Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster, 2015-03-20

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

All times shown according to UTC.

Time Nick Message
00:15 wkf joined #gluster
00:17 rjoseph|afk joined #gluster
00:26 gnudna joined #gluster
00:27 gnudna Hi guys is  there any performance benefits using Libgfapi with qemu libvirt
00:30 gnudna i got a test vm working but i am not sure if there are obvious benefits
00:33 badone__ joined #gluster
00:36 wkf joined #gluster
00:39 RicardoSSP joined #gluster
00:39 RicardoSSP joined #gluster
00:42 wkf joined #gluster
01:01 gnudna left #gluster
01:04 harish_ joined #gluster
01:30 siel joined #gluster
01:30 siel joined #gluster
02:13 bharata-rao joined #gluster
02:13 bharata_ joined #gluster
02:14 _polto_ joined #gluster
02:26 harish joined #gluster
02:35 haomaiwa_ joined #gluster
02:35 cyberbootje joined #gluster
02:36 meghanam joined #gluster
02:43 lyang0 joined #gluster
02:45 sputnik13 joined #gluster
02:45 DV joined #gluster
02:45 Telsin joined #gluster
02:58 hagarth JustinClift: peer probe is owned by glusterd devs
03:07 bharata__ joined #gluster
03:09 gnudna joined #gluster
03:10 coredump joined #gluster
03:11 bharata_ joined #gluster
03:11 gnudna hi guy is it normal that when a replicated volume comes back up due to heal all vm's b
03:12 gnudna become unresponsive ^
03:12 gnudna noticed my test vm with libgfapi did not exhibit this behavior
03:17 chirino joined #gluster
03:20 plarsen joined #gluster
03:27 p0licy joined #gluster
03:32 Apeksha joined #gluster
03:37 sputnik13 joined #gluster
03:54 kumar joined #gluster
03:55 bharata-rao joined #gluster
03:55 sputnik13 joined #gluster
04:03 atinmu joined #gluster
04:04 dusmant joined #gluster
04:05 gnudna left #gluster
04:12 _zerick_ joined #gluster
04:12 punit_ raghu, good morning
04:16 sputnik13 joined #gluster
04:19 kanagaraj joined #gluster
04:20 RameshN joined #gluster
04:23 kdhananjay joined #gluster
04:25 ppai joined #gluster
04:32 ndarshan joined #gluster
04:36 sputnik13 joined #gluster
04:41 badone__ joined #gluster
04:43 gildub joined #gluster
04:46 anoopcs joined #gluster
04:46 rafi joined #gluster
04:46 spandit joined #gluster
04:46 sputnik13 joined #gluster
04:50 schandra joined #gluster
04:53 kshlm joined #gluster
04:55 anrao joined #gluster
04:56 punit_ kshlm,https://www.mail-archive.com/gluster-users@gluster.org/msg19490.html
04:56 ppp joined #gluster
04:57 lalatenduM joined #gluster
05:00 sputnik13 joined #gluster
05:01 meghanam joined #gluster
05:03 sputnik13 joined #gluster
05:03 smohan joined #gluster
05:11 itpings joined #gluster
05:13 lalatenduM joined #gluster
05:15 stickyboy joined #gluster
05:15 Apeksha joined #gluster
05:17 lyang0 joined #gluster
05:22 lalatenduM joined #gluster
05:25 ashiq joined #gluster
05:27 Manikandan joined #gluster
05:27 Manikandan_ joined #gluster
05:33 bharata-rao joined #gluster
05:37 Guest42798 joined #gluster
05:39 topshare joined #gluster
05:40 stickyboy joined #gluster
05:40 aravindavk joined #gluster
05:46 _polto_ joined #gluster
05:47 sputnik13 joined #gluster
05:47 overclk joined #gluster
05:59 XpineX joined #gluster
06:01 Bhaskarakiran joined #gluster
06:02 vimal joined #gluster
06:02 raghu joined #gluster
06:04 raghu punit_: did setting the ping-timeout to 0 and restarting the volume help you? If not, can you provide me the logs?
06:04 punit_ raghu, no it didn't help me
06:04 punit_ raghu,which logs you want ??
06:05 an joined #gluster
06:06 raghu the client logs, vdsm logs, brick logs and glusterd logs
06:07 gem joined #gluster
06:09 punit_ raghu, gluster info : http://paste.ubuntu.com/10632428/
06:11 punit_ raghu, status :-http://paste.ubuntu.com/10632433/
06:11 sputnik13 joined #gluster
06:11 kshlm joined #gluster
06:13 foster joined #gluster
06:14 punit_ raghu, please find the logs here : http://www.filedropper.com/logs_2
06:15 gem joined #gluster
06:16 hgowtham joined #gluster
06:18 ashiq joined #gluster
06:20 hchiramm joined #gluster
06:20 punit_ raghu, i think the disconneting connection from the client is the main issue..which can see in the brick logs
06:21 punit_ raghu, but there is no problem in the network....even the brick diconnect from the localhost such as cpu01 to cpu01 disconnect
06:24 raghu punit_: hmm. The disconnects are the ones which are causing the main problems. Still wondering why the client should disconnect. I thought setting ping-timeout to 0 might help as in that case, client wont be doing any ping to the server. Let me check what is happening
06:24 punit_ raghu, ok thanks
06:25 jackdpeterson joined #gluster
06:25 jackdpeterson Hey all, looking for some documentation on why / where I can locate the 'service glusterd restart' type of commands on Ubuntu 14.04 using the 3.6 PPA
06:26 raghu punit_: the client log file you sent is very recent one probably created after log rotate. Can you send the previous client log as well? It would be having the same name with a large number appended to its name at the end
06:26 jackdpeterson -- looking to see how I can restart the glusterd, glusterfsd services as necessary without bouncing the whole box or manually backgrounding processes
06:27 raghu punit_: if possible, send everything that is there in /var/log/glusterfs
06:29 kumar joined #gluster
06:31 oxae joined #gluster
06:36 hgowtham joined #gluster
06:37 Manikandan joined #gluster
06:38 ashiq joined #gluster
06:39 Philambdo joined #gluster
06:40 deepakcs joined #gluster
06:41 punit_ raghu,please find here http://www.filedropper.com/gluster
06:42 SOLDIERz joined #gluster
06:43 raghu punit_: thanks :)
06:45 anil_ joined #gluster
06:48 an joined #gluster
06:49 dusmant joined #gluster
06:55 raghu punit_: one more thing. Can you please give the output of the command "dmesg" and the file "/var/log/messages"
06:57 lifeofguenter joined #gluster
06:58 gildub joined #gluster
06:59 Bardack joined #gluster
06:59 punit_ raghu: both are here http://www.filedropper.com/dmesg
07:00 raghu punit_: and the output of the command "gluster peer status"
07:02 punit_ raghu: peer status here http://paste.ubuntu.com/10632690/
07:05 fandi joined #gluster
07:07 fandi joined #gluster
07:10 fandi joined #gluster
07:12 nbalacha joined #gluster
07:14 fandi joined #gluster
07:15 hchiramm joined #gluster
07:17 fandi joined #gluster
07:22 fandi joined #gluster
07:22 jtux joined #gluster
07:26 fandi joined #gluster
07:26 [Enrico] joined #gluster
07:28 atalur joined #gluster
07:30 fandi joined #gluster
07:33 hchiramm joined #gluster
07:33 fandi joined #gluster
07:35 anil_ joined #gluster
07:35 ppai joined #gluster
07:37 ghenry joined #gluster
07:37 ghenry joined #gluster
07:37 fandi joined #gluster
07:40 R0ok_ joined #gluster
07:44 badone_ joined #gluster
07:46 dusmant joined #gluster
07:46 DV joined #gluster
07:49 R0ok_ joined #gluster
07:55 kows joined #gluster
07:56 kovshenin joined #gluster
07:57 kows Hello :) I'm running glusterfs-3.6.2 and I'm looking for a wait to lower logging for cli.log which fills up of debug logs like "D [registry.c:408:cli_cmd_register] 0-cli: Returning 0"
07:57 kows is it there a way to change (lower) log priority for the cli?
07:59 anil_ joined #gluster
08:07 DV joined #gluster
08:08 [Enrico] joined #gluster
08:14 ashiq joined #gluster
08:18 nangthang joined #gluster
08:18 topshare joined #gluster
08:24 oxae joined #gluster
08:25 punit_ raghu: is there any thing more i can provide to solve this issue ?
08:30 fsimonce joined #gluster
08:31 chirino joined #gluster
08:31 hagarth joined #gluster
08:33 bala joined #gluster
08:40 _polto_ joined #gluster
08:44 hchiramm joined #gluster
08:46 dusmant joined #gluster
08:56 raghu punit_: can you please turn off quorum (both server and client) and restart the volume?
08:59 [Enrico] joined #gluster
09:07 TvL2386 joined #gluster
09:07 liquidat joined #gluster
09:09 jflf joined #gluster
09:10 deniszh joined #gluster
09:11 ira joined #gluster
09:12 Norky joined #gluster
09:18 tanuck joined #gluster
09:19 tanuck_ joined #gluster
09:19 ctria joined #gluster
09:21 Pupeno joined #gluster
09:27 tanuck joined #gluster
09:30 ppai joined #gluster
09:35 _polto_ joined #gluster
09:40 badone_ joined #gluster
09:50 nangthang joined #gluster
09:51 Slashman joined #gluster
09:52 punit_ raghu, please let me know how to turn off the quorum...is it ok if i will remove it from this file /var/lib/glusterd/groups/virt
09:52 atalur joined #gluster
09:55 harish_ joined #gluster
10:03 punit_ raghu, i have turn off the quorum (server and client) and resart the volume
10:11 punit_ raghu, but still the same issue...
10:21 topshare joined #gluster
10:31 rjoseph joined #gluster
10:36 R0ok_ joined #gluster
10:36 RameshN joined #gluster
10:37 anil_ joined #gluster
10:44 dusmant joined #gluster
10:50 rafi1 joined #gluster
10:53 ppai joined #gluster
10:53 diegows joined #gluster
10:58 doekia joined #gluster
11:00 atalur_ joined #gluster
11:02 jmarley joined #gluster
11:03 lyang0 joined #gluster
11:04 hagarth joined #gluster
11:07 DV_ joined #gluster
11:15 firemanxbr joined #gluster
11:16 rjoseph joined #gluster
11:25 schandra1 joined #gluster
11:27 schandra1 left #gluster
11:28 schandra1 joined #gluster
11:32 ppai joined #gluster
11:36 AndroUser2 joined #gluster
11:40 bennyturns joined #gluster
11:40 AndroUser2 My location: Bengaluru, India
11:41 haomaiwa_ joined #gluster
11:43 Pupeno joined #gluster
11:43 Pupeno joined #gluster
11:44 SOLDIERz joined #gluster
11:47 fsimonce joined #gluster
11:50 lifeofguenter joined #gluster
11:55 LebedevRI joined #gluster
11:57 bala joined #gluster
11:57 edwardm61 joined #gluster
11:58 rafi joined #gluster
12:19 plarsen joined #gluster
12:23 raghu punit_: just do this 1) gluster volume set <volume name> cluster.server-quorum-type none 2) gluster volume set <volume name> cluster.quorum-type none
12:24 kovsheni_ joined #gluster
12:25 ppai joined #gluster
12:32 kdhananjay joined #gluster
12:33 topshare joined #gluster
12:34 kdhananjay joined #gluster
12:35 rjoseph joined #gluster
12:40 RameshN joined #gluster
12:43 schandraMob joined #gluster
12:44 punit_ raghu: i turnoff the quorum but still the VM can not poweron and in the bricks logs same dissconnetion error..
12:47 malevolent joined #gluster
12:47 xavih joined #gluster
12:47 hagarth joined #gluster
12:47 topshare joined #gluster
12:52 ppai joined #gluster
12:55 wkf joined #gluster
12:56 topshare joined #gluster
12:56 dgandhi joined #gluster
12:57 topshare joined #gluster
12:58 rafi joined #gluster
13:04 schandraMob joined #gluster
13:08 nbalacha joined #gluster
13:10 ckotil joined #gluster
13:16 hchiramm joined #gluster
13:18 luis_silva joined #gluster
13:28 theron joined #gluster
13:28 _Bryan_ joined #gluster
13:34 georgeh-LT2 joined #gluster
13:45 hamiller joined #gluster
13:59 rafi joined #gluster
14:02 squizzi joined #gluster
14:08 ildefonso joined #gluster
14:15 topshare joined #gluster
14:16 Norky joined #gluster
14:16 topshare joined #gluster
14:19 sputnik13 joined #gluster
14:28 ppai joined #gluster
14:30 RameshN joined #gluster
14:34 glusterbot News from newglusterbugs: [Bug 1183764] [SNAPSHOT]: fails to create on thin LV <https://bugzilla.redhat.com/show_bug.cgi?id=1183764>
14:37 topshare joined #gluster
14:41 gnudna joined #gluster
14:52 T3 joined #gluster
14:57 R0ok_ joined #gluster
15:00 topshare joined #gluster
15:01 coredump joined #gluster
15:10 rwheeler joined #gluster
15:12 harish_ joined #gluster
15:13 axolx joined #gluster
15:14 axolx Hi, I’m trying to replace a brick in a pool with one that has a different hostname and I can’t get to the instructions. The Gluster doc point to this link, which 404’s
15:14 axolx http://community.gluster.org/q/a-replica-node-has-failed-completely-and-must-be-replaced-with-new-empty-hardware-how-do-i-add-the-new-hardware-and-bricks-back-into-the-replica-pair-and-begin-the-healing-process/
15:15 axolx Could someone point me to documentation on how to replace a brick from a dead node with one with a new hostname?
15:15 sputnik13 joined #gluster
15:21 sputnik13 joined #gluster
15:26 meghanam joined #gluster
15:28 sputnik13 joined #gluster
15:30 deniszh joined #gluster
15:35 deniszh left #gluster
15:35 deniszh joined #gluster
15:52 coredump joined #gluster
15:52 hamiller joined #gluster
16:01 theron__ joined #gluster
16:04 glusterbot News from resolvedglusterbugs: [Bug 1204008] Sqlite3 library required even after disable-tiering is given during configure <https://bugzilla.redhat.com/show_bug.cgi?id=1204008>
16:06 RameshN joined #gluster
16:08 schandraMob joined #gluster
16:19 coredump joined #gluster
16:33 bennyturns joined #gluster
16:36 abyss^ joined #gluster
16:42 saltlake joined #gluster
16:44 veonik joined #gluster
16:44 Pupeno joined #gluster
16:44 Pupeno joined #gluster
16:46 ildefonso joined #gluster
16:53 abyss^ joined #gluster
16:58 jhc76 joined #gluster
17:03 jhc76 glusterfs-builds.log is growing very rapid on one of the bricks. The contents are full of the following. https://gist.github.com/anonymous/49af16a0fa7ef3dc7fd0 I'd appreciate any insight to this matter.
17:03 jhc76 the file grows up to 50GB in one day.
17:04 glusterbot News from newglusterbugs: [Bug 1204247] mmap failed for CEN and END part of zip file <https://bugzilla.redhat.com/show_bug.cgi?id=1204247>
17:14 ndevos jhc76: well, the log-level for that brick process is set to TRACE, you cant get it more verbose than that
17:15 ndevos jhc76: that option can be set as a volume option, "gluster volume info $VOLUME" would show diagnostics.brick-log-level in that case
17:16 ndevos you can unset it with: gluster volume reset $VOLUME diagnostics.brick-log-level
17:16 ndevos or, maybe something else triggered that option to be set, maybe it was passed on the commandline?
17:17 jhc76 ndevos: that's a good possibility. I was out on paternity leave and handed this off to other admin.
17:17 ndevos for example, a fuse mountpoint can have: mount -t glusterfs -o log-level=TRACE ....
17:17 ndevos jhc76: it's the " T " in the logs that give it away ;)
17:18 ndevos INFO is the default, and those messages are logged with " I "
17:18 jhc76 :) thank you sir
17:18 jhc76 my gluster volume info https://gist.github.com/anonymous/a508cb83c7609859b67e
17:18 overclk joined #gluster
17:18 roost joined #gluster
17:21 hagarth joined #gluster
17:22 ndevos jhc76: okay, so no log-level increase there, maybe on the commandline?
17:23 jhc76 ndevos: I ran the gluster log-level reset command, I now see different output. but still in T mode. https://gist.github.com/anonymous/6b89e0ea5aed101dfdd4
17:24 ndevos jhc76: hmm, I'm not sure about it, can you check the commandline of the brick process?
17:25 ndevos jhc76: you can identify the process by the logfile it writes to, that is passed on the commandline too
17:27 jhc76 ndevos: commandline of the brick process?
17:27 ndevos jhc76: yes, try something like this: ps ax | grep glusterfsd
17:27 anrao joined #gluster
17:28 ndevos jhc76: if that does not list the process, then you could also grep for all 'gluster' processes
17:29 ndevos jhc76: actually, what is the filename of that log? it could well be a client-side process and not a brick
17:30 jhc76 only processes I see are these https://gist.github.com/anonymous/f28a6ea0504242d56ae4
17:30 jhc76 file name is glusterfs-builds.log
17:31 ndevos oh, right, yes, you mentioned that before :)
17:31 ndevos do you have a /glusterfs/builds mountpoint? that seems the most likely case
17:32 jhc76 I'm using samba to mount glusterfs
17:32 ndevos aha, then the TRACE is set in the smb.conf where you export that volume
17:34 ndevos something like "gluster:loglevel = TRACE", I think, but I'm not a samba genius
17:34 jhc76 ndevos: you are absolutely right! I set it there and I was looking at gluster config... I'm so dumb. Need to reboot my brain
17:34 jhc76 I set the log level at 10 in samba
17:34 jhc76 thank you sir!
17:34 glusterbot News from newglusterbugs: [Bug 1194640] Tracker bug for Logging framework expansion. <https://bugzilla.redhat.com/show_bug.cgi?id=1194640>
17:34 ndevos oh, nice, thats solved then :)
17:37 sputnik13 joined #gluster
17:38 sputnik13 left #gluster
17:39 lalatenduM joined #gluster
17:45 abyss^ joined #gluster
17:47 anil_ joined #gluster
17:47 fandi joined #gluster
17:56 gh5046 joined #gluster
17:58 gh5046 Is there any reason why the glusterfs-epel.repo file under the 3.5 download directory doesn't point to 3.5?
17:58 gh5046 http://download.gluster.org/pub/gluster/glusterfs/3.5/LATEST/CentOS/glusterfs-epel.repo
17:58 gh5046 The contents point to pub/gluster/glusterfs/LATEST instead of pub/gluster/glusterfs/3.5
17:58 shaunm joined #gluster
17:58 shaunm joined #gluster
17:59 gh5046 If I use that file as is it'll install 3.6.x instead of 3.5.x
17:59 gh5046 It's easy enough to modify, but it strikes me as odd.
18:00 ndevos gh5046: oh, that is definitely wrong then, please file a bug against the projcet-infrastructure for that
18:00 glusterbot https://bugzilla.redhat.com/enter_bug.cgi?product=GlusterFS
18:01 gh5046 Ah, okay. I will do that. Thanks!
18:02 morse joined #gluster
18:04 kkeithley_ don't file a BZ. I've fixed the repo file(s)
18:05 gh5046 Nice!
18:06 gh5046 I see 3.5/LATEST/...repo is updated.  The 3.5/3.5.x/...repo files are still incorrect.
18:06 kkeithley_ check again. ;-)
18:08 gh5046 The others (3.5.1, 3.5.2, 3.5.3) are pointing to 3.5/LATEST.
18:08 gh5046 Is that how you want it?
18:08 gh5046 (Not complaining, just curious)
18:08 kkeithley_ I think so, yes
18:11 gh5046 Cool!  Thanks for fixing that!
18:13 its_pete joined #gluster
18:21 Arminder- joined #gluster
18:27 Arminder joined #gluster
18:28 Arminder- joined #gluster
18:31 theron joined #gluster
18:39 elitecoder JoeJulian: I have a theoretical question. If XFS were to lose a file with two gluster nodes in a replication config, would self-healing result in deletion or duplication?
18:39 elitecoder (One node's underlying filesystem, XFS loses a file)
18:40 elitecoder Also is GlusterGS compatible with EXT4 now?
18:40 wkf joined #gluster
18:42 gh5046 elitecoder: ext4 support has apparently been fixed as of >= 3.4.0 and 3.31.  http://review.gluster.org/#/c/4711/
18:43 elitecoder gh5046: thanks!
18:44 gh5046 Maybe whoever does the GlusterFS blog could do a post on it someday :)
18:44 elitecoder lol
18:48 Arminder joined #gluster
18:50 elitecoder sigh
18:50 elitecoder W: Failed to fetch http://ppa.launchpad.net/semiosis/ubuntu-glusterfs-3.5/ubuntu/dists/trusty/main/binary-amd64/Packages  404  Not Found
18:50 JustinClift semiosis: ^
18:50 gnudna so is there an issue with xfs and glusterfs?
18:51 elitecoder gnudna: Why do you ask?
18:51 gnudna also on the original mount can i i use noatime etc..
18:51 gnudna just seen people mentioning files missing
18:51 elitecoder Other than myself?
18:51 gnudna and the fs seem to always be xfs
18:51 kkeithley_ ,,(ppa)
18:51 glusterbot The official glusterfs packages for Ubuntu are available here: 3.4: http://goo.gl/M9CXF8 3.5: http://goo.gl/6HBwKh 3.6: http://goo.gl/XyYImN -- See more PPAs for QEMU with GlusterFS support, and GlusterFS QA releases at https://launchpad.net/~gluster -- contact semiosis with feedback
18:51 eclectic joined #gluster
18:52 gnudna btw i have been reading mail thread so as far as i know i could be readin your thread ;)
18:52 elitecoder I have no thread so it wasn't me
18:52 elitecoder could you point me to it?
18:52 gnudna at the moment i cannot i do not have it on hand
18:52 kkeithley_ elitecoder: ,,(ppa)
18:52 glusterbot elitecoder: The official glusterfs packages for Ubuntu are available here: 3.4: http://goo.gl/M9CXF8 3.5: http://goo.gl/6HBwKh 3.6: http://goo.gl/XyYImN -- See more PPAs for QEMU with GlusterFS support, and GlusterFS QA releases at https://launchpad.net/~gluster -- contact semiosis with feedback
18:53 gnudna my question is what benefits are there to using libgfapi over the regular mount using glusterfs
18:53 gnudna in my case related to kvm images
18:54 kkeithley_ if you're using libvirt, libgfapi bypasses the fuse mount, eliminating two copies between kernel and userspace, and associated context switching
18:54 tdasilva joined #gluster
18:55 semiosis :O
18:55 semiosis JustinClift: i deleted the old ~semiosis PPAs.  they haven't been updated in a long time and were deprecated, replaced by the ~gluster ppas
18:57 JustinClift elitecoder: ^
18:57 JustinClift Ahhh, already done :)
18:57 JustinClift Ths :)
18:57 JustinClift Thx even
18:58 elitecoder Thanks guys. I'm running ubuntu 14 and i use apt on them. I had it setup to upgrade with everything else then I got that error
18:58 elitecoder kinda trying to sort out what to do now
18:59 Arminder joined #gluster
19:01 gh5046 Question about GlusterFS and Firewalls: Is every file read/write a separate packet?
19:02 Arminder- joined #gluster
19:03 gnudna thanks kkeithley_
19:05 Arminder joined #gluster
19:11 Pupeno joined #gluster
19:11 Pupeno joined #gluster
19:14 Arminder joined #gluster
19:14 ricky-ticky joined #gluster
19:14 elitecoder If I were to upgrade to gluster 3.6 from 3.5.2, should I expect any problems?
19:15 roost joined #gluster
19:17 tdasilva joined #gluster
19:20 tdasilva joined #gluster
19:22 Pupeno_ joined #gluster
19:31 Arminder joined #gluster
19:40 gh5046 I ask because I have to use a firewall, but I'm wondering if it's worth confiuring notrack on glusterfs traffic.
19:49 Pupeno joined #gluster
19:51 Pupeno__ joined #gluster
19:53 o5k joined #gluster
19:54 Pupeno joined #gluster
19:54 _polto_ joined #gluster
19:55 ctria joined #gluster
19:55 o5k nobody can help me with this ? (issue presented yesterday here: http://irclog.perlgeek.de/gluster/2015-03-19#i_10306953 )
19:57 Arminder joined #gluster
19:58 Arminder- joined #gluster
19:58 ira joined #gluster
19:58 elitecoder o5k_: i'd try but busy with my isssue atm
20:03 fyxim joined #gluster
20:03 wkf joined #gluster
20:04 Pupeno_ joined #gluster
20:09 plarsen joined #gluster
20:10 samsaffron___ joined #gluster
20:16 Arminder joined #gluster
20:17 oxae joined #gluster
20:17 JustinClift elitecoder: Wait until GlusterFS 3.5.3 is out, just to be safe
20:18 _polto_ joined #gluster
20:18 gnudna JustinClift i thought version 3.6.2 is considered stable
20:19 gh5046 ?
20:19 gh5046 # rpm -aq glusterfs
20:19 gh5046 glusterfs-3.5.3-1.el6.x86_64
20:20 gnudna well i used epel repo and i am running 3.6.2
20:20 gh5046 That's latest.  The latest stable is 3.5.3.
20:21 gnudna glusterfs-3.6.2-1.el7.x86_64
20:22 gnudna http://download.gluster.org/pub/gluster/glusterfs/LATEST/CentOS/epel-6/x86_64/
20:22 elitecoder JustinClift: Ok thanks. I'll use the 3.5 ppa for now then.
20:22 hamiller left #gluster
20:22 * JustinClift is just paranoid
20:22 JustinClift ;)
20:23 gh5046 Oh, not out for debian/ubuntu?  I'm slow.
20:26 wkf joined #gluster
20:28 o5k elitecoder, ok, i'll try to do what I can , once you are done from your current issue, let's keep in touch about live migration issue :)
20:28 Arminder joined #gluster
20:29 elitecoder o5k: Couldn't you just add more bricks?
20:29 o5k elitecoder, ??
20:30 elitecoder You mean migrating TO gluster?
20:32 o5k I will make my choice based on many points: which one would perform live migration better, at the same time i think it would be more interesting to use libgfapi (if kvm would be better than xen for the VM live migration)
20:32 elitecoder this doesn't sound very gluster related lol
20:33 o5k let's say, i will be using libgfapi, and use gluster as block storage
20:34 Arminder joined #gluster
20:36 o5k I should benchmark both solutions in order to get which one would provide better results
20:36 o5k so i'm here to get some help about using gluster as a block storage , to create VMs directly on it with qemu
20:37 o5k I haven't found a fresh tutorial to set up gluster this way and set up some VMs on it
20:39 o5k i'm using glusterfs 3.6.2
20:43 o5k elitecoder, can u suggest some resources to get started with this ?
20:46 elitecoder Is gluster running yet?
20:47 axolx left #gluster
20:47 o5k yes
20:47 elitecoder oh wow i see what you're saying
20:47 elitecoder i don't know how to use gluster that way
20:48 elitecoder I mount it as a filesytem
20:48 o5k i can use it that way, use a splitted volume as the files (virtual hard drives of VMs) are big
20:50 o5k but, seeing the effort made by gluster developers to make it possible to use it as block storage, and the gain in performance it would make for the VMs, i want to learn how to do it :)
20:50 o5k elitecoder, thanks anyway ;-)
20:52 elitecoder oh cool.
20:53 Pupeno joined #gluster
20:54 SOLDIERz joined #gluster
20:59 gnudna left #gluster
21:08 elitecoder Is GlusterFS 3.6 really unstable?
21:09 gh5046 Well, it's in base for CentOS 6 :)
21:19 elitecoder Oh... well wtf
21:20 elitecoder I suppose either way I'll upgrade to latest 3.5 and do 6 later
21:21 kkeithley_ No, I wouldn't say it's unstable. Note that the gluster packages in base RHEL and CentOS are the RHS client-side (only) bits, a.k.a. downstream.  That's based on 3.6.x community, or upstream sources. Community RPMs are what's available on download.gluster.org
21:21 elitecoder ok
21:21 kkeithley_ Also community dpkgs for Debian on download.gluster.org in addition to community dkpgs for Ubuntu in the PPA.
21:22 _polto_ joined #gluster
21:23 gh5046 Is it safe to use 3.6 client software with a 3.5 server?
21:23 elitecoder I was about to ask the same thing haha
21:26 kkeithley_ I can never remember. I googled.  E.g. http://www.mail-archive.com/gluster-users@gluster.org/msg18904.html, seems to say  that 3.5 server and 3.6 client will work
21:28 kkeithley_ AFR changed in 3.6. A 3.5 client won't have the correct AFR bits to match a 3.6 server.
21:28 kkeithley_ But a 3.6 client will have the necessary bits to work with a 3.5 server.
21:28 kkeithley_ YMMV
21:33 Alpinist joined #gluster
21:34 wkf joined #gluster
21:34 elitecoder ok
21:34 elitecoder nice
21:41 rotbeard joined #gluster
22:05 fattaneh joined #gluster
22:12 tdasilva joined #gluster
22:16 gh5046 left #gluster
22:21 gildub joined #gluster
22:29 fattaneh left #gluster
22:33 YogeeBear joined #gluster
22:38 YogeeBear left #gluster
23:17 anil_ joined #gluster
23:22 plarsen joined #gluster
23:34 Ozy_ joined #gluster
23:35 Ozy_ hello. gluster newbie searching for some simple answers. anyone online ?
23:42 badone_ joined #gluster
23:44 Ozy_ heelp ?
23:47 elitecoder Ozy_: ask your question and maybe someone will be able to answer it
23:48 elitecoder People watch the channel and if they see something they know will just answer, but are unlikely to answer the first two things you said
23:48 Ozy_ (irc n00b too it seems :))
23:50 elitecoder : D
23:50 Ozy_ i have two small servers, each with a raid 1 array. i want to get them into a HA san for Hyper-V machines. i thought installing Gluster, and then using both iSCSI targets as one multipath device for Windows. Would that work ?
23:51 Ozy_ Do i miss keepalived / heartbeat in that kind of setup ? And last but not least, does gluster support this setup as active-active SAN ?
23:53 elitecoder I can tell you that you can use multiple boxes as one drive
23:54 elitecoder I use two servers' 40gb partitions in replication mode, so it shows up on clients as 40gb but it replicates between eachother
23:54 elitecoder if one node goes down the other is there and availableimmediately
23:55 Ozy_ i read something about writeback time. I mean gluster says that the data is written even if it isn't yet. Does that create any trouble with an active-active scenario ?

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