Camelia, the Perl 6 bug

IRC log for #gluster, 2013-03-31

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

All times shown according to UTC.

Time Nick Message
00:00 lh joined #gluster
00:02 joehoyle joined #gluster
00:07 lh joined #gluster
00:30 lh joined #gluster
00:39 threesome At http://joejulian.name/blog/glust‚Äčerfs-replication-dos-and-donts/, what does the following mean: "GlusterFS is latency dependent. Since self-heal checks are done when establishing the FD and the client connects to all the servers in the volume simultaneously, high latency (mult-zone) replication is not normally advisable. Each lookup will query both sides of the replica. A simple directory listing of 100 files across a 200ms co
00:39 threesome nnection would require 400 round trips totaling 80 seconds. A single drupal page could take around 20 minutes." When reading a file it needs to read all the replicas rather than just one?
00:40 glusterbot <http://goo.gl/JmECY> (at joejulian.name)
01:04 _br_ joined #gluster
01:04 lh joined #gluster
01:05 premera yeah, looks this way, it reads metadata from all the replicas first
01:10 premera in addition, according to that article, metadata reads are done one after another (not in parallel), so the more replicas you have the longer the operation is going to take
01:11 _br_ joined #gluster
01:31 Rydekull joined #gluster
03:02 rastar joined #gluster
04:00 premera joined #gluster
04:41 premera left #gluster
05:01 _br_ joined #gluster
05:09 _br_ joined #gluster
05:16 _br_ joined #gluster
05:31 vshankar joined #gluster
08:00 lalatenduM joined #gluster
08:24 juhaj Ok, I got my fs fixed, but now no client can write to it, except those on the hosts also hosting the bricks!
08:26 samppah are they able to mount it?
08:33 rosmo #gluster.fi? :)
08:33 kevein joined #gluster
08:37 sohoo joined #gluster
08:39 sohoo is it possible to mix replica pairs volume space, for example 2 bricks with 2 TB eaqch in replica mode and another server with 2 replica bricks of 1TB each and how will it affect the load
08:42 joehoyle joined #gluster
08:46 samppah rosmo ;)
08:47 avati joined #gluster
08:48 a2_ joined #gluster
09:04 rosmo samppah: noh, perustin sen :-)
09:19 ricky-ticky joined #gluster
09:22 shylesh joined #gluster
09:28 lalatenduM joined #gluster
09:46 camel1cz joined #gluster
09:51 camel1cz1 joined #gluster
09:58 hateya joined #gluster
10:05 shylesh joined #gluster
10:08 lh joined #gluster
11:00 GLHMarmot joined #gluster
11:35 Slydder joined #gluster
11:36 Slydder hey all. anyone around today?
13:01 juhaj They're able to mount it, but then "[fuse-bridge.c:1678:fuse_create_cbk] 0-glusterfs-fuse: 4: /rigelroot => -1 (Read-only file system)"
13:01 juhaj when I try to write
13:10 lh joined #gluster
13:12 hateya joined #gluster
13:27 yinyin joined #gluster
13:37 pmuller_ joined #gluster
13:51 juhaj And on the server side, I get "W [socket.c:1494:__socket_proto_state_machine] 0-socket.management: reading from socket failed. Error (Transport endpoint is not connected), peer (10.0.0.2:1021)"
13:51 glusterbot juhaj: That's just a spurious message which can be safely ignored.
13:52 juhaj I distrust glusterbot here, as these two occur at precisely the same moment
13:57 juhaj Ok, got one step further, for some reason "mount -o rw ..." calls "/usr/sbin/glusterfs --log-level=INFO --read-only"
13:57 juhaj No wonder my mount is read-only... But why would -o rw call --read-only?
14:06 juhaj Ok, so after making sure it's not mounted --read-only, I get to a point where root and uid-owner can write, members of the owning group cannot (except the user who also owns the file)
14:07 juhaj Another member of the owning group cannot even read files (I made everything 770 there: if I allow world to access, then another user can access, too)
14:07 juhaj So there's something wrong with group member access
14:39 camel1cz joined #gluster
14:39 camel1cz left #gluster
14:51 joe joined #gluster
14:55 piotrektt_ joined #gluster
15:14 _br_ joined #gluster
15:21 _br_ joined #gluster
15:26 _br_ joined #gluster
15:29 juhaj Oh, and from the group members can access the files, but only from the servers that have the bricks
15:31 _br_ joined #gluster
16:10 lh joined #gluster
16:25 joe joined #gluster
16:26 _br_ joined #gluster
16:29 raghug joined #gluster
16:30 _br_ joined #gluster
16:31 juhaj Hm? Anybody ideas?
16:42 lh joined #gluster
17:04 lh joined #gluster
17:33 andreask joined #gluster
17:36 raghug joined #gluster
17:39 lh joined #gluster
18:00 lh joined #gluster
18:44 joehoyle joined #gluster
18:57 premera joined #gluster
19:16 rosmo joined #gluster
19:34 RobertLaptop joined #gluster
19:38 premera joined #gluster
19:47 johnmorr joined #gluster
20:54 ricky-ticky joined #gluster
21:45 _br_ joined #gluster
21:52 _br_ joined #gluster
21:58 _br_ joined #gluster
21:59 lh joined #gluster
22:02 _br_ joined #gluster
22:02 _br_ joined #gluster
22:29 sohoo joined #gluster
22:33 sohoo im quite disapointed with gluster self heal, i hear about the improvemnts off evry version but i cannot reboot without issues(real issue) the main probelm i guess is that file created while replica was down then come back up, self replication doesnt happne or doesnt happen for long time to accept, clients machines keep requesting data from the live node again but the files are not there thus
22:33 sohoo getting tons of errors, untill complete filesystem manuala data truncat made on the failed node then its starting to replicate all back. but its crazy it happen on every reboot. whay kind of self heal is it?
22:34 sohoo can anybody explain some more on the self-heal after replica set reboot? the docs say its perfect :)
22:36 sohoo over all gluster do what its meant to but the healing procces that come after each reboot is realy crazy
22:36 sohoo anybody?
22:37 sohoo :)
22:38 sohoo compile debug version?
22:38 sohoo no tools?
22:40 sohoo we are willing to make the tools and give it to the comunity ofcours but some guidlines will be helpfull seens there is no good documentation
22:42 ChikuLinu__ joined #gluster
22:45 sohoo nobody had this issue? real strange
22:53 mtanner joined #gluster
22:55 tryggvil joined #gluster
22:56 sohoo thanks a lot
23:05 johnmorr joined #gluster
23:08 mtanner joined #gluster
23:40 sohoo joined #gluster
23:44 fidevo joined #gluster
23:55 Cenbe_ joined #gluster

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