[00:53] hi all. i'm struggling to get crashdump working on ubuntu 13.10 running as a virtualbox guest on a mac. trying to follow https://wiki.ubuntu.com/Kernel/CrashdumpRecipe [00:53] i found someone having similar problems at https://bugs.launchpad.net/ubuntu/+source/kexec-tools/+bug/1235616 [00:53] Launchpad bug 785394 in grub (Ubuntu Saucy) "duplicate for #1235616 Hard-coded crashkernel=... memory reservation in /etc/grub.d/10_linux is insufficient" [Medium,In progress] [00:54] anyone have advice for me? should this work, and i need to just keep futzing, or are there known issues? === gerald is now known as Guest89221 [07:22] yaaaaaaaaaaaaaaaaawwwwwnnnn... :) [07:22] * ppisati goes for a lot of coffee... [07:46] * smb joins the cup-ride [11:44] smb, so i have simple-f cause i am on -5 [11:44] E: DRIVER=simple-framebuffer [11:44] i'd not see that if i updated you say yes ? [11:44] apw, yep [11:44] ok [11:44] then all is good [11:45] apw, It was the updated kernel that caused me to fight efi-framebuffer (as simple was gone) [11:45] ahh heh [11:46] anyhow when that version hits, i would like to know if that works ok on your radeon box [11:46] or whether i need to fix vesafb [11:48] apw, So I ran with that changed locally and this makes vesafb still load but the idev event has not the primary env var set [11:49] probably what we expect [11:49] Still comes up, got that weird login bug again... but that could be fluke [12:08] well thats something [12:11] rsalveti, hey ... mako, we have just gotten a security team patch set for this kernel, but there is a version just sitting in our PPA for your testing, shuld i just replace it and we reset from there or do you want to do something with it first [12:24] apw, So a qemu uri for virsh is "virsh -c qemu+ssh://@/system [12:38] apw: better to just replace it with the security fix, and rebase the tree with the current changes [12:39] apw: as we're not done yet with android 4.4 [12:42] rsalveti, this is bringing in yama not a security fix, so i don't need to drop it on wahts in the archive at all [12:42] rsalveti, i was more thinking shall i update the 4.4 version in the PPA with it [12:42] apw: oh, then yeah [12:42] that would be better [12:42] and easier [12:42] ok ... thanks will do that [13:17] jjohansen, I don't think the config patch for bug #1270215 is doing the right thing according to the bug reporter. [13:17] Launchpad bug 1270215 in linux (Ubuntu Precise) "kernel 3.13.0-4.19~precise1-generic: no internet via ethernet or WiFi" [High,Confirmed] https://launchpad.net/bugs/1270215 [13:51] bjf, did that Trusty AA oops go away in CI testing ? [13:53] rtg, don't know for sure http://ci.ubuntu.com/sru/trusty/version/3.13.0-7.26-generic/ [13:54] bjf, is the dashboard accurate now ? [13:54] psivaa, have you seen that panic with the new trusty kernel? [13:54] rtg, no, but it tells you that the tests ran to completion and results got posted [13:55] ok [13:55] bjf: not yet, there was a temp archive issue and cobbler in the morning. the tests are running on amd64 machines now. [13:55] psivaa, cool thanks [13:56] yw, i'll keep you posted if i see that again [13:56] psivaa, appreciated. you can let me or any of us know in this channel [13:56] bjf: ack [14:10] Does anybody know if the kernel in the netboot package for 12.04.4 will finally have support for the Intel I210/I217 NICs? [14:11] those drivers have been available forever [14:13] That would make life soo much easier for folks that have gigantic clusters of boxes with those NICs in them [14:14] DW-10297, if you can tell us what the driver for that is, we can check [14:17] indeed, we ought to be able to netboot an Intel NIC [14:17] yeah when those nics came out 18 months ago i requested this and i was told to shove it =D [14:18] and several releases of the netboot package came and went [14:18] DW-10297, did you file a bug about it? [14:18] and if so what was the number, so i can see what happeneed to it [14:19] it might have been helpful to test the beta images a couple of weeks ago, then we could have made sure [14:20] Intel(R) PRO/1000 Network Driver - 2.1.4-k in 13.04 sees the target NICs [14:20] what does lsmod say is bound to it [14:21] it says used by 0 [14:22] but in dmesg when the kernel starts its indicating that it's using that driver for the NIC [14:25] on a supermicro server board that has Intel Corporation I210 Gigabit (lspci) it uses igb 5.0.5-k (in centos) [14:26] so i guess as long as it has at least 2.1.4-k of e1000e and igb 5.0.5-k (or whatever) it should work fine [14:27] by the way here is the bug that someone else filed for this a lonnggg time ago [14:27] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1182878 [14:27] Launchpad bug 1182878 in linux (Ubuntu) "i210/i217 unsupported by igb / e1000e driver in precise" [Medium,Invalid] [14:27] and it was marked invalid [14:27] for no reason === ghostcube_ is now known as ghostcube [14:31] apparently this should've been fixed in 12.04.3 [14:32] but i can confirm it doesnt work [14:32] at least not in the netboot kernel [14:32] even the jan 10 release of the netboot package [14:34] DW-10297, according to comment 7 of that bug, the person that filed the bug says it is working for him and it's no longer valid [14:34] it might work if you install it from a cd [14:35] DW-10297, that was done june of last year. if you disagreed, you could have commented on the bug or opened your own [14:35] DW-10297, it would sure help if you would start a bug and attach a dmesg to it so we can start looking at some specifics. Intel typically has really good support for their drivers. [14:36] its not a problem with intel's drivers, it's that the kernel in http://archive.ubuntu.com/ubuntu/dists/precise-updates/main/installer-amd64/current/images/netboot/ubuntu-installer/amd64/ doesn't have the right driver [14:39] DW-10297, a 3.2 kernel likely _won't_ have the right driver. if you have the same HW as in bug 1182878, then you'll need at least a 3.5 kernel, i.e., 12.04.2 [14:39] Launchpad bug 1182878 in linux (Ubuntu) "i210/i217 unsupported by igb / e1000e driver in precise" [Medium,Invalid] https://launchpad.net/bugs/1182878 [14:39] rtg: Okay, i'm not arguing that. I'm trying to figure out why the netboot images aren't kept up to date with whats going on in the actual kernel [14:40] I asked #ubuntu-installer, they told me to ask you guys [14:40] DW-10297, ok, I'm just researching that. I admit I've never net booted 12.04.2 [14:40] in fact, I rarely netboot at all [14:40] its slightly useful in clouds [14:44] I just found it odd that there was just a new release of ubuntu installer netboot on Jan 10 and it still didnt include support for Haswell [14:44] DW-10297, I'm sure there is a way to do it, but it may require you to configure your own Cobbler or MAAS servers [14:44] it is likely missing from the d-i configuration, the upstream peeps just love to change the names of their drivers for shits and giggles, and then they get lost [14:46] I think Debian 7 may actually have the driver in their installer as well. I can check if it helps [14:48] anyway I will just cross my fingers and hope that the 12.04.4 netboot works =D [14:49] if you could test that and if it does not file a bug against 'linux' (ubuntu-bug linux) and put all the details in it, which driver it is using etc then we might be able to fix it [14:50] i will just see if 12.04.4's install fixes it first, if not i'll try it on debian and see how that goes [14:50] .4 should be out sometime soon i would hope [14:51] DW-10297, today is the release day for .4 [14:51] DW-10297, and why a bug report before today would have been helpful [14:52] yeah I knew that but i didnt want to be like.. if they ever publish it to the mirrors haha =) [14:52] I guess I just assumed that it would make sense to refactor the kernels used in the installer every once in awhile to add new drivers [14:52] even RHEL does that [14:52] (surprisingly) [14:52] smb: hello [14:53] smb: i'm still unable to reproduce bug 1273386 , i was running those scripts in a loop [14:53] Launchpad bug 1273386 in neutron "Neutron namespace metadata proxy triggers kernel crash on Ubuntu 12.04/3.2 kernel" [Critical,Triaged] https://launchpad.net/bugs/1273386 [14:55] smb: i have the devstack setup in a VM i could give access to somebody who may have more experience with setting it up, maybe i'm missing some condition [14:59] arges, I just started to deploy a devstack setup to one of my VMs. In parallel do you want to remind Salvatore we might have better chances with some independent reproducer he said he may provide? [15:00] smb: antoher option is to give him access to the VM to get it set up properly [15:00] smb: cause all we care about is getting a proper crash dump [15:00] arges, Yeah true [15:00] i did get to learn a bit about devstack which was fun : ) [15:00] :) [15:02] smb: blah my vm got filled up... need to fix that [15:02] maybe that's why it never reproduced [15:03] ogasawara, such charts ... so pretty.. https://wiki.ubuntu.com/Kernel/Support https://imgflip.com/i/6p4hs [15:05] chiluk, makes it pretty clean, huh ? [15:05] clear* [15:05] yeah ogasawara did an awesome job. [15:06] chiluk, you should try her pizza ! [15:06] heh [15:07] chiluk: hehe, "very schedule", "so color" [15:07] ogra_, that was _my_ pizza by the way. [15:08] ogasawara, glad I could make you smile [15:08] :D [15:08] rtg: isn't that was great managers do, take all the credit :) [15:08] now we just need a translation into japanese [15:16] jjohansen, https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1270215/comments/33 [15:16] Launchpad bug 1270215 in linux (Ubuntu Precise) "kernel 3.13.0-4.19~precise1-generic: no internet via ethernet or WiFi" [High,Confirmed] [15:17] rtg: is he using the lts-backport kernel? [15:17] jjohansen, indeed [15:18] okay, I'll look. I did test that it was working before I sent you a pull request [15:18] DW-10297, apw: pretty sure http://us.archive.ubuntu.com/ubuntu/dists/precise-updates/main/installer-amd64/current/images/saucy-netboot/ is what you actually want [15:18] jjohansen, I checked that the config option is 'no' and that he has the right kernel [15:19] rtg: okay, I'm installing it [15:20] that should be the enablement stack netboot for 12.04.4 there [15:20] i've never used it but thats what the release notes say at least [15:21] Sarvatt, but that will give you a saucy user space [15:21] Sarvatt, oh, nm [15:21] rtg: it's in precise, they refresh them for the hwe kernels [15:21] that'd be nice to have on http://cdimage.ubuntu.com/netboot/ eh? [15:22] Sarvatt, right. [15:22] I knew that stuff had to bethere somewhere [15:40] rtg: bjf: the kernel panic that we saw with trusty kernel in our systems is not occurring with 3.13.0-7.26 [15:40] psivaa, cool, thanks for the note [15:40] jjohansen, ^^ [15:41] https://jenkins.qa.ubuntu.com/view/Trusty/view/SRU/ (the green balls knowingly lie because there are test failures in the console logs) [16:42] rtg, hi :), are you aware of the broken headers package of the 3.13 precise kernel builds? "linux-headers-3.13.0-7_3.13.0-7.26~precise1_all.deb (5.2 KiB)" [16:43] ricotz, broken how ? [16:43] like being empty ;) [16:44] checking... [16:44] iirc it has been for all 3.13 builds for precise so far [16:44] referring too https://launchpad.net/~canonical-kernel-team/+archive/ppa/+packages [16:44] I could swear I checked them once. [16:45] indeed, totally vacant. [16:47] rtg, thanks for looking [16:48] smb: do you know if it's possible to get crashdumps out of ec2? [16:49] sforshee, yes ... no [16:50] smb: that's too bad [17:10] ricotz, fixed the headers problem. next upload ought to correct it. [17:11] guess I'd better check foranything else I've missed [17:13] rtg, great :), may i ask when this upload will happen? [17:13] ricotz, I imagine 3.13.2 will publish today, so we'll likely upload tomorrow [17:13] rtg, alright, thanks! [17:22] apw, I suspect I don't need do_tools_hyperv=true for armhf, right ? (Trusty LTS) [17:24] rtg, right amd64 only, they don't do 32 bit [17:32] apw, tilting at lttng in trusty [17:34] rtg, heh good luck with that :) [17:35] apw, I had the most problems with armhf last time. something about the compiler version that it hated. [17:35] deep joy indeed [17:39] rsalveti, ok the updated -mako is built in the ckt ppa [17:40] rsalveti, security would like this added to all your 3.4 kernels, so i would like to spin them all, presumably to the PPA as well? [17:45] apw: yeah [17:46] rsalveti, ok cool, i'll let you know when they are in there [17:46] apw: great, thanks [17:49] bjf, am i right in thinking that we have stables outstanding on saucy and raring? that we didn't respin with those in [17:50] apw, correct, i only respun with the upstream fix for the critical CVE [17:50] bjf, ta, trying to make sense of the inbalance in the cve tracker [17:51] apw, is LP all better today? [17:51] bjf, i'd say i have had a regular level timeouts again, so i guess its "ok" [17:51] ack [18:07] anyone here well-versed in the memory cgroup's memory.swappiness contraints? [18:08] Documentation/cgroups/memory.txt says you can't set it if you have use_hierarchy set and (a) you're not root or (b) you have children. Seems like that pretty much means you can't set it if you're using hierarchy, at all. [18:09] ok code confirms it... nm i guess [18:11] cking, just had a crash due to overheat on my x201. I'm not sure if thermd had crashed or not though. I'll apply the latest updates and keep an eye on it. [18:11] jsalisbury, i'll email you some more notes on how to tweak this [18:11] hm, thermd? maybe something i could use [18:11] cking, cool, thanks [18:12] heh, cool, no pun intended ;-) [18:13] hopefully it will be cool(er) [20:10] infinity: saucy uploaded, and awaiting build [20:11] zequence: \o/ [20:11] zequence: Let me score those up so we can get some results. [20:48] rtg: hey === zz_mwhudson is now known as mwhudson === chrisccoulson_ is now known as chrisccoulson [22:29] arges, yo