Switch to DuckDuckGo Search
   July 2, 2019
< | 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
[00:01:18] *** sarvet <sarvet!~sarvet@dslb-094-220-207-064.094.220.pools.vodafone-ip.de> has joined #illumos
[00:02:28] *** sarvet <sarvet!~sarvet@dslb-094-220-207-064.094.220.pools.vodafone-ip.de> has quit IRC (Remote host closed the connection)
[00:12:30] *** andy_js <andy_js!~andy@94.12.186.242> has quit IRC (Quit: andy_js)
[00:18:14] *** gh34 <gh34!~textual@cpe-184-58-181-106.wi.res.rr.com> has quit IRC (Quit: Textual IRC Client: www.textualapp.com)
[00:19:57] <gvm> I've successfully put together a script that reproducibly demonstrates the hole_birth issue that I had a few days ago which caused zfs send/receive failure and incomplete scrubs being recorded as complete.
[00:20:09] <gvm> To definitively prove that this is down to feature@hole_birth, I've been trying unsuccessfully to create a pool on OI Hipster20190511 without feature@hole_birth (or, at least, with feature@hole_birth disabled).
[00:20:22] <gvm> "zpool create -o feature@hole_birth=disabled" gives "property 'feature@hole_birth' can only be set to 'enabled'".
[00:20:31] <gvm> Creating pools with specific feature sets isn't something I've experimented with previouly, but I'd thought that this was the whole point of having feature sets. Maybe not.
[00:20:41] <gvm> I came across this posting by CyberJock that states that it's not: https://www.ixsystems.com/community/threads/zpool-features.18651/#post-103157 . Is this really true??
[00:20:54] <gvm> If I start with a pool created by a (very) old version and then individually zpool upgrade every other feature flag except this one - would that work?
[00:28:07] <LeftWing> gvm: Yes, I think today you can start with a v28 pool and move forward
[00:30:45] <gvm> LeftWing: But is there really no way to create a v28 pool on 20190511?
[00:31:24] <LeftWing> I think you can with some set of flags?
[00:31:39] <LeftWing> I'm sure I saw some suggestion of being able to specify the version number
[00:31:47] <LeftWing> With zpool create
[00:34:03] <gvm> I think I might have found it - man page says the "-d" option is to create with no feature flags at all
[00:34:55] <gvm> So create with "-d" then add features. Would be nice to have an "all features except X" option, but...
[00:35:44] <gvm> Couldn't find anything about specifying numbers. I'm assuming that "-d" will create a version 5000.
[00:38:12] *** arekinath <arekinath!~arekinath@pdpc/supporter/professional/arekinath> has quit IRC (Remote host closed the connection)
[00:38:12] *** melloc <melloc!~codytimec@timecube.club> has quit IRC (Remote host closed the connection)
[00:39:08] <gvm> So the only way to do an "-except feature X" is to create a pool, see what disabled features it has, and enable all except the ones you don't want. And don't make any mistakes, because it's one-way!
[00:39:38] *** melloc <melloc!~codytimec@timecube.club> has joined #illumos
[00:39:53] *** arekinath <arekinath!~arekinath@pdpc/supporter/professional/arekinath> has joined #illumos
[00:41:19] <gvm> OK. It's doable. Just.
[00:46:45] *** aszeszo <aszeszo!~aszeszo@unaffiliated/aszeszo> has quit IRC (Quit: aszeszo)
[00:49:19] <gvm> Anyhow - I managed to put together a script that reproduces the zfs send/receive and incomplete scrub reported as successful problems that we discussed previously. It shows up the problem occuring after a fairly lengthy sequence of operations on a group of pools. It runs OK (i.e. it fails as expected) in a VM, though. Just want to add a final check to show that it doesn't fail if...
[00:49:21] <gvm> ...feature@hole_birth is disabled.
[01:25:03] *** merzo <merzo!~merzo@228-31-133-95.pool.ukrtel.net> has quit IRC (Ping timeout: 245 seconds)
[01:38:55] *** merzo <merzo!~merzo@95-54-132-95.pool.ukrtel.net> has joined #illumos
[02:17:33] *** pmooney <pmooney!~pmooney@67-4-175-230.mpls.qwest.net> has quit IRC (Quit: WeeChat 2.4)
[02:25:44] <richlowe> there's literally no way to use flar anymore, right?
[02:25:47] <richlowe> even on tribblix or something.
[02:30:27] *** MilkmanDan <MilkmanDan!~dan@wilug/expat/MilkmanDan> has quit IRC (Ping timeout: 245 seconds)
[02:32:06] *** MilkmanDan <MilkmanDan!~dan@wilug/expat/MilkmanDan> has joined #illumos
[03:00:38] * bdha would sure hope not.
[03:07:39] <richlowe> I just stumbled into the libc hooks we still have
[03:13:22] <richlowe> every time I have to touch Targetdirs I long for mtree
[03:13:40] <richlowe> tsoome: you're a resident BSD apologist currently, any interest in dropping mtree in tools, andu sing it in place of Targetdirs? :)
[03:20:00] *** ed209 <ed209!~ed209@165.225.128.67> has quit IRC (Remote host closed the connection)
[03:20:08] *** ed209 <ed209!~ed209@165.225.128.67> has joined #illumos
[03:36:40] *** baojg <baojg!~baojg@162.243.44.213> has joined #illumos
[04:12:38] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has quit IRC (Remote host closed the connection)
[04:12:55] <jbk> building this makes me wonder how insane multi-threaded ld would be :)
[04:15:30] <richlowe> not insane, just likely not helpful.
[04:16:38] <richlowe> I guess depending on how excessively crazy the input is, bits would go well.
[04:20:30] <richlowe> jbk: profile it, but a lot of the things that seem heaviest would also be manipulating the output image and hold the biggest lock
[04:21:02] <richlowe> but in your case for all input processing we could go infinitely wide, because you have _so_ many sections.
[04:21:37] <richlowe> output I don't think we could go much more than a thread per output section
[04:56:17] *** shruti` <shruti`!shruti@nat/redhat/x-vtduyvegkkdyvkit> has quit IRC (Ping timeout: 245 seconds)
[05:01:51] *** shruti <shruti!shruti@nat/redhat/x-lgewypwoxlapxzzz> has joined #illumos
[05:43:09] *** despair86 <despair86!~despair@2600:1700:8f61:163f:2ad:24ff:fe90:9f5d> has joined #illumos
[06:08:43] *** xanderlent <xanderlent!~xander@flanflanflan.xanderlent.net> has quit IRC (Quit: ZNC - https://znc.in)
[06:14:50] *** funnyguy <funnyguy!~funnyguy@74.207.236.96> has joined #illumos
[06:26:44] *** xanderlent <xanderlent!xander@flanflanflan.xanderlent.net> has joined #illumos
[06:46:00] *** despair86 <despair86!~despair@2600:1700:8f61:163f:2ad:24ff:fe90:9f5d> has quit IRC (Read error: Connection reset by peer)
[06:46:02] *** despair_ <despair_!~despair@2600:1700:8f61:163f:2ad:24ff:fe90:9f5d> has joined #illumos
[06:53:57] *** triffid <triffid!triffid@lovecraft-ipv6.mcclung.systems> has joined #illumos
[06:55:14] <tsoome> richlowe: not before I am done with installboot update:)
[07:04:56] *** despair_ <despair_!~despair@2600:1700:8f61:163f:2ad:24ff:fe90:9f5d> has quit IRC (Quit: Leaving)
[07:05:39] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Quit: tsoome)
[07:05:48] *** hemi770 <hemi770!~hemi666@unaffiliated/hemi770> has quit IRC (Ping timeout: 252 seconds)
[07:07:28] *** hemi770 <hemi770!~hemi666@unaffiliated/hemi770> has joined #illumos
[07:20:16] *** BOKALDO <BOKALDO!~BOKALDO@81.198.159.225> has joined #illumos
[07:50:04] <sjorge> gvm I think you can specify -o version=28 on zpool create
[07:56:37] *** jimklimov <jimklimov!~jimklimov@ip-86-49-254-8.net.upcbroadband.cz> has quit IRC (Ping timeout: 258 seconds)
[08:06:25] *** shruti <shruti!shruti@nat/redhat/x-lgewypwoxlapxzzz> has quit IRC (Ping timeout: 248 seconds)
[08:06:38] *** shruti <shruti!shruti@nat/redhat/x-pibppwannjujlueq> has joined #illumos
[08:16:30] *** tsoome <tsoome!~tsoome@138-103-235-80.sta.estpak.ee> has joined #illumos
[08:26:10] *** aszeszo <aszeszo!~aszeszo@unaffiliated/aszeszo> has joined #illumos
[08:36:22] *** alanc <alanc!~alanc@129.157.69.51> has quit IRC (Remote host closed the connection)
[08:36:49] *** alanc <alanc!~alanc@129.157.69.51> has joined #illumos
[09:05:19] *** jelmd <jelmd!~purple@port-ip-88-150-100-107.reverse.mdcc-fun.de> has quit IRC (Ping timeout: 244 seconds)
[09:18:36] *** andy_js <andy_js!~andy@94.12.186.242> has joined #illumos
[09:21:56] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has joined #illumos
[09:39:31] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has joined #illumos
[09:47:56] <ptribble> richlowe: we have no way to create a flar archive any more
[09:48:13] <ptribble> but there's code to install a flar into a zone
[09:48:22] <ptribble> usr/src/lib/brand/shared/zone/common.ksh
[10:15:09] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has quit IRC (Read error: Connection reset by peer)
[10:15:39] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has joined #illumos
[10:27:21] *** man_u <man_u!~manu@manu2.gandi.net> has joined #illumos
[10:34:11] *** merzo <merzo!~merzo@95-54-132-95.pool.ukrtel.net> has quit IRC (Ping timeout: 244 seconds)
[10:46:42] *** aszeszo <aszeszo!~aszeszo@unaffiliated/aszeszo> has quit IRC (Ping timeout: 245 seconds)
[11:01:14] *** aszeszo <aszeszo!~aszeszo@unaffiliated/aszeszo> has joined #illumos
[11:13:03] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has quit IRC (Quit: Leaving)
[11:35:37] <gvm> sjorge: Yes, you're right - Thanks! - this does work. Looking back, it is in the man page too, under the list of properties that can be set and changed. It mainly talks about upgrading pools in that section, which is how I missed it, but it is there.
[11:35:56] <gvm> Actually, I don't need to go as far back as v28 for this particular issue, since it appears as late as 2015.
[11:36:31] <gvm> And latest test shows that it occurs even when feature@hole_birth is disabled, which is a bit unexpected.
[11:52:56] *** tsoome <tsoome!~tsoome@138-103-235-80.sta.estpak.ee> has quit IRC (Quit: tsoome)
[11:53:08] *** merzo <merzo!~merzo@185.39.197.205> has joined #illumos
[11:59:37] *** aszeszo <aszeszo!~aszeszo@unaffiliated/aszeszo> has quit IRC (Ping timeout: 245 seconds)
[12:06:56] *** merzo <merzo!~merzo@185.39.197.205> has quit IRC (Ping timeout: 258 seconds)
[12:09:55] *** baojg <baojg!~baojg@162.243.44.213> has quit IRC (Remote host closed the connection)
[12:13:33] *** aszeszo <aszeszo!~aszeszo@unaffiliated/aszeszo> has joined #illumos
[12:20:00] *** ed209 <ed209!~ed209@165.225.128.67> has quit IRC (Remote host closed the connection)
[12:20:07] *** ed209 <ed209!~ed209@165.225.128.67> has joined #illumos
[12:22:37] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has joined #illumos
[12:26:34] *** merzo <merzo!~merzo@185.39.197.205> has joined #illumos
[12:38:37] *** baojg <baojg!~baojg@162.243.44.213> has joined #illumos
[12:43:09] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[12:59:55] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has quit IRC (Read error: Connection reset by peer)
[13:00:23] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has joined #illumos
[13:01:45] *** elegast <elegast!~elegast@83-161-171-20.mobile.xs4all.nl> has joined #illumos
[13:09:01] *** merzo <merzo!~merzo@185.39.197.205> has quit IRC (Ping timeout: 246 seconds)
[13:15:08] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has quit IRC (Read error: Connection reset by peer)
[13:15:44] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has joined #illumos
[13:26:38] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Quit: tsoome)
[13:30:03] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[13:30:07] *** kvik <kvik!~kvik@unaffiliated/kvik> has joined #illumos
[13:41:18] *** amrfrsh <amrfrsh!~Thunderbi@190.2.145.106> has quit IRC (Ping timeout: 245 seconds)
[13:45:18] *** BOKALDO <BOKALDO!~BOKALDO@81.198.159.225> has quit IRC (Quit: Leaving)
[13:45:35] *** baojg <baojg!~baojg@162.243.44.213> has quit IRC (Remote host closed the connection)
[13:46:55] *** BOKALDO <BOKALDO!~BOKALDO@87.110.91.52> has joined #illumos
[14:01:58] *** mno-hime <mno-hime!~mno-hime@94.142.238.232> has quit IRC (Quit: Leaving)
[14:02:14] *** merzo <merzo!~merzo@185.39.197.205> has joined #illumos
[14:04:46] *** mno-hime <mno-hime!~mno-hime@94.142.238.232> has joined #illumos
[14:07:47] *** gh34 <gh34!~textual@cpe-184-58-181-106.wi.res.rr.com> has joined #illumos
[14:46:17] *** Kurlon_ <Kurlon_!~Kurlon@cpe-67-253-141-249.rochester.res.rr.com> has quit IRC (Ping timeout: 245 seconds)
[15:07:05] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has quit IRC (Remote host closed the connection)
[15:09:04] *** hugoAgogo <hugoAgogo!~hugoAgogo@148.252.250.12> has joined #illumos
[15:26:21] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has joined #illumos
[15:47:39] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[15:53:10] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Ping timeout: 246 seconds)
[16:04:43] *** amrfrsh <amrfrsh!~Thunderbi@host-62514ecc6afffe842a0f.ip6.padnet.de> has joined #illumos
[16:05:52] *** MilkmanDan <MilkmanDan!~dan@wilug/expat/MilkmanDan> has quit IRC (Ping timeout: 245 seconds)
[16:08:39] *** MilkmanDan <MilkmanDan!~dan@wilug/expat/MilkmanDan> has joined #illumos
[16:27:28] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[16:28:48] *** khng300 <khng300!~khng300@unaffiliated/khng300> has quit IRC (Ping timeout: 245 seconds)
[16:50:01] *** khng300 <khng300!~khng300@unaffiliated/khng300> has joined #illumos
[16:51:43] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has quit IRC (Remote host closed the connection)
[16:57:11] *** khng300 <khng300!~khng300@unaffiliated/khng300> has quit IRC (Remote host closed the connection)
[16:57:28] *** pmooney <pmooney!~pmooney@67-4-175-230.mpls.qwest.net> has joined #illumos
[17:00:10] *** spectral` <spectral`!~julian@os.k/spectral> has quit IRC (Quit: Free ZNC ~ Powered by LunarBNC: https://LunarBNC.net)
[17:00:52] *** elegast <elegast!~elegast@83-161-171-20.mobile.xs4all.nl> has quit IRC (Ping timeout: 245 seconds)
[17:07:28] *** spectral` <spectral`!~julian@spectral.powered.by.lunarbnc.net> has joined #illumos
[17:07:28] *** spectral` <spectral`!~julian@spectral.powered.by.lunarbnc.net> has quit IRC (Changing host)
[17:07:28] *** spectral` <spectral`!~julian@os.k/spectral> has joined #illumos
[17:07:43] *** khng300 <khng300!~khng300@unaffiliated/khng300> has joined #illumos
[17:08:24] *** jimklimov <jimklimov!~jimklimov@31.7.243.238> has quit IRC (Ping timeout: 244 seconds)
[17:13:14] *** merzo <merzo!~merzo@185.39.197.205> has quit IRC (Read error: Connection reset by peer)
[17:13:31] *** merzo <merzo!~merzo@185.39.197.205> has joined #illumos
[17:17:21] *** baojg <baojg!~baojg@162.243.44.213> has joined #illumos
[17:41:58] *** MarcelT <MarcelT!~marcel@tortuga.telka.sk> has quit IRC (Ping timeout: 258 seconds)
[17:42:31] *** MarcelT <MarcelT!~marcel@tortuga.telka.sk> has joined #illumos
[18:13:59] *** aszeszo <aszeszo!~aszeszo@unaffiliated/aszeszo> has quit IRC (Remote host closed the connection)
[18:14:10] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 258 seconds)
[18:16:59] *** aszeszo <aszeszo!~aszeszo@unaffiliated/aszeszo> has joined #illumos
[18:34:51] *** hugoAgogo <hugoAgogo!~hugoAgogo@148.252.250.12> has quit IRC (Ping timeout: 268 seconds)
[18:39:22] *** man_u <man_u!~manu@manu2.gandi.net> has quit IRC (Quit: man_u)
[18:41:31] *** merzo <merzo!~merzo@185.39.197.205> has quit IRC (Ping timeout: 246 seconds)
[18:42:32] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[19:18:57] *** merzo <merzo!~merzo@185.39.197.205> has joined #illumos
[19:19:45] *** amrfrsh <amrfrsh!~Thunderbi@host-62514ecc6afffe842a0f.ip6.padnet.de> has quit IRC (Quit: amrfrsh)
[19:20:59] *** amrfrsh <amrfrsh!~Thunderbi@host-62514ecc6afffe842a0f.ip6.padnet.de> has joined #illumos
[19:38:56] *** amrfrsh <amrfrsh!~Thunderbi@host-62514ecc6afffe842a0f.ip6.padnet.de> has quit IRC (Ping timeout: 252 seconds)
[19:40:59] *** amrfrsh <amrfrsh!~Thunderbi@190.2.145.106> has joined #illumos
[19:44:15] <nahamu> Nexenta was acquired by DDN: https://nexenta.com/company/media/press-releases/ddn-completes-acquisition-nexenta
[19:44:36] <nahamu> Oh, old-ish news.
[19:44:43] <LeftWing> Yeah that happened a little while ago
[19:51:43] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 245 seconds)
[19:55:04] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[19:55:58] <gitomat> [illumos-gate] 11266 mandoc(1) does not support a RETURN VALUES section for 7I man pages -- Jason King <jason.king at joyent dot com>
[20:09:03] <richlowe> ptribble: There's libc hooks in lib/libc/i386/etc too
[20:09:09] <richlowe> to deal with the mounts.
[20:51:32] *** neirac <neirac!~cneira@190.162.109.53> has joined #illumos
[21:07:07] *** MilkmanDan <MilkmanDan!~dan@wilug/expat/MilkmanDan> has quit IRC (Ping timeout: 245 seconds)
[21:09:11] *** MilkmanDan <MilkmanDan!~dan@wilug/expat/MilkmanDan> has joined #illumos
[21:14:38] *** shruti <shruti!shruti@nat/redhat/x-pibppwannjujlueq> has quit IRC (Ping timeout: 245 seconds)
[21:15:22] *** kvik <kvik!~kvik@unaffiliated/kvik> has quit IRC (Ping timeout: 244 seconds)
[21:15:28] *** kvik <kvik!~kvik@unaffiliated/kvik> has joined #illumos
[21:16:55] *** ikonia <ikonia!~irc@unaffiliated/ikonia> has quit IRC (Ping timeout: 244 seconds)
[21:17:14] *** ikonia <ikonia!~irc@unaffiliated/ikonia> has joined #illumos
[21:20:07] *** shruti <shruti!shruti@nat/redhat/x-usfuihonselusgrl> has joined #illumos
[21:25:03] *** shruti <shruti!shruti@nat/redhat/x-usfuihonselusgrl> has quit IRC (Ping timeout: 268 seconds)
[21:25:20] *** BOKALDO <BOKALDO!~BOKALDO@87.110.91.52> has quit IRC (Quit: Leaving)
[21:25:20] *** shruti <shruti!shruti@nat/redhat/x-coubdpmngywheaqi> has joined #illumos
[21:53:22] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 245 seconds)
[21:54:18] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #illumos
[22:00:22] *** khng300 <khng300!~khng300@unaffiliated/khng300> has quit IRC (Remote host closed the connection)
[22:04:42] *** khng300 <khng300!~khng300@unaffiliated/khng300> has joined #illumos
[22:20:00] *** ed209 <ed209!~ed209@165.225.128.67> has quit IRC (Remote host closed the connection)
[22:20:21] *** ed209 <ed209!~ed209@165.225.128.67> has joined #illumos
[22:32:01] *** merzo <merzo!~merzo@185.39.197.205> has quit IRC (Ping timeout: 248 seconds)
[22:33:40] <igork> richlowe: i see i was not crazy to ask about one default.run for zfs tests :)
[22:34:09] *** khng300 <khng300!~khng300@unaffiliated/khng300> has quit IRC (Ping timeout: 248 seconds)
[22:36:06] <igork> next one will be migration /usr/bin/ksh -> /bin/ksh - for compatibility with others platforms
[22:37:41] *** khng300 <khng300!~khng300@unaffiliated/khng300> has joined #illumos
[22:52:20] *** Kurlon_ <Kurlon_!~Kurlon@bidd-pub-04.gwi.net> has joined #illumos
[22:56:04] *** _Tenchi_ <_Tenchi_!~phil@d-207-255-80-203.paw.cpe.atlanticbb.net> has quit IRC (Read error: Connection reset by peer)
[22:56:16] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 272 seconds)
[22:56:24] <richlowe> You're assuming me asking something means it's not crazy :)
[22:56:48] <rmustacc> It's what we've asked for all the other runfiles, so it's not unreasonable.
[22:56:56] *** Kurlon_ <Kurlon_!~Kurlon@bidd-pub-04.gwi.net> has quit IRC (Ping timeout: 268 seconds)
[22:57:20] *** _Tenchi_ <_Tenchi_!~phil@d-207-255-80-203.paw.cpe.atlanticbb.net> has joined #illumos
[23:16:17] *** gh34 <gh34!~textual@cpe-184-58-181-106.wi.res.rr.com> has quit IRC (Ping timeout: 245 seconds)
[23:18:39] <andyf> Can anyone help me troubleshoot a SuperMicro EPYC system that is resetting during boot? I have console access and it does not drop into kmdb, it just appears to reset (although the iKVM could be hiding something). rmustacc suggested putting breakpoints on the trap handlers, but I don't know which.
[23:19:41] *** khng300 <khng300!~khng300@unaffiliated/khng300> has quit IRC (Ping timeout: 258 seconds)
[23:20:13] *** khng300 <khng300!~khng300@unaffiliated/khng300> has joined #illumos
[23:20:42] <rmustacc> Well, the first thing I'd see is if the IPMI log says anything.
[23:21:32] <rmustacc> In terms of the fault handlers, usually if something is rather messed up, we'll end up in a double fault handler, fault again, which causes a triple fault and reset.
[23:21:42] *** MilkmanDan <MilkmanDan!~dan@wilug/expat/MilkmanDan> has quit IRC (Ping timeout: 245 seconds)
[23:22:48] *** MilkmanDan <MilkmanDan!~dan@wilug/expat/MilkmanDan> has joined #illumos
[23:22:54] <rmustacc> So I would probably try something like on tr_syserrtrap
[23:23:19] <andyf> The IPMI SEL does not seem to contain anything useful - just power on, power off events
[23:23:33] <rmustacc> OK
[23:24:26] <andyf> I'll try a bp on that and see how far it gets
[23:25:01] <rmustacc> Not sure if that'll fire or not.
[23:25:15] <rmustacc> Otherwise, maybe turning on some amount of trap debugging and other stuff would be the next step.
[23:25:50] <andyf> I've tried various breakpoints on things called from main() up to `sched()` and it always caught it. One problem is that I don't seem to have keyboard input once it drops into kmdb
[23:26:06] <andyf> well, I do for the initial kmdb after a debug boot, but not when it stops on a breakpoint
[23:26:13] <rmustacc> Wait, you're using the iKVM?
[23:26:16] <rmustacc> Yeah, don't do that. :/
[23:26:34] <rmustacc> You want to use ttyb or another one and use the IPMI serial console.
[23:26:41] <rmustacc> If you can.
[23:26:52] <andyf> I'll see - it's a users system, he's just given me temporary access
[23:27:07] <rmustacc> I understand.
[23:27:15] <rmustacc> Just tryin to be honest about it being a very hard thing to debug via.
[23:27:24] <rmustacc> Because of the console related issues.
[23:27:37] <rmustacc> As the keyboard is emulated USB, not emulated PS/2.
[23:28:13] <andyf> It depends on whether ipmi is allowed from my address, I will certainly try
[23:28:24] <rmustacc> I understand.
[23:28:54] <rmustacc> Sorry I don't have a better initial suggestion.
[23:29:44] <jbk> some supermicro systems have a watchdog timer built in who's behavior is controlled by a jumper
[23:29:51] <andyf> Any help is welcome, thanks - syserrtrap did not fire
[23:29:54] <jbk> could that be something worth checking/eliminating?
[23:30:26] <rmustacc> That's true, there is a BIOS option that controls that.
[23:31:09] <jbk> on mine, I believe i can set a jumper to have it send an NMI if enabled (one of these days I was going to look at writing a driver for it)
[23:32:15] <jbk> I think all of the support for watchdog timers in illumos is on the sparc side (I don't recall seeing any drivers for x86)
[23:33:00] <jbk> but anyway, that might be something to double check
[23:33:24] <andyf> It looks like I only have access to 443 so far.. I'll drop him a line asking for ipmi
[23:38:50] <andyf> rmustacc - do you have any pointers on turning on trap debugging etc?
[23:42:06] *** Kurlon <Kurlon!~Kurlon@cpe-67-253-141-249.rochester.res.rr.com> has joined #illumos
[23:42:21] <rmustacc> Thre's a variable tdebug.
[23:42:26] <rmustacc> It will slow things down a lot.
[23:42:32] <rmustacc> Also lodebug and faultdebug.
[23:43:25] <rmustacc> tdebug basically causes one to show all the trap register states and other information anytime we take a trap.
[23:43:33] *** jellydonut <jellydonut!~quassel@s91904425.blix.com> has quit IRC (Quit: jellydonut)
[23:43:37] <rmustacc> Other things that could be useful are turning on more advanced moddebug.
[23:43:58] <rmustacc> Other random things that you can try is turning off multiple CPUs, kpti, or some other things I guess.
[23:45:02] <andyf> Thanks again, that gives me some more things to try while I wait for SOL access
[23:45:18] <rmustacc> Sorry, I don't really have a great methodology.
[23:45:30] *** jellydonut <jellydonut!~quassel@s91904425.blix.com> has joined #illumos
[23:46:18] <andyf> Debugging is more of an art anyway. I'll learn quite a lot just by reading trap.c now so I appreciate the pointers
[23:48:07] *** khng300 <khng300!~khng300@unaffiliated/khng300> has quit IRC (Ping timeout: 246 seconds)
[23:50:31] *** tg2 <tg2!~tg2@205.204.66.35> has joined #illumos
[23:52:22] *** khng300 <khng300!~khng300@unaffiliated/khng300> has joined #illumos
top

   July 2, 2019
< | 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