Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2014-08-14

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

All times shown according to UTC.

Time Nick Message
00:10 hchiramm_ joined #gluster-dev
00:37 systemonkey joined #gluster-dev
01:14 Yuan__ joined #gluster-dev
01:15 Yuan_ joined #gluster-dev
01:26 bala joined #gluster-dev
02:47 Yuan_ joined #gluster-dev
02:48 Yuan__ joined #gluster-dev
03:09 bharata-rao joined #gluster-dev
03:35 hchiramm_ joined #gluster-dev
03:48 kdhananjay joined #gluster-dev
03:51 msvbhat_ joined #gluster-dev
03:52 nixpanic_ joined #gluster-dev
03:52 nixpanic_ joined #gluster-dev
03:53 spandit joined #gluster-dev
03:53 tg2 joined #gluster-dev
03:55 ilbot3 joined #gluster-dev
03:55 Topic for #gluster-dev is now Gluster Development Channel - http://gluster.org | For general chat go to #gluster | Patches - http://review.gluster.org/ | Channel Logs - https://botbot.me/freenode/gluster-dev/ & http://irclog.perlgeek.de/gluster-dev/
04:10 ilbot3 joined #gluster-dev
04:10 Topic for #gluster-dev is now Gluster Development Channel - http://gluster.org | For general chat go to #gluster | Patches - http://review.gluster.org/ | Channel Logs - https://botbot.me/freenode/gluster-dev/ & http://irclog.perlgeek.de/gluster-dev/
04:18 Yuan_ joined #gluster-dev
04:18 Yuan__ joined #gluster-dev
04:39 anoopcs joined #gluster-dev
04:45 ppai joined #gluster-dev
04:51 atinmu joined #gluster-dev
04:52 kdhananjay joined #gluster-dev
04:53 Rafi_kc joined #gluster-dev
04:53 nishanth joined #gluster-dev
04:55 jiffin joined #gluster-dev
04:58 bala joined #gluster-dev
05:02 Yuan_ joined #gluster-dev
05:02 Yuan__ joined #gluster-dev
05:05 skoduri joined #gluster-dev
05:08 itisravi_ joined #gluster-dev
05:33 lalatenduM joined #gluster-dev
05:37 atalur joined #gluster-dev
05:38 overclk joined #gluster-dev
05:44 hagarth joined #gluster-dev
06:03 raghu joined #gluster-dev
06:03 ppai joined #gluster-dev
06:49 ppai joined #gluster-dev
07:00 Yuan_ joined #gluster-dev
07:01 Yuan__ joined #gluster-dev
07:07 Yuan_ joined #gluster-dev
07:24 rgustafs joined #gluster-dev
07:44 ppai joined #gluster-dev
07:53 ppai joined #gluster-dev
08:20 ppai joined #gluster-dev
08:33 bharata-rao joined #gluster-dev
08:39 ppai joined #gluster-dev
09:10 vimal joined #gluster-dev
09:35 bharata-rao joined #gluster-dev
10:31 edward1 joined #gluster-dev
10:37 ira joined #gluster-dev
10:45 kkeithley1 joined #gluster-dev
11:07 kkeithley_ hagarth: brick .vol files are created (in /var/lib/glusterd/vols/$volname/...) when the volume is created.   But the nfs.vol and glustershd.vol files are not created until the volume is started!  Is that really intentional? Do you know what the rationale is for that? (I'm testing some things and this has tripped me up.)
11:07 kkeithley_ Or anyone else? Do you know why?
11:09 kkeithley_ Seems a little bit odd
11:25 skoduri joined #gluster-dev
11:32 ndevos kkeithley_: I think nfs.vol gets re-created on every glusterd start...
11:33 ndevos but yes, maybe it needs at least one volume where nfs is not disabled, hmm
11:33 * ndevos doesnt know the reasoning behind it and points to the glusterd guys
11:37 kkeithley_ I don't have nfs disabled.  I'm working on automating some dynamic analysis (e.g. valgrind). I'm starting the brick processes "manually", i.e. not with `gluster volume start ...`.  Since there isn't a volume start command the nfs-server.vol file doesn't get created. Or the glustershd-server.vol file. But without an nfs-server.vol file the nfs glusterfs process doesn't work.
11:38 kkeithley_ Not a big deal, I can just squirrel away the *-server.vol files and copy them into place.
11:39 kkeithley_ I don't have a problem with (re)creating a new nfs-server.vol file every time the volume is restarted. It just seems strange to create some .vol files one place, and the others at a different place.
11:39 ndevos oh, but isnt there a --run-with-valgrind switch somethere, or maybe it was an xlator option?
11:39 kkeithley_ I've never seen that
11:41 kkeithley_ but I see it now, now that I've looked for it.
11:41 ndevos seems to be an option for /etc/glusterd/glusterd.vol ?
11:41 kkeithley_ that's okay, there are other tools too
11:42 kkeithley_ comment says runs bricks.  Nothing about the nfs glusterfs
11:42 kkeithley_ maybe I'm taking that too literally
11:42 ndevos yeah, I'd like to be able to run nfs with valgrind
11:42 kkeithley_ that's okay though, I have other tools in mind too
11:42 kkeithley_ besides valgrind
11:43 kkeithley_ cachegrind, helgrind, gcc-4.8 and 4.9 fsanitize
11:43 ndevos valgrind has (too) many options, it would be nice to be able to tune those
11:45 kkeithley_ gcc-4.9 has many more -fsanitize options than 4.8
11:46 lalatenduM ndevos, I have done some changes to the wiki page of Bug Triage
11:46 lalatenduM e.g http://www.gluster.org/community/documentatio​n/index.php/Bug_triage#Bug_Triage_Guidelines
11:48 lalatenduM ndevos, changed spacing , text it some places , added text to make it simpler . I think to me it look simpler and easier to follow then before . but it is just or me you also think so ?
11:49 * ndevos reads up
11:49 lalatenduM kkeithley, kindly take a look if you have some spare time :)
11:49 lalatenduM and let me know
11:49 kkeithley_ is BDB really the _Berkeley db_ backend storage?
11:50 kkeithley_ I was thinking the "bd" xlator was a pseudo block device for things like qemu/kvm
11:52 lalatenduM hchiramm, ^^ you too :)
11:52 ndevos lalatenduM: looks good to me, I've read the Bug_Triage_Guidelines paragraph and like it, did you change other parts too?
11:52 lalatenduM ndevos, yes
11:52 kkeithley_ BD translator V2 - Exports Block devices on server side as regular files to client
11:52 lalatenduM ndevos, added a section called " Example of Triaged Bugs "
11:53 lalatenduM also taken the "Keywords section" to lower end
11:53 lalatenduM added a new keyword as "Tracking"
11:54 kkeithley_ oh, docs/legacy/user-guide.info speaks to some old BDB xlator that used BDB as a storage back end.
11:55 kkeithley_ <obiwan>this isn't the xlator you're looking for</obiwan>
11:57 ndevos lalatenduM: "Example of Triaged Bugs" is more of just a list, and if we dont keep on triaging, the list will be empty after a while
11:58 ndevos lalatenduM: other than that, I dont have any remarks :)
11:59 lalatenduM ndevos, agree, but the other options did not workout for example if I take all possible bug states in the search , it is getting old bugs ( 2 to 3 years old)
11:59 lalatenduM I think previously alos they used triaged keyword
11:59 lalatenduM so I might point to a bug which did not follow the guidelines not decided by us
12:00 lalatenduM s/did not/did/
12:00 kkeithley_ At this point you might want to use mainline/3.6/3.5 for examples instead of mainline/3.5/3.4. In case someone takes things too literally and omits ever cloning a bug against 3.6.
12:01 ndevos lalatenduM: ah, I guess some bugs from RHS that got cloned can have a Triaged keyword too
12:01 lalatenduM kkeithley, good point
12:03 kkeithley_ And I'm not a grammar expert, couldn't tell you the difference between a gerund and a direct object, but I think "bug triage" is (also) correct, and sounds better to my ear, than "bug triaging"
12:03 kkeithley_ IMO
12:04 lalatenduM kkeithley, agree, and fixed :), you have to refresh the browser
12:09 lalatenduM ndevos, cool
12:11 kkeithley_ lalatenduM++
12:11 glusterbot kkeithley_: lalatenduM's karma is now 18
12:16 lalatenduM kkeithley, ndevos thanks. Now I have to compose a mail :(, because of some reason my mails to gluster-users/devel regarding bug triage did not much response , so not very positive about the response
12:17 kkeithley_ I think it looks pretty good.
12:17 lalatenduM thanks, hoping for some good response this time
12:17 kkeithley_ yeah, don't feel bad if you don't get much response. that's pretty typical
12:18 kkeithley_ in my experience
12:19 ndevos lalatenduM: I wonder whats the best approach to get some responses... it must be super easy for new contributors to get started, but how to convince the, it's so easy?
12:21 lalatenduM ndevos, I think easy and interesting , you need a motivation to do something, so doing triage should appeal to community
12:21 lalatenduM not sure how to put it in a better way
12:21 ndevos lalatenduM: yeah, I think so too, it would be great if you can express that in your email :)
12:23 lalatenduM ndevos, haha, may be I should talk about little bit of open source / community philosophy, LOL
12:24 ndevos lalatenduM: oh, maybe! and include something like THIS IS HOW *YOU* CAN CONTRIBUTE TO GLUSTER
12:24 itisravi joined #gluster-dev
12:25 ndevos lalatenduM: and if that works out good, you can apply for the community leader position :D
12:26 * ndevos isnt sure if that has been filled already
12:26 lalatenduM ndevos, LOL good idea :p
12:28 kkeithley_ it's filled, he starts next week
12:29 * misc know the name
12:29 kkeithley_ johnmark has told me his name at least twice, but I don't remember
12:29 kkeithley_ I do remember that he was at SGI when they open sourced xfs
12:30 tdasilva joined #gluster-dev
12:35 kkeithley_ And more recently he was (one of) Adobe's open source community people
12:35 * kkeithley_ wonders why johnmark isn't jumping in here
12:52 Yuan_ joined #gluster-dev
13:11 bala joined #gluster-dev
13:13 hagarth joined #gluster-dev
13:20 JustinClift ndevos: The smoke and regression tests used by the Rackspace nodes are here: https://forge.gluster.org/gluster-patch-acceptance​-tests/gluster-patch-acceptance-tests/trees/master
13:20 Yuan_ joined #gluster-dev
13:21 JustinClift ndevos: It was based on the stuff on build.gluster.org, but has since been cleaned up and improved.
13:21 kkeithley1 joined #gluster-dev
13:30 shyam joined #gluster-dev
13:57 bala joined #gluster-dev
14:08 johnmark kkeithley_: Dave McAlister
14:11 wushudoin| joined #gluster-dev
14:26 JustinClift johnmark: We're announcing that publicly now?
14:27 skoduri joined #gluster-dev
14:27 johnmark JustinClift: he publicly announced he's leaving
14:28 johnmark we'll make a more public announcement when he starts
14:28 JustinClift Cool
14:28 JustinClift In that case: https://www.linkedin.com/in/davemc :)
14:32 bala joined #gluster-dev
14:32 lalatenduM JustinClift++
14:32 glusterbot lalatenduM: JustinClift's karma is now 8
14:33 lalatenduM JustinClift, thanks :) there are so-many of them in linkdin I was confused
14:34 JustinClift :)
14:35 JustinClift johnmark: How'd that GlusterFS distribution concept of yours go?
14:36 JustinClift Doesn't seem like anyone really picked it up?
14:39 ndevos JustinClift: ah, so I can add my new bug-version-git-branch check jenkins job there?
14:39 lalatenduM JustinClift, johnmark creating a distribution is lots of work, creating , maintaining so many packages itself is difficult
14:39 JustinClift ndevos: Yep
14:40 lalatenduM ndevos, regarding bug-version-git-branch, what abt bugs only applicable to stable branches but not applicable to master branch
14:40 JustinClift lalatenduM: I think that in early conversation about it, we discovered that someone (ndevos?) had already been separately packaging many of the projects.
14:40 JustinClift I'm not sure it was ndevos though :)
14:41 ndevos lalatenduM: the script does not check dependencies of bugs, it only checks if the change+branch match the bug+version
14:41 lalatenduM ndevos, cool
14:41 ndevos lalatenduM: so, a change for release-3.5 needs a bug that is filed against 3.5
14:42 lalatenduM ndevos, yeah make sense
14:42 ndevos lalatenduM: a change for master, needs a bug for mainline or pre-release
14:42 lalatenduM ndevos, cool :)
14:42 lalatenduM ndevos++
14:42 glusterbot lalatenduM: ndevos's karma is now 17
14:45 johnmark JustinClift: I think it was
14:45 johnmark ndevos: you're the one wiht all the packaged RPMs for stuff, right?
14:45 johnmark lalatenduM: and with dockit, we have an easy way to distribute this stuff
14:45 johnmark I think...
14:45 johnmark easier, I suppose
14:46 ndevos johnmark: uh, yeah, for many things, but maybe not the latest versions
14:46 johnmark ok
14:48 lalatenduM johnmark, may be with docker we can package them together , but from RPM distribution point of view it is difficult, at least we need one guy full time :)
14:49 johnmark lalatenduM: yeah, so forget RPM for the moment :)
14:49 johnmark lalatenduM: let's use dockit
14:50 ndevos JustinClift: okay, pushed the bug-check scripts
14:52 lalatenduM johnmark, hmm, that is interesting.
14:53 johnmark lalatenduM: could be.
15:02 JustinClift ndevos: Should the Rackspace nodes be updated with it, in case you want to run those scripts from them in a job?
15:02 ndevos JustinClift: those two scripts are triggered like smoke.sh
15:03 JustinClift ... which reminds me, I need to rebuild one of the slave nodes which harsha killed a while ago
15:03 ndevos JustinClift: but, they could be run on any slave that has the python-bugzilla package installed (currently build.g.o only?)
15:03 JustinClift ndevos: Sure.  What I'm saying is they're in the git repo at the moment, but not on any of the slave nodes.  Should I pull them down to the slave nodes
15:03 JustinClift k
15:04 JustinClift I'll refresh the slave nodes in a bit later on today.  It's overdue anyway :)
15:04 ndevos JustinClift: up to you, install python-bugzilla and run those scripts on other nodes, or keep them on b.g.o only, I have no preference
15:04 JustinClift k :)
15:05 JustinClift johnmark: In the GlusterFS distribution, I was kinda thinking it'd be an .ISO or similar.
15:05 ndevos JustinClift: btw, those are *Jenkins* scripts, like listed in the Jenkins webui
15:05 JustinClift johnmark: Thinking how to provide OSX client binaries to people for GlusterFS 3.6.x releases
15:06 JustinClift ndevos: Ahhh
15:06 JustinClift Gotcha
15:06 ndevos JustinClift: :)
15:10 deepakcs joined #gluster-dev
15:12 johnmark JustinClift: ISO would be fine
15:12 johnmark JustinClift: but a docker container would be sooooo trendy ;)
15:12 johnmark but yes, agree re: OSX client binaries
15:37 ndevos lalatenduM: can you check the last lines of http://build.gluster.org/job/compare-​bug-version-and-git-branch/29/console ?
15:37 edward2 joined #gluster-dev
15:37 lalatenduM ndevos, checking
15:37 ndevos lalatenduM: I've re-triggered the smoke test, I guess it should succeed next time
15:38 lalatenduM ndevos, oops , I missed to clone the bug, will do that now
15:38 ndevos lalatenduM++ thanks :)
15:38 glusterbot ndevos: lalatenduM's karma is now 19
15:39 ndevos lalatenduM: you can then update the commit message of your patch in the gerrit webui, and send a backport for the 3.5 bug later
15:40 bala joined #gluster-dev
15:41 lalatenduM ndevos, done , thanks for reminding me :)
15:41 lalatenduM ndevos, 3.5 bug now depends on mainline bug
15:41 ndevos lalatenduM: ok!
16:16 nishanth joined #gluster-dev
16:27 ndevos JustinClift: uhm, what to do when the NetBSD smoke test hangs? http://build.gluster.org/jo​b/netbsd-smoke2/139/console
16:34 ndevos I think I'll just hit the [x] button and click retrigger...
16:35 ndevos ah, that seems to do it :) http://build.gluster.org/jo​b/netbsd-smoke2/140/console
17:19 Humble joined #gluster-dev
17:27 bala joined #gluster-dev
17:30 johnmark misc: yt?
17:49 MacWinner joined #gluster-dev
18:34 bala joined #gluster-dev
18:38 _Bryan_ joined #gluster-dev
22:51 jobewan joined #gluster-dev
23:13 Yuan_ joined #gluster-dev

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