=== kallesbar_ is now known as kallesbar | ||
mattgphoto | Hello everyone! I've got a bit of an issue and need to know how to go about troubleshooting it, was hoping someone here could help. How do I determine the cause of a Call Trace? | 17:29 |
---|---|---|
mattgphoto | For example: What is causing this call trace to happen when my Ubuntu VM is under load/transferring data between samba shares (as client): https://i.imgur.com/sBNZ2My.png | 17:29 |
mattgphoto | This is on 18.04.1 server btw | 17:29 |
TJ- | mattgphoto: which kernel "uname -r" ? | 17:30 |
mattgphoto | One second, lemme shut it off, power back on :-) | 17:31 |
mattgphoto | ~$ uname -r | 17:32 |
mattgphoto | 4.15.0-39-generic | 17:32 |
TJ- | mattgphoto: I see that bug for v4.14 through v4.17 reported for other distros, so it looks like a kernel regression | 17:34 |
TJ- | mattgphoto: I'd recommend upgrading to the v4.18 HWE kernel (linux-image-generic-hwe-18.04-edge) | 17:35 |
lotuspsychje | old, but perhaps related? https://bugs.launchpad.net/ubuntu/+source/gvfs/+bug/1075923 | 17:35 |
ubottu | Launchpad bug 1075923 in gvfs (Ubuntu Precise) "nautilus hangs copying large directories from a samba share" [High,Confirmed] | 17:35 |
mattgphoto | @TJ-, how do I do that? | 18:01 |
mattgphoto | hrm, I see the instructions for 16.04... https://wiki.ubuntu.com/Kernel/LTSEnablementStack | 18:02 |
TJ- | mattgphoto: "apt list linux-image-generic-hwe-18.04-edge" - if you see it, then "apt install ..." | 18:05 |
mattgphoto | Gotcha. :-) @TJ- Thank you! Here's hoping this fixes it! | 18:06 |
mattgphoto | Um, nothing's listed. | 18:06 |
mattgphoto | It just returns: 'Listing... Done'. And returns to command prompt | 18:07 |
TJ- | mattgphoto: OK, so you need to enable the component it is in | 18:07 |
TJ- | right, it's in bionic-proposed so you need to enable that in apt's sources | 18:08 |
TJ- | mattgphoto: "echo 'deb http://gb.archive.ubuntu.com/ubuntu bionic-proposed main' | sudo dd of=/etc/apt/sources.list.d/bionic-proposed.list" then "sudo apt update" | 18:09 |
mattgphoto | so just `sudo add-apt-repository bionic-proposed`? | 18:09 |
mattgphoto | ah | 18:09 |
TJ- | mattgphoto: maybe you don;t want the "gb" mirror though :) | 18:09 |
mattgphoto | oh, I'll assume us... one sec | 18:10 |
mattgphoto | hrm, it'd be available on any mirror? | 18:11 |
mattgphoto | oh, just archive works too | 18:13 |
mattgphoto | So pardon if I ask a bunch of questions here, the release of ubuntu versions isn't something I'm quite familiar with. Is the HWE kernel something you just stay on until next point release? or? | 18:16 |
TJ- | mattgphoto: the idea is that newer hardware, better performance, and so on can be supported on older releases. As 18.04 is now to be supported for 10 years we wouldn't expect to remain on the 4.15 kernel :) | 18:41 |
mattgphoto | gotcha | 18:41 |
TJ- | mattgphoto: I'd hope a stable release update to the mainline kernel might also bring in the fix for your issue to the 4.15 kernel at some point | 18:41 |
mattgphoto | gotcha | 18:42 |
mattgphoto | we'll see | 18:42 |
TJ- | mattgphoto: we'd have to figure out which commit the fix was first though; I couldn't find anything so it may have been fixed incidentally in other code | 18:42 |
TJ- | mattgphoto: the other reports I found said v4.18 had solved it | 18:42 |
mattgphoto | gotcha | 19:02 |
mattgphoto | so basically a normal 'apt-get dist-upgrade' would do it once it's figured out? | 19:02 |
WBILL | shouldn't this on eliner in my crontab work to just reboot my server | 19:48 |
WBILL | <WBILL> < 52 13 * * * root shutdown -r now > | 19:48 |
WBILL | <WBILL> minu the <>symbols of course | 19:48 |
blackflow | WBILL: might need full path to shutdown | 20:35 |
mattgphoto | New issue! heh. How would I go about troubleshooting this? https://i.imgur.com/Web3Xh4.png | 21:38 |
mattgphoto | (I find some humor in the jiffies term btw) | 21:39 |
mattgphoto | This just seems like an amalgamation of problems now :-/ | 21:41 |
mybalzitch | I wonder if its io related | 21:50 |
mattgphoto | I don't wonder myself | 21:53 |
mattgphoto | I tried having this VM in it's own iSCSI LUN on ESXi, and ran into the previous problem... | 21:54 |
mattgphoto | now it's in a NFS Store... | 21:54 |
mattgphoto | (There isn't enough space on this server to house the TB's of storage heh) | 21:56 |
mattgphoto | although | 21:57 |
mattgphoto | I guess I could put just the VM itself on the local storage to the VM | 21:57 |
mattgphoto | woah damn https://i.imgur.com/UZmRhDH.png | 22:08 |
mybalzitch | dat iowait | 22:09 |
mattgphoto | https://i.imgur.com/WJTgBiJ.png | 22:10 |
mattgphoto | how can you tell iowait in that screenshot? | 22:10 |
mybalzitch | that would be my first guess as to why your load average is past 60 on a 4 core system | 22:11 |
mattgphoto | oh gotcha heh | 22:11 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!