Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2015-02-04

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

All times shown according to UTC.

Time Nick Message
00:48 _Bryan_ joined #gluster-dev
00:50 jamesc|2 joined #gluster-dev
02:17 rafi joined #gluster-dev
03:04 kshlm joined #gluster-dev
03:44 itisravi joined #gluster-dev
03:50 atinmu joined #gluster-dev
04:13 shubhendu_ joined #gluster-dev
04:14 bala joined #gluster-dev
04:26 Manikandan joined #gluster-dev
04:38 anoopcs joined #gluster-dev
04:41 jiffin joined #gluster-dev
04:59 gem joined #gluster-dev
04:59 nkhare joined #gluster-dev
05:00 spandit joined #gluster-dev
05:01 soumya_ joined #gluster-dev
05:01 kdhananjay joined #gluster-dev
05:02 aravindavk joined #gluster-dev
05:03 prasanth_ joined #gluster-dev
05:10 ppai joined #gluster-dev
05:13 pranithk joined #gluster-dev
05:15 shubhendu_ joined #gluster-dev
05:25 ndarshan joined #gluster-dev
05:45 atinmu joined #gluster-dev
05:46 kanagaraj joined #gluster-dev
05:59 vikumar joined #gluster-dev
06:06 aravindavk joined #gluster-dev
06:15 vikumar joined #gluster-dev
06:17 anrao joined #gluster-dev
06:20 overclk joined #gluster-dev
06:28 atinmu joined #gluster-dev
07:02 raghu joined #gluster-dev
07:17 soumya_ joined #gluster-dev
07:21 anrao joined #gluster-dev
07:27 shubhendu_ joined #gluster-dev
07:49 bala joined #gluster-dev
07:59 kanagaraj joined #gluster-dev
08:10 tdasilva joined #gluster-dev
08:11 kkeithley1 joined #gluster-dev
08:12 nishanth joined #gluster-dev
08:27 ws2k3 joined #gluster-dev
08:28 ws2k3 joined #gluster-dev
08:29 rjoseph joined #gluster-dev
08:35 lalatenduM joined #gluster-dev
08:36 shubhendu joined #gluster-dev
08:36 hagarth joined #gluster-dev
08:49 kdhananjay joined #gluster-dev
08:52 rjoseph joined #gluster-dev
09:04 hagarth joined #gluster-dev
09:24 rjoseph joined #gluster-dev
09:31 kanagaraj joined #gluster-dev
09:45 PaulCuzner joined #gluster-dev
10:08 ppai joined #gluster-dev
10:20 kkeithley1 joined #gluster-dev
10:24 kkeithley1 joined #gluster-dev
10:38 bala joined #gluster-dev
10:46 hchiramm joined #gluster-dev
10:53 ndevos hagarth: will you send out a reminder for the meeting, or shall I do that?
10:53 atinmu joined #gluster-dev
10:53 hagarth ndevos: please go ahead
10:54 ndevos hagarth: ok
10:58 ndevos REMINDER: in ~60 minutes from now, the weekly Gluster Community meeting starts in #gluster-meeting
11:02 soumya__ joined #gluster-dev
11:04 jiffin1 joined #gluster-dev
11:12 hagarth joined #gluster-dev
11:15 kanagaraj joined #gluster-dev
11:21 atinmu joined #gluster-dev
11:21 hchiramm joined #gluster-dev
11:45 ndevos REMINDER: in ~15 minutes from now, the weekly Gluster Community meeting starts in #gluster-meeting
12:00 ndevos REMINDER: the weekly Gluster Community meeting starts now in #gluster-meeting
12:00 jdarcy joined #gluster-dev
12:04 rjoseph joined #gluster-dev
12:10 anrao joined #gluster-dev
12:10 soumya_ joined #gluster-dev
12:11 jiffin joined #gluster-dev
12:22 mikedep333 joined #gluster-dev
12:26 ndarshan joined #gluster-dev
12:27 gem joined #gluster-dev
12:36 overclk joined #gluster-dev
12:38 gothos lalatenduM: Hey, about the rpm spec problem with logrotate we talked about @fosdem... can you point me to the location of the spec filed which is used for the centos 6 rpms @download.gluster.org?
12:39 lalatenduM gothos, sure
12:40 lalatenduM gothos, here it is http://pkgs.fedoraproject.org/cgit/glusterfs.git
12:40 hchiramm gothos, https://github.com/gluster/glusterfs/blob/master/glusterfs.spec.in
12:41 lalatenduM gothos, we use fedora dist git  for building el rpms and fedora
12:41 lalatenduM hchiramm, d.g.o rpms have spec file from fedora dist git
12:42 hchiramm true.. but we are keeping both in sync
12:42 hchiramm or trying to keep it in recent versions
12:43 lalatenduM hchiramm, yeah we try to , it is not 100% equal
12:43 gothos lalatenduM: okay, thx. the wrong logrotate stuff is clearly in there. can I just send you a patch for the file in the fedora git or do I have to do threw their bugzilla or something?
12:43 kkeithley_ _trying_ to keep them in sync.  I don't believe they're quite the same yet
12:43 gothos hchiramm: it's not in sync at the moment
12:43 gothos true
12:43 hchiramm gothos, yep
12:43 hchiramm gothos, send a patch to upstream
12:43 lalatenduM gothos, nice , you can file a bug against gluster mainline and I will pick from there to put in el6 rpms
12:44 lalatenduM gothos++
12:44 glusterbot lalatenduM: gothos's karma is now 4
12:44 gothos lalatenduM: okay, I'll do that
12:44 lalatenduM gothos, awesome :)
12:44 ira joined #gluster-dev
12:44 hchiramm kkeithley, yeah :)
12:44 kkeithley_ we could get the patches I've had in gerrit for months reviewed.
12:44 kkeithley_ hint hint
12:45 hchiramm \o :P
12:45 lalatenduM kkeithley, I think you need to rebase them, isn;t it?
12:45 kkeithley_ probably
12:45 kkeithley_ by now
12:45 lalatenduM kkeithley, and ndevos or you can merge those as you guys are sub-maintainers for build system :)
12:47 kkeithley_ yes, I think I heard a rumor recently that I've been granted privs to do that. I hadn't been aware
12:52 gothos btw. do you know why --disable-syslog is set for rhel6 in spec? since rsyslog is used for rhel6
12:54 lalatenduM gothos, nope , this is something was present when i started pkging, kkeithley, might have info abt on that
12:54 kkeithley_ maybe it wasn't always in rhel6
12:55 kkeithley_ I don't remember ever changing it.
12:55 gothos AFAIR 6.0 had rsyslog, it was in the switch from 5 -> 6, but I'll check
12:57 kkeithley_ in any event it's there now
12:59 gothos yeah, and changing the spec might even break the currently expected behaviour
12:59 anoopcs joined #gluster-dev
13:12 hagarth joined #gluster-dev
13:18 rjoseph joined #gluster-dev
13:20 ndevos gothos: --disable-syslog on rhel6 was set because it would otherwise require rsyslog-mmjson (or something like that), which is not in RHEL/CentOS
13:20 glusterbot ndevos: gothos's karma is now 3
13:21 ndevos uh, what?!
13:21 ndevos gothos++ I did not intend to do that :)
13:21 glusterbot ndevos: gothos's karma is now 4
13:25 lalatenduM ndevos, gothos yes we have a bug for removing rsyslog-mmjson dependency from spec file
13:27 kkeithley_ rsyslog-mmjsonparse
13:27 lalatenduM kkeithley, yes
13:27 lalatenduM I am not able to find the bug now, but I have seen it sometime back
13:29 kkeithley_ comment says "because RHEL6 doesn't have rsyslog-mmcount also
13:35 gothos okay, thx. that clears it up. I'll take a closer look tomorrow.
13:38 JustinClift Ugh, looks like our Jenkins infra atm is in a really bad state
13:38 * JustinClift is rebooting nodes now
13:40 JustinClift Oh interesting.  /var/log/glusterfs has filled the partition on several of the nodes (17GB of logs)
13:43 gothos what distributions are you using on the nodes?
13:43 lalatenduM JustinClift, lots of folks are looking for you, kshlm is one of them :)
13:45 JustinClift lalatenduM: I got to the hotel in London not long ago.
13:45 JustinClift lalatenduM: Now on laptop while eating food.
13:45 JustinClift kshlm: Ping?
13:48 JustinClift lalatenduM: Thx btw :)
13:49 lalatenduM JustinClift, yeah we guessed so, it is #crazyeventweek :)
13:49 JustinClift ;)
13:52 ndevos gothos: most use CentOS-6, but there are also some NetBSD and FreeBSD
13:52 ndevos gothos: build.gluster.org is the instance, not sure what it reveals, but you can have a look there
13:53 kshlm joined #gluster-dev
13:53 hchiramm JustinClift, kshlm is here
13:54 ndevos JustinClift: oh man, you did not reboot those systems with my awesome reboot job: http://build.gluster.org/job/reboot-vm/
13:54 kshlm JustinClift, I have two things.
13:55 kshlm One. I need liburcu to be installed on the Jenkins slaves.
13:55 JustinClift ndevos: Ahhh, I couldn't remember the URL
13:56 JustinClift ndevos: Doing it via the rackspace gui atm.  slave23 is not coming back from a soft boot.
13:56 JustinClift 26 and 26 are rebooting now.  Hopefully they come back
13:56 JustinClift kshlm: Just saw your email about that.
13:56 ndevos JustinClift: thats is your chance to try the script on slave23!
13:56 kshlm Two. Can you help with our GSOC org proposal and can you become the admin for the the organization?
13:56 gothos ndevos: JustinClift: well, if they are using the rpm's with the fedora spec then logrotate is broken, which might be the reason for the size of the logs
13:56 JustinClift ndevos: It's not responding to pings either, so good luck? :)
13:57 ndevos JustinClift: its doing a hard reboot through the RAX-API :D
13:57 JustinClift gothos: Ahhh, that's good info
13:57 JustinClift ndevos: Cool.  Smart approach.  Ok, will now use your script. :)
13:58 ndevos gothos: well... the slaves get glusterfs installed from the sources for each regression test, its nothing that runs very long - and the jobs should cleanup the logs
13:59 ndevos ... "and the jobs should cleanup the logs", isnt that right, JustinClift?
13:59 JustinClift Hmmm, I don't actually remember.  It's been a *long* time since I looked at the ob scripting. :/
13:59 ndevos well, the gluster logs should probably even get written to some non /var location
13:59 JustinClift s/ob/job/
13:59 JustinClift Yeah, I'm thinking that now too. ;)
14:00 kshlm They archive the logs, and probably should be clearing the logs after.
14:00 ndevos JustinClift: any idea what filled up /var/log ?
14:00 JustinClift 14GB /var/log/glusterfs/glusterfs.log file
14:00 * ndevos *cough*
14:01 JustinClift (or similar filename, I'm only remembering that
14:01 JustinClift )
14:01 JustinClift 17GB on one of the nodes
14:02 JustinClift Slave23,26,28 not coming back from hard reboot
14:02 ndevos that is sad :-/
14:02 JustinClift I'll nuke them and setup new replacement nodes.  Gimme a few minutes.
14:02 JustinClift Meh
14:03 ndevos you got that ansible scripted yet?
14:03 JustinClift It's still on my ToDo list ;(
14:03 kshlm JustinClift, looks like you're busy with other stuff right now. I'm leaving for home now, we can continue later.
14:04 JustinClift kshlm: The library is installed now
14:04 JustinClift kshlm: Well, on the nodes that are online
14:04 ndevos JustinClift: the -devel version?
14:04 JustinClift yeah
14:04 ndevos hehe, thanks :)
14:05 * JustinClift just cut-n-pasted the name in kshlm's email
14:05 ndevos also on NetBSD / FreeBSD ?
14:05 JustinClift Nope.  We'll need to prod Manu a bit
14:06 JustinClift I suspect he's a bit peeved, since I've not been around to assist/do-stuff with Jenkins for ages.  Especially after the good work he'd been putting in. :(
14:06 shaunm joined #gluster-dev
14:11 kanagaraj joined #gluster-dev
14:12 bala joined #gluster-dev
14:12 JustinClift ls -la
14:12 JustinClift Heh
14:16 anoopcs joined #gluster-dev
14:20 anoopcs_ joined #gluster-dev
14:23 Manikandan_ joined #gluster-dev
14:23 JustinClift ndevos: This is your area of knowledge isn't it? http://www.gluster.org/pipermail/gluster-users/2015-February/020462.html
14:24 JustinClift ndevos: This is from the Corvid Tech guys, which is an er... "high touch" upstream user.
14:27 shyam joined #gluster-dev
14:31 anoopcs joined #gluster-dev
14:32 nishanth joined #gluster-dev
14:32 anoopcs joined #gluster-dev
14:33 anoopcs joined #gluster-dev
14:34 ndevos JustinClift: uh, through Samba? I know about ACLs and groups and stuff, but little about Samba
14:34 anoopcs joined #gluster-dev
14:40 lalatenduM JustinClift, ira  is ur guy:)
14:45 ndevos lalatenduM: no, it seems you are! I just responded to the email and included a link to your blog :D
14:47 JustinClift ndevos: Thanks. :)
14:48 ndevos JustinClift: sure, np!
14:49 JustinClift ls -la
14:49 JustinClift Gah
14:53 lalatenduM ndevos, haha :) , I dont know much abt grps
14:54 ndevos lalatenduM: the nice thing is "that it does not matter" :)
14:58 lalatenduM ndevos, cool
15:13 nishanth joined #gluster-dev
15:16 shubhendu joined #gluster-dev
15:17 kshlm joined #gluster-dev
15:20 wushudoin joined #gluster-dev
15:38 kshlm JustinClift, thanks for installing the library
15:38 kshlm JustinClift++
15:38 glusterbot kshlm: JustinClift's karma is now 33
15:40 kkeithley_ JustinClift: has dlambrig or jfernand asked you to to put sqlite rpms on jenkins machines?
15:45 hagarth joined #gluster-dev
15:51 JustinClift kkeithley_: Not yet
15:51 JustinClift kkeithley_: Shouldn't be hard to do though
15:51 kkeithley_ I think they want sqlite. Better wait for them to tell you themselves.
15:51 JustinClift The rpms available are probably not the fastest tho
15:52 JustinClift There are good improvements with the lastest version of SQLite (3.8.8.2) compared to previous
15:52 JustinClift But "good enough" ;)
15:52 kkeithley_ good enough for regression testing
15:52 JustinClift kkeithley_: Sure, np
16:05 JustinClift Hmmm, this one's a weird failure.  Bad paths in the regression script maybe? http://build.gluster.org/job/rackspace-regression-2GB-triggered/4034/console
16:08 soumya__ joined #gluster-dev
16:12 semiosis kkeithley_: so i have a vagrant config that builds a box using ubuntu trusty that can build packages for all ubuntu & debian releases
16:13 semiosis kkeithley_: need to make a few more tweaks to the script then i'll work on hooking it in to jenkins
16:13 semiosis kkeithley_: ran into problems building from git though.  currently only building from the release tarballs on download.g.o
16:15 JustinClift semiosis: Sounds like we'll be able to get a regression test for Ubuntu happening, if the building-from-git problems can be solved?
16:15 semiosis i'm sure that can be solved.  i've built from git before but it's been a long time
16:16 semiosis probably some weird autoconf version issue.  i'm not a C guy
16:16 JustinClift Ahhh yeah.  Hopefully just a matter of remembering some details then. :)
16:17 semiosis was hoping maybe kkeithley could do a build from git on his debian box and give me some pointers
16:17 semiosis :)
16:17 JustinClift :)
16:17 semiosis get it?  pointers?
16:17 semiosis hahaha
16:17 JustinClift Here's a free one: 0
16:17 JustinClift :D
16:32 ndevos semiosis: maybe https://forge.gluster.org/bugzappers/nightly-builds/blobs/master/autobuild-copr.sh#line220 ?
16:32 ndevos that is what the nightly builds for the RPMs do to get a version
16:32 semiosis interesting
16:34 JustinClift Hmmm, *some* of the regression tests are failing at build time
16:38 bala joined #gluster-dev
16:39 JustinClift ndevos: Interesting script failure there? http://build.gluster.org/job/reboot-vm/27/console
16:42 * JustinClift just killed the 1118 cores on slave22  Hopefully the node will function better now
16:54 hagarth joined #gluster-dev
16:54 ndevos JustinClift: yeah, that happens on occasion, web access from build.gluster.org seems to get hijacked by a proxy of some kind :-/
16:55 jobewan joined #gluster-dev
16:55 ndevos JustinClift: there was an email about it to the -infra list, but no response yet
16:55 ndevos *HINT*
17:02 JustinClift Ahhh.  I saw the header but didn't read it
17:02 JustinClift I sometimes get similar behaviour on my home machines when loading web pages.  eg they load the gluster.org web page instead of the currect URL
17:03 JustinClift I think the gluster.org DNS server is doing a wildcard DNS thing.  So *.gluster.org goes to the www.gluster.org server
17:04 JustinClift (my home setup is using *.uk.gluster.org)
17:05 ndevos JustinClift: well, the script really does not use any gluster.org address, so it must be something on build.gluster.org or its hypervisor/network
17:05 gem joined #gluster-dev
17:07 JustinClift Yeah, no argument.  Was "just saying" :)
17:23 gem_ joined #gluster-dev
17:24 jobewan joined #gluster-dev
17:37 JustinClift k, I think those new slave nodes are setup and working ok now
17:38 JustinClift Turns out there was a bunch of post-setup adjustment we've been doing over time, but hadn't added to the setup scripting
17:38 JustinClift It's now been added to the scripting
17:38 JustinClift Ansible later ;)
17:55 jamesc|2 joined #gluster-dev
18:29 lalatenduM joined #gluster-dev
18:59 shaunm joined #gluster-dev
19:04 ira joined #gluster-dev
19:13 shaunm joined #gluster-dev
19:41 jobewan joined #gluster-dev
19:49 bala joined #gluster-dev
19:49 kkeithley1 joined #gluster-dev
20:43 bala joined #gluster-dev
22:54 jobewan joined #gluster-dev

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