20:57:29] <kdavyd> haj, yea based on what you said it sounds like hardware/driver, not software
[20:57:49] <mui> http://xkcd.com/364/
[20:57:50] <mui> http://xkcd.com/364/
[20:57:51] <mui> http://xkcd.com/364/
[20:57:52] <mui> http://xkcd.com/364/
[20:58:18] <kdavyd> :)
[20:58:25] <mui> sorry
[20:58:34] <mui> cat found second mouse button
[20:58:53] <haj> mui: I'll try without multipathing
[20:58:56]
[21:01:06] <haj> kdavyd: I'll check later.. I'm going to check firmware revisions as well.. Had a problem that was caused by firmware last year... worse than this one though.. ;)
[21:01:19] <kdavyd> ... or, of course, it could just be ZFS
[21:02:13] <kdavyd> but there is insufficient data to tell.
[21:02:39] *** smrt has quit IRC
[21:02:57] *** smrt has joined #openindiana
[21:05:09] <haj> kdavyd: That should be test-able... I can just run my testscript that hangs the NFS-server on local disk.
[21:07:04]
[21:07:09] <kdavyd> dtrace -n 'profile-1001 { @[stack()] = count(); }'
[21:07:19] <kdavyd> Sample for about 60 seconds, then Ctrl-C
[21:07:22] <haj> just a minute.. i'm booting the servers
[21:07:48] *** spanglywires has quit IRC
[21:07:49] <haj> kdavyd: I'll get a sample for you.. :)
[21:08:43] <haj> that multipath disable is taking its time
[21:10:43] *** enmand has quit IRC
[21:17:16] *** enmand has joined #openindiana
[21:18:24] *** voidcoder has joined #openindiana
[21:19:40] <haj> mui: I seems to get way worse when disabling multipath... So I guess I must be a controller issue
[21:20:58] <kdavyd> LSIutil to the rescue, then :)
[21:21:22]
[21:22:26] *** _404_Error_ has quit IRC
[21:22:26] *** _404_Error_ has joined #openindiana
[21:23:22] *** OMV-User has quit IRC
[21:23:54] *** smrt has quit IRC
[21:24:12] *** smrt has joined #openindiana
[21:24:26] <haj> kdavyd: still want the trace?
[21:28:55] *** OMV-User has joined #openindiana
[21:29:57] <haj> kdavyd: actually it seems to be discs.. I have a couple of bad ones it would seem
[21:30:47] *** Monte has quit IRC
[21:36:05]
[21:38:54] *** smrt has quit IRC
[21:39:12] *** smrt has joined #openindiana
[21:42:57] <Kurlon> Can Grub boot a deduped and compressed root pool?
[21:43:07] <kdavyd> Kurlon, sure
[21:43:36] <Kurlon> Cool, might be a way to save some wear and tear on my USB flash.
[21:43:50] <kdavyd> Whether you want dedup on your rpool is a separate discussion topic - but yes, grub can boot it.
[21:45:12] <Kurlon> The downsides would be increased CPU and RAM use, but for a 16GB pool it can't be that much, right? Dedupped swap could be interesting I suppose...
[21:45:47] <Triskelios> downside would be poor write performance
[21:45:56] <Triskelios> unless you have enough memory to keep the DDT in cache
[21:46:04]
[21:46:29] <Kurlon> Considering the flash tops out at 1.5MB/sec I don't think I can hurt my write perf too much more can I? :D
[21:46:40] <kdavyd> Kurlon, actually... you can :)
[21:46:56] <Kurlon> Bugger... so maybe just compression then.
[21:47:17]
[21:49:02] <Tenzer> You should never use dedup for performance, only in order to save space, and only if your data really can be deduped
[21:49:44] *** smrt has quit IRC
[21:49:52] <haj> kdavyd: lsiutil shows errors on 5 out of 44 disks.. about half of them are used disks, the other half new.
[21:50:02] *** smrt has joined #openindiana
[21:52:30] <kdavyd> haj, clear the errors, do some writes, see if errors reoccur. Also, LSIutil will not only show errors on disks - it will show all PHYs, which includes expanders
[21:55:00] <haj> way ahead of you this time
[21:55:24] <haj> already cleared the errors... and they do reoccur
[21:56:13] <haj> hm .. but ok, didn't check if it was actually disks
[21:56:17] <haj> :)
[21:57:06] *** TomJ has joined #openindiana
[22:02:56] <cis> brb
[22:03:03] *** cis has quit IRC
[22:03:38] *** init3 has quit IRC
[22:12:10] *** enmand has quit IRC
[22:12:11] *** Kurlon has quit IRC
[22:12:11] *** Kurlon_ has joined #openindiana
[22:22:13] <haj> kdavyd: wonder if I'm reading this correctly.. link 4 must be one of the cables? http://v3.nathue.dk/~haj/sas-errors
[22:22:22] <haj> phy 3 that is
[22:22:45] <haj> i'm talking nonsense
[22:24:07] *** Kurlon has joined #openindiana
[22:24:40] <haj> errors on phy 0-3 on handle 0009 would look like a bad connection.. could be a cable? I have 4 cables between the controllers and the sas chassis
[22:25:21] <haj> might be overkill with 4 cables....
[22:27:10] *** Kurlon_ has quit IRC
[22:29:30] <haj> have to go.. :)
[22:29:33] <haj> *away*
[22:32:22] <Kurlon> How hard is it to setup multipath on OI anyway? I've got a couple FC chassis to play with, and a good dozen single and dual port FC HBAs to shove into my test system.
[22:33:29] *** timclassic has quit IRC
[22:35:55] <matsim> Kurlon: as target or as initiator? - I merely know the role as COMSTAR target
[22:36:33] *** Seony has joined #openindiana
[22:36:35] <Kurlon> I think initiator, I want to increase the machine's bandwidth to the chassis by using multiple links.
[22:36:51] <Kurlon> (I never remember which is which when talking target / initiator)
[22:37:12] <lennard> server/client = target/initiator
[22:37:22] <lennard> although officially you cant compare them I think :P
[22:37:28] <Kurlon> Can I use OI/COMSTAR to make LUNs on FC ports to feed to other computers?
[22:38:03] <matsim> that is possible if your HBAs are let's say >=4G QL or Emulex that I know of
[22:38:36] <Kurlon> Nope, I've got 1G and 2G QLogics only.
[22:38:57] <Kurlon> I *MIGHT8 have some 4G QLogics to play with but I'm not counting on it.
[22:39:38] <matsim> every QL with 4G that I know should be able to run with hte qlt (target mode) module instead of qlc
[22:40:13] <matsim> if you're speaking about 1G and 2G then it's possibly PCI-X - and PCI-X HBAs are quite affordable now on second-hand market
[22:40:22] *** ozquera has quit IRC
[22:41:07] <Kurlon> Yup, all the gear I have is PCI-X. So, are you saying there are different SFPs for target mode vs client mode on the 4Gs?
[22:42:30] <matsim> no you have the very same card but in the case of QL, you tell OI to load simply another driver for this type of HBAs or even per port
[22:43:13] <Kurlon> Ah, that could be cool to play with at some point.
[22:43:21] <matsim> it's only that the chip on the 4G and 8G HBAs are detected by qlt driver
[22:43:54] *** nvez has quit IRC
[22:46:50] *** flyz has quit IRC
[22:49:07] <matsim> Kurlon: So if you have a chassis with FC discs, and want them to be multipathed, look for MPXIO
[22:49:11] *** raichoo has joined #openindiana
[22:49:29] <Kurlon> Cool, I'll give that a go Monday
[22:49:52] <matsim> it doesn't seem trivial though, sometimes people had to switch from round-robin to failover only config to get it working
[22:52:58] <kdavyd> round-robin to logical-block might be another thing to try
[22:54:42]
[22:56:02]
[22:56:49] *** smrt has quit IRC
[22:57:07] *** smrt has joined #openindiana
[22:58:23]
[22:58:38] *** heldus has joined #openindiana
[23:00:13] *** DocHoliday has joined #openindiana
[23:15:37] *** init3 has joined #openindiana
[23:17:13] *** init3 has quit IRC
[23:24:11] *** init3 has joined #openindiana
[23:31:18] *** enmand has joined #openindiana
[23:35:52] *** enmand has quit IRC
[23:39:19] <finnland_> has anyone gotten USB in virtual working in a zone?
[23:39:31] <finnland_> vitualbox even
[23:41:26] <tomww> you want USB in a Vbox guest *or* want a vbox instance running in a non-global-zone and have access to USB ports of the hosts's mainboard ?
[23:42:54] <finnland_> the second one
[23:43:14] <finnland_> vbox in zone having access to physical usb
[23:44:52] <finnland_> i have /dev/vboxusbmon available in the zone but VBoxManage list usbhost always shows <none> but are shown in global
[23:46:42] <Triskelios> VBox has to change the driver configuration to capture USB devices; you can't do that in a non-global zone
[23:47:50] <finnland_> hmm, so wha's the nonsens e in the manual with adding vboxusbmon to the zone?
[23:49:49] <Triskelios> finnland_: oh, huh, guess they added zone support for that
[23:50:13] <Triskelios> it probably needs a service in the GZ
[23:51:50] <finnland_> i have the virtualbox/zoneaccess:default service running in global so that should be enough
[23:55:22] <tomww> <device match="/dev/vboxusbmon"/> <device match="/dev/vboxdrv"/> <device match="/dev/usb"/>
[23:56:01] <finnland_> mkay, so /dev/usb is also needed
[23:56:21] <finnland_> much thanks
top

   February 11, 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 | >