Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster, 2013-12-24

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

All times shown according to UTC.

Time Nick Message
00:12 itch_ joined #gluster
00:13 dreadnoughtgod joined #gluster
00:14 RicardoSSP joined #gluster
01:18 mattappe_ joined #gluster
01:45 bala joined #gluster
01:53 k4nar joined #gluster
01:54 TonySplitBrain joined #gluster
01:56 k4nar_ joined #gluster
02:02 TonySplitBrain Hello! I have split-brain, can someone help me?
02:04 TonySplitBrain I'm first time on IRC ever.. Don't know how to ask the right way :-(
02:04 k4nar joined #gluster
02:08 TonySplitBrain My `gluster volume heal data info split-brain` gives this: http://pastie.org/private/g68macqbn8amayji8ioiq
02:08 glusterbot Title: Private Paste - Pastie (at pastie.org)
02:10 TonySplitBrain I'm trying to use healer.py script, but it says 'Client volume not found'
02:16 TonySplitBrain it's replicated volume with two bricks
02:18 TonySplitBrain Now i'm backing up files on the volume, and then will stop gluster server on one node and will completely delete brick dir on that node.
02:19 TonySplitBrain Maybe there are some more gently and accurate ways to troubleshoot splitbrain of the root dir?
02:21 TonySplitBrain All i need is to say to gluster 'this node is right, use it as master white resolving this conflict.
02:22 k4nar_ joined #gluster
02:24 k4nar joined #gluster
02:58 kshlm joined #gluster
03:07 flrichar joined #gluster
03:32 mattappe_ joined #gluster
03:34 RameshN joined #gluster
03:35 mattapperson joined #gluster
03:40 psyl0n joined #gluster
03:45 kanagaraj joined #gluster
03:48 shubhendu joined #gluster
03:50 prasanth_ joined #gluster
04:04 itisravi joined #gluster
04:07 davinder joined #gluster
04:13 JoeJulian TonySplitBrain: When I've gotten that, since it's a directory, I just choose one single brick to leave alone and change each trusted.afr.* xattr to 0x0 on all the others.
04:20 TonySplitBrain So the brick with trusted.afr... attrs set will become the master copy?
04:20 TonySplitBrain Do I need to stop gluster server(s) before/after?
04:21 TonySplitBrain What action sequence do you recommend?
04:21 TonySplitBrain And Joe, BTW thanks for your answer!
04:22 JoeJulian The non-zero will become the "master" copy and replicate any changes that might be to the other bricks.
04:23 JoeJulian No stopping /should/ be necessary (I never do).
04:23 thogue joined #gluster
04:23 TonySplitBrain Maybe you have some helper sctipt?
04:24 dusmant joined #gluster
04:25 JoeJulian I don't, sorry.
04:25 TonySplitBrain I think I need something like 'find /brick -exec setfattr blabla' ... Is it right direction?
04:25 ngoswami joined #gluster
04:26 JoeJulian I just "getfattr -m trusted.afr $brick" to get which xattrs it has, then "setfattr -n $afrxattr -v 0x0 $brick"
04:26 JoeJulian for each trusted.afr xattr.
04:27 JoeJulian Since I do it so seldom, and have so few bricks to do it on, it's always been an occasion where it would take longer to script it than to just do it.
04:27 TonySplitBrain Should I do it only on /brick dir, or on every dir/file recursively?
04:27 JoeJulian Looks (from your paste) like just the brick directory should be sufficient.
04:29 TonySplitBrain There are actually some other dirs, also repeated a couple of times each. I just don't care of them, because thay are replaceble.
04:30 JoeJulian Well, each directory that shows as split-brain could be handled that way.
04:31 DV__ joined #gluster
04:36 dusmant joined #gluster
04:52 ababu joined #gluster
04:54 vpshastry joined #gluster
04:55 bala joined #gluster
04:57 TonySplitBrain_ joined #gluster
04:58 MiteshShah joined #gluster
04:58 vikhyat joined #gluster
04:59 ndarshan joined #gluster
05:00 vpshastry joined #gluster
05:10 psharma joined #gluster
05:11 aravindavk joined #gluster
05:18 flakrat joined #gluster
05:19 davinder2 joined #gluster
05:21 mattapperson joined #gluster
05:25 MiteshShah joined #gluster
05:38 shylesh joined #gluster
05:39 lalatenduM joined #gluster
05:45 lalatenduM joined #gluster
05:45 CheRi joined #gluster
05:52 aravindavk joined #gluster
05:59 saurabh joined #gluster
06:02 dreadnoughtgod joined #gluster
06:04 hagarth joined #gluster
06:04 rich joined #gluster
06:06 itch_ joined #gluster
06:07 kshlm joined #gluster
06:12 satheesh joined #gluster
06:12 aravindavk joined #gluster
06:12 psyl0n joined #gluster
06:13 shubhendu joined #gluster
06:25 nshaikh joined #gluster
06:28 mohankumar joined #gluster
06:30 askb joined #gluster
06:34 hagarth joined #gluster
06:34 itisravi_ joined #gluster
06:36 ndarshan joined #gluster
06:37 itisravi_ joined #gluster
06:44 vimal joined #gluster
06:45 shubhendu joined #gluster
06:49 jbautista- joined #gluster
07:02 mtanner_ joined #gluster
07:04 hagarth1 joined #gluster
07:05 anands joined #gluster
07:06 rastar joined #gluster
07:09 itisravi_ joined #gluster
07:11 ababu joined #gluster
07:17 ngoswami joined #gluster
07:19 ricky-ti1 joined #gluster
07:22 itisravi joined #gluster
07:23 itisravi_ joined #gluster
07:36 ekuric joined #gluster
07:41 davinder joined #gluster
07:44 hflai joined #gluster
07:57 ctria joined #gluster
08:25 shylesh joined #gluster
08:40 ntenisOT joined #gluster
08:41 ntenisOT hello all.. I have installed glustefs 3.4.1 on Ubuntu 12.04 and I have the following problem : http://pastebin.com/vpEqrEMd . Can you please help me? I have already tried https://lists.gnu.org/archive/html/​gluster-devel/2012-12/msg00031.html
08:41 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.
08:42 vpshastry joined #gluster
08:42 ntenisOT hello all.. I have installed glustefs 3.4.1 on Ubuntu 12.04 and I have the following problem : http://paste.ubuntu.com/6628011/ . Can you please help me? I have already tried https://lists.gnu.org/archive/html/​gluster-devel/2012-12/msg00031.html
08:42 glusterbot Title: Ubuntu Pastebin (at paste.ubuntu.com)
08:45 psharma joined #gluster
08:53 meghanam joined #gluster
08:54 ricky-ticky joined #gluster
08:55 zwu joined #gluster
08:56 samppah ntenisOT: are you using glusterfs with qemu?
08:57 nshaikh joined #gluster
08:57 ntenisOT @samppah em.. I have a clean Ubuntu Server installation on VCenter, not qemu
08:58 samppah ntenisOT: ok, that log is from client side, right?
08:59 ntenisOT yes.. when I try: sudo mount -t glusterfs dfs1.rayzit.com:/dfs /mnt/glusterfs
09:01 samppah ntenisOT: can you send output from server of following commands: gluster vol info and gluster vol status
09:04 rastar joined #gluster
09:06 ababu joined #gluster
09:06 glusterbot New news from resolvedglusterbugs: [Bug 834729] gluster volume remove-brick defaults to force commit, and causes data loss <https://bugzilla.redhat.com/show_bug.cgi?id=834729>
09:08 hagarth1 joined #gluster
09:08 ntenisOT @samppah: Please check http://paste.ubuntu.com/6628117/
09:09 ntenisOT @samppah: Do I have to use sudo mount -t glusterfs dfs1.rayzit.com:/data /mnt/glusterfs or sudo mount -t glusterfs dfs1.rayzit.com:/dfs /mnt/glusterfs
09:10 ntenisOT @samppah: sudo mount -t glusterfs dfs1.rayzit.com:/dfs /mnt/glusterfs does not present any error but I cannot access my files..
09:13 samppah ntenisOT: dfs1.rayzit.com:/dfs is correct as dfs is name of the volume
09:13 samppah ntenisOT: is client ip listed in auth.allow: 194.42.17.181,194.42.17.171,194.42.17.172 ?
09:13 ntenisOT yes..
09:13 samppah is there a firewall between those machines? hardware or iptables or something else?
09:14 samppah ntenisOT: oh
09:14 ntenisOT I have add iptables filters..
09:14 ntenisOT dfs1.rayzit.com:/dfs on /mnt/glusterfs type fuse.glusterfs (rw,default_permissions,al​low_other,max_read=131072)
09:14 samppah gluster vol info shows that status is created.. it should be started.. can you try to gluster vol start dfs
09:15 samppah hmm
09:15 ntenisOT ls -l /mnt/glusterfs total 0
09:16 ntenisOT I have created a file to the /data directory of the servers..
09:16 ntenisOT do I have to recreate it from the client in order to be able to see it?
09:17 samppah yes, you have to access it through glusterfs mount point
09:17 samppah reading directly from bricks is fine, but writing should happen through glusterfs mount point only
09:17 ntenisOT @samppah : thanks a lot mate ;) it works now!
09:17 samppah ntenisOT: great :)
09:18 ntenisOT @sampaah: one last question. Is it possible to write as a regular user instead of root?
09:19 samppah ntenisOT: yes that's possible.. you should be able to chmod glusterfs to what's needed
09:19 ntenisOT @samppah: can you give me an example? or a document to read about this?
09:20 samppah or use gluster vol set dfs storage.owner-uid 123 to set wanted uid for it
09:23 samppah ntenisOT: if you have gluster mounted at /mnt/glusterfs you should be able to create directory /mnt/glusterfs/user and then chown user /mnt/glusterfs/user etc.
09:24 ntenisOT thanks ;)
09:24 ntenisOT I found it
09:24 ntenisOT @sampaah: thanks once again ;)
09:24 samppah ntenisOT: no problem :)
09:29 CheRi joined #gluster
09:36 lalatenduM joined #gluster
09:37 atrius joined #gluster
09:48 rastar joined #gluster
10:00 rotbeard joined #gluster
10:35 hagarth joined #gluster
10:36 RameshN joined #gluster
10:53 MiteshShah joined #gluster
11:19 itch_ joined #gluster
11:20 sticky_afk joined #gluster
11:21 stickyboy joined #gluster
11:27 meghanam joined #gluster
11:28 TonySplitBrain joined #gluster
11:33 TonySplitBrain hello dear all... is it possible to clean the list that 'gluster volume heal VOLUME info healed' (or '... split-brain') gives me?
11:43 rastar joined #gluster
11:52 TonySplitBrain in other words: i did what JoeJulian told me here (thanks); so how could i know that i solved all splitbrains and the volume is in ordnung?
12:05 satheesh joined #gluster
12:39 vikhyat TonySplitBrain: once you have solved the problem, 'gluster volume heal VOLUME info healed' (or '... split-brain') will not give you any files in its output
12:53 TonySplitBrain vikhyat: but what mean multiple lines for same dir in 'gluster volume heal data info split-brain' output?
12:54 TonySplitBrain i got 150 records for just two dirs
12:55 satheesh joined #gluster
12:56 TonySplitBrain is it mean that i got 150 different problems with that dirs? 8-0
12:56 TonySplitBrain i still can't get any logig in gluster 'heal' output..
12:59 vpshastry left #gluster
13:00 vikhyat TonySplitBrain: yup may be !
13:00 vikhyat TonySplitBrain: http://review.gluster.org/​#/c/4216/1/doc/split-brain.txt
13:00 glusterbot Title: Gerrit Code Review (at review.gluster.org)
13:00 vikhyat TonySplitBrain: you can take help of this doc !
13:01 vikhyat TonySplitBrain: I am sure it will help you !
13:02 RameshN joined #gluster
13:12 satheesh joined #gluster
13:12 TonySplitBrain vikhyat: thanks.. sure, it will help, when i will understand all the data-metadata-xattrs stuff
13:16 TonySplitBrain sorry to say, but now i understand that Gluster is defintly not production ready
13:17 TonySplitBrain it lacks essential docs and tools, or they are buried deep
13:18 TonySplitBrain vikhyat: without your link, i was unable to find this info and 'bless' method (fill try it now)
13:25 TonySplitBrain vikhyat: i'm trying to use this method http://review.gluster.org/#/c/4132/ , but getting "No such attribute" error :-(
13:25 glusterbot Title: Gerrit Code Review (at review.gluster.org)
13:29 bala joined #gluster
13:34 TonySplitBrain i'm using Gluster 3.4.1-2 on Debian. Is it possible that my version does not supports such 'bless' method? How to know it?
13:35 aravindavk joined #gluster
13:36 sroy_ joined #gluster
13:53 davinder joined #gluster
13:54 gluslog_ joined #gluster
13:57 mkzero_ joined #gluster
13:57 pat__ joined #gluster
13:57 wgao__ joined #gluster
13:58 klaxa_ joined #gluster
13:58 tru_tru_ joined #gluster
14:04 B21956 joined #gluster
14:10 jiphex_ joined #gluster
14:10 al joined #gluster
14:17 tru_tru joined #gluster
14:28 getup- joined #gluster
14:31 CheRi joined #gluster
14:37 zwu joined #gluster
14:55 XpineX_ joined #gluster
15:06 anands joined #gluster
15:20 itch_ joined #gluster
15:27 vpshastry joined #gluster
15:29 anands joined #gluster
15:29 vpshastry left #gluster
15:30 XpineX__ joined #gluster
15:44 eclectic joined #gluster
15:48 zerick joined #gluster
15:48 zapotah joined #gluster
15:48 rPawel joined #gluster
15:50 rPawel hi chaps, any chance someone could tell me if it is ok to mix gluster server 32 with 64 bit versions ?
15:51 jag3773 joined #gluster
15:51 XpineX_ joined #gluster
15:57 XpineX__ joined #gluster
16:06 ppai joined #gluster
16:12 XpineX_ joined #gluster
16:16 zapotah joined #gluster
16:19 diegows joined #gluster
16:24 zapotah joined #gluster
16:25 itch_ joined #gluster
16:28 ron-slc joined #gluster
16:33 zerick joined #gluster
16:38 XpineX__ joined #gluster
16:42 johnbot11 joined #gluster
17:03 Mo__ joined #gluster
17:18 mattappe_ joined #gluster
17:22 jag3773 joined #gluster
17:24 sticky_afk joined #gluster
17:25 stickyboy joined #gluster
17:43 mattappe_ joined #gluster
17:44 mattappe_ joined #gluster
17:45 XpineX_ joined #gluster
17:45 hagarth joined #gluster
17:52 davinder joined #gluster
18:00 mattappe_ joined #gluster
18:03 hagarth joined #gluster
18:04 mattapperson joined #gluster
18:46 JoeJulian TonySplitBrain: restart all glusterd to clear the split-brain log.
19:09 davinder joined #gluster
19:11 TonySplitBrain JoeJulian: thanks, but i moved real data from gluster now. Needs much more testing to go to production.
19:11 TonySplitBrain I'll sleep and than play with brains on testing cluster.
20:00 TonySplitBrain joined #gluster
20:20 zapotah joined #gluster
20:20 zapotah joined #gluster
20:24 TonySplitBrain dear all, is it possible to specify alternative ip / hostname for node? it could be reached on multiple interfaces
20:51 RedShift joined #gluster
20:52 TonySplitBrain oh, i found Justin's proposal for connection groups, that's it
21:19 calum_ joined #gluster
21:23 daMaestro joined #gluster
21:31 rotbeard joined #gluster
21:43 fidevo joined #gluster
22:24 blook joined #gluster

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