Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster, 2014-09-27

| Channels | #gluster index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:09 fubada joined #gluster
00:27 Pupeno joined #gluster
00:59 MacWinner joined #gluster
01:20 n-st joined #gluster
01:30 Alssi_ joined #gluster
03:03 rjoseph joined #gluster
03:09 MacWinner joined #gluster
03:16 bala joined #gluster
03:28 hagarth joined #gluster
04:17 johnmwilliams__ joined #gluster
04:27 fyxim__ joined #gluster
04:28 suliba_ joined #gluster
04:33 frankS2 joined #gluster
04:42 lezo__ joined #gluster
04:43 XpineX joined #gluster
04:45 elico joined #gluster
04:57 anoopcs joined #gluster
05:05 anoopcs joined #gluster
05:16 adamdrew joined #gluster
05:31 partner cool, we're still up
05:32 partner 00:55 < partner> 16:23 < partner> upgraded yesterday from 3.3.2 to 3.4.5 and now our logs are flooded with the "disk layout missing" with following "mismatching layout", like this: http://pastie.org/private/48otxyeuda3mfima7jnmq
05:32 glusterbot Title: Private Paste - Pastie (at pastie.org)
05:33 partner nobody has any idea of why?
05:38 partner oh pardon my language, i guess the issues took over me and got frustrated
05:39 partner the issue is currently "solved" by simply changing log level into warning to prevent servers from filling up the /var
05:40 partner also running the fix-layout over the night but it will take probably weeks or months to complete
05:40 partner while i really don't see the need for it, we did not change anything but version
05:45 haomaiwang joined #gluster
07:24 dmachi joined #gluster
07:35 frankS2 joined #gluster
07:56 bjuanico joined #gluster
08:01 bjuanico joined #gluster
08:12 ndevos partner: I'm not sure if missed a step from this blog post? http://vbellur.wordpress.com/2013/​07/15/upgrading-to-glusterfs-3-4/
08:15 ndevos partner: "disk layout missing" suggests that the dht (for distribute) extended attributes are not available for all/some of the directories on the bricks
08:16 ndevos that is strange, because 3.3 uses the same attributes for all I know
08:21 harish joined #gluster
08:22 ndevos partner: oh, the "disk layout missing" can also happen after directories were created directly on the bricks (not through a mountpoint)
08:47 partner bricks not touched straight
08:47 jiffin joined #gluster
09:03 partner afaik the upgrade was made "online", doing apt-get install for all the gluster packages, host by host
09:04 partner hosts provide a mix of distributed and replicated volumes, no dist-repl
09:29 glusterbot New news from resolvedglusterbugs: [Bug 1119827] Brick goes offline unexpectedly <https://bugzilla.redhat.co​m/show_bug.cgi?id=1119827>
10:05 soumya__ joined #gluster
10:21 ekuric joined #gluster
10:35 RobertLaptop joined #gluster
10:38 lanning joined #gluster
10:45 LebedevRI joined #gluster
11:07 azar joined #gluster
11:55 fubada joined #gluster
12:07 diegows joined #gluster
12:12 Alssi_ joined #gluster
12:14 sprachgenerator joined #gluster
12:21 and` joined #gluster
12:25 and` joined #gluster
12:37 ekuric joined #gluster
12:43 milka joined #gluster
12:43 fubada joined #gluster
12:52 chirino joined #gluster
13:04 P0w3r3d joined #gluster
13:25 glusterbot New news from newglusterbugs: [Bug 883785] RFE: Make glusterfs work with FSCache tools <https://bugzilla.redhat.com/show_bug.cgi?id=883785>
13:29 glusterbot New news from resolvedglusterbugs: [Bug 871986] [RFE] Striping should have optional parity/checksum <https://bugzilla.redhat.com/show_bug.cgi?id=871986>
13:34 burn420 joined #gluster
13:53 chirino joined #gluster
13:55 glusterbot New news from newglusterbugs: [Bug 922542] [FEAT] Please add support to replace multiple bricks at a time. <https://bugzilla.redhat.com/show_bug.cgi?id=922542> || [Bug 905747] [FEAT] Tier support for Volumes <https://bugzilla.redhat.com/show_bug.cgi?id=905747> || [Bug 914874] Enhancement suggestions for BitRot hash computation <https://bugzilla.redhat.com/show_bug.cgi?id=914874> || [Bug 928781] hangs when mount a volume at own brick <https://b
14:02 fubada joined #gluster
14:21 chirino joined #gluster
14:51 atrius` joined #gluster
15:00 pkoro joined #gluster
15:03 haomaiwa_ joined #gluster
16:26 glusterbot New news from newglusterbugs: [Bug 1147156] AFR client segmentation fault in afr_priv_destroy <https://bugzilla.redhat.co​m/show_bug.cgi?id=1147156>
16:38 chirino joined #gluster
16:42 elico joined #gluster
16:43 d-fence joined #gluster
17:05 R0ok__50 joined #gluster
17:26 glusterbot New news from newglusterbugs: [Bug 958781] KVM guest I/O errors with xfs backed gluster volumes <https://bugzilla.redhat.com/show_bug.cgi?id=958781>
17:31 abyss^^ joined #gluster
17:31 dtrainor joined #gluster
17:32 dtrainor mornin.  i'm trying to use a gluster filesystem as a filesystem directory for storage using libvirt.  Right when my installer is about to write to that disk, I get a pretty obscure error that I'm not able to decipher.  I've read a few mailing list posts about something similar but not exactly this error - http://pastebin.com/rFVf9Nb4
17:32 dtrainor The brick logs have a very, very similar error
17:34 ndevos dtrainor: maybe www.gluster.org/community/documentatio​n/index.php/Libgfapi_with_qemu_libvirt contains some points that you have missed?
17:34 dtrainor I'll review it, not sure how i missed that!
17:35 dtrainor Oh, right, I'm trying to not use it that way so I don't need to create a new volume
17:35 dtrainor I'm trying to use an existing volume that's already mounted on disk, on the hypervisor
17:36 dtrainor I just.... want to use a directory on disk, as I would anywhere else
17:38 portante joined #gluster
17:38 ndevos dtrainor: okay, I dont think that should cause any issues... do you have corresponding errors in the brick logs? ,,(paste)
17:46 cfeller joined #gluster
17:46 glusterbot joined #gluster
17:46 torbjorn__ joined #gluster
17:46 samppah joined #gluster
17:46 tru_tru joined #gluster
17:46 saltsa joined #gluster
17:46 klaas joined #gluster
17:46 mikedep333 joined #gluster
17:46 Bardack joined #gluster
17:46 coredumb joined #gluster
17:46 stickyboy joined #gluster
17:46 d4nku_ joined #gluster
17:46 Lee- joined #gluster
17:46 edong23_ joined #gluster
17:46 radez_g0n3 joined #gluster
17:46 tobias- joined #gluster
17:46 prasanth|brb joined #gluster
17:46 Guest44047 joined #gluster
17:46 neoice joined #gluster
17:47 dtrainor sure do, they're very similar.  coming up.
17:47 dtrainor http://pastebin.com/eZEddJxc
17:47 diegows joined #gluster
17:48 dtrainor i see rpc errors, i do have insecure = on
17:48 glusterbot Please use http://fpaste.org or http://paste.ubuntu.com/ . pb has too many ads. Say @paste in channel for info about paste utils.
17:48 dtrainor Noted
17:49 dtrainor selinux in permissive, too
17:50 ndevos dtrainor: nothing in the logs make it clear to me... what version are you running?
17:50 dtrainor 3.5.1
17:55 dtrainor i have like a million updates pending on that host, let me update it and try this again
17:55 dtrainor maybe there's a libvirt update
17:56 glusterbot New news from newglusterbugs: [Bug 976129] Upstream Gluster could do with a decent Feature Matrix on the website <https://bugzilla.redhat.com/show_bug.cgi?id=976129> || [Bug 988367] 32 bit Applications not working properly despite mount option enable-ino32 <https://bugzilla.redhat.com/show_bug.cgi?id=988367> || [Bug 985390] Posix ACL not supported via fuse-client <https://bugzilla.redhat.com/show_bug.cgi?id=985390>
17:57 ndevos dtrainor: if an update does not help, you could try to disable performance.write-behind
17:57 dtrainor hmm, how do you figure that'd help? did you see any clues in the logs?
17:58 ndevos well, write-behind had an issue with sparse files and extending writes, that would result in a write (or read) after the End-Of-File -> EINVAL
18:00 ndevos at least one of those bugs should be fixed in 3.5.1, but there may be some other corner cases that have a problem (bug 1058663)
18:00 glusterbot Bug https://bugzilla.redhat.com:​443/show_bug.cgi?id=1058663 high, medium, ---, ndevos, CLOSED CURRENTRELEASE, Sporadic SIGBUS with mmap() on a sparse file created with open(), seek(), write()
18:00 ndevos well, mmap() would get a SIGBUS, but the underlaying error would be a EINVAL on read/write
18:00 dtrainor hah!
18:01 dtrainor well. this is a great time for gluster to do a rebalance so...
18:02 glusterbot New news from resolvedglusterbugs: [Bug 985085] Install virt.group file into /var/lib/glusterd/groups/ <https://bugzilla.redhat.com/show_bug.cgi?id=985085>
18:04 samkottler joined #gluster
18:04 samkottler joined #gluster
18:04 chirino joined #gluster
18:06 dtrainor disabling performance.write-behind didn't work
18:09 ndevos could you check the size of the file on the brick?
18:10 ndevos the logs from the brick mention "write failed: offset", is that offset higher than the size of the file?
18:13 dtrainor oh.  one sec.
18:13 dtrainor it's a 10g file, eventually going to have a few 100G+ files in there, of VMs
18:14 dtrainor no, the offsets are smaller
18:14 n-st joined #gluster
18:16 lalatenduM joined #gluster
18:20 dtrainor gonna reboot this fileserver, let's hope it comes back :)
18:23 ndevos well, maybe you have the same issue as this one: https://bugzilla.redhat.com​/show_bug.cgi?id=958781#c26
18:23 glusterbot Bug 958781: medium, unspecified, ---, gluster-bugs, NEW , KVM guest I/O errors with xfs backed gluster volumes
18:24 ndevos comment #24 might give you some ideas that you could check, https://bugzilla.redhat.com​/show_bug.cgi?id=958781#c24
18:24 glusterbot Bug 958781: medium, unspecified, ---, gluster-bugs, NEW , KVM guest I/O errors with xfs backed gluster volumes
18:25 ndevos I am not sure if the filesystem sector size really makes that difference, but it could be related to the alignment of the read/writes
18:26 ndevos if you think you hit the same issue, feel free to post the related information in the bug report
18:27 dtrainor will do
18:27 dtrainor i don't know how i missed that bug either.  i promise, i did my research :)
18:29 dtrainor yeah, i can do that
18:30 ndevos well, there are way too many bugs, I wont blame you on missing any of those :)
18:32 dtrainor haha
18:34 ndevos also, you may be interested in http://review.gluster.org/4232 , current versions already have the 'virt' group file, you should only need to execute the command to enable it for the volume
18:34 glusterbot Title: Gerrit Code Review (at review.gluster.org)
18:35 ndevos thats it for today from me, I'll be gone now, good luck!
18:43 sijis joined #gluster
18:48 dtrainor thanks for the pointers ndevos
18:48 dtrainor i need to go play outside, i'll update the bug with my 3.5.1 results later.
19:00 fattaneh1 joined #gluster
19:02 glusterbot New news from resolvedglusterbugs: [Bug 1006698] glfs_preadv_async fails for large read with io-cache xlator on <https://bugzilla.redhat.co​m/show_bug.cgi?id=1006698> || [Bug 851237] [FEAT] Add gluster cli interface for setting a 'group' of options in one go. <https://bugzilla.redhat.com/show_bug.cgi?id=851237>
19:13 jvdm i'm trying geo-rep between two hosts, but i need to put gluster inside a chroot (debian wheezy in this case); the chroots are up, but geo-rep volume can't start, since the ssh server running on the nodes are outside the chroot -- does anyone have a recommendation or links to relevant documentation?
19:14 jvdm or, anyone with a similar setup to provide me some kick-off steps?
19:17 jvdm oh, and doing regular clustering (i just did replicas) between the two chroots works.
19:20 rafi1 joined #gluster
19:36 Pupeno joined #gluster
19:57 ThatGraemeGuy joined #gluster
20:02 bjuanico joined #gluster
20:06 RioS2 joined #gluster
20:07 RioS2 joined #gluster
20:20 cyberbootje joined #gluster
20:33 davdunc` joined #gluster
20:34 capri joined #gluster
20:36 JustinCl1ft joined #gluster
20:37 georgeh_ joined #gluster
20:38 kke joined #gluster
20:39 bjuanico joined #gluster
20:40 johndescs joined #gluster
20:40 davdunc`` joined #gluster
20:42 VeggieMeat_ joined #gluster
20:43 nated joined #gluster
20:43 l0uis_ joined #gluster
20:43 NuxRo joined #gluster
20:43 foster_ joined #gluster
20:43 bfoster_ joined #gluster
20:43 oxidane_ joined #gluster
20:44 eryc joined #gluster
20:44 delhage joined #gluster
20:44 partner joined #gluster
20:45 morse joined #gluster
20:46 social joined #gluster
21:22 justinmburrous joined #gluster
21:43 fubada joined #gluster
21:54 wgao_ joined #gluster
21:57 glusterbot New news from newglusterbugs: [Bug 1141379] Geo-Replication - Fails to handle file renaming correctly between master and slave <https://bugzilla.redhat.co​m/show_bug.cgi?id=1141379>
21:59 fubada joined #gluster
22:12 fubada joined #gluster
22:28 fubada joined #gluster
22:36 chirino joined #gluster
22:49 fubada joined #gluster
23:18 fubada joined #gluster
23:19 diegows joined #gluster
23:24 cyberbootje joined #gluster
23:24 chirino joined #gluster
23:32 fubada joined #gluster
23:43 fubada joined #gluster
23:44 coredump joined #gluster
23:59 fubada joined #gluster

| Channels | #gluster index | Today | | Search | Google Search | Plain-Text | summary