Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2016-08-22

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

All times shown according to UTC.

Time Nick Message
00:12 shyam joined #gluster-dev
00:50 inevity joined #gluster-dev
01:46 nishanth joined #gluster-dev
03:23 kshlm joined #gluster-dev
03:42 sanoj joined #gluster-dev
04:00 atinm joined #gluster-dev
04:08 itisravi joined #gluster-dev
04:28 itisravi joined #gluster-dev
04:29 shubhendu joined #gluster-dev
04:42 karthik_ joined #gluster-dev
04:53 ashiq joined #gluster-dev
05:02 kdhananjay joined #gluster-dev
05:13 ndarshan joined #gluster-dev
05:15 Manikandan joined #gluster-dev
05:19 atinm itisravi, can you please ack http://review.gluster.org/#/c/15188 ?
05:21 itisravi atinm: just did
05:22 atinm itisravi, thanks!
05:22 atinm itisravi++
05:22 glusterbot atinm: itisravi's karma is now 19
05:22 itisravi atinm: :)
05:22 itisravi np
05:31 aspandey joined #gluster-dev
05:33 rafi joined #gluster-dev
05:37 nishanth joined #gluster-dev
05:37 ramky joined #gluster-dev
05:42 kotreshhr joined #gluster-dev
05:55 nthomas joined #gluster-dev
05:56 asengupt joined #gluster-dev
05:59 Saravanakmr joined #gluster-dev
06:01 Bhaskarakiran joined #gluster-dev
06:12 hgowtham joined #gluster-dev
06:13 msvbhat joined #gluster-dev
06:14 mchangir joined #gluster-dev
06:22 Manikandan kotreshhr++
06:22 glusterbot Manikandan: kotreshhr's karma is now 7
06:27 atalur joined #gluster-dev
06:30 itisravi joined #gluster-dev
06:31 kaushal_ joined #gluster-dev
06:33 kdhananjay joined #gluster-dev
06:33 nigelb kshlm: Please welcome "Worker Ant" https://bugzilla.redhat.com​/show_bug.cgi?id=1368349#c2 :)
06:33 glusterbot Bug 1368349: unspecified, unspecified, ---, aspandey, MODIFIED , tests/bugs/cli/bug-1320388.t: Infrequent failures
06:34 ankitraj joined #gluster-dev
06:35 kshlm Welcome "Worker Ant"
06:35 kshlm nigelb++ Awesome!
06:35 glusterbot kshlm: nigelb's karma is now 26
06:36 nigelb kshlm: Though, now we have to be really careful about private bugs.
06:36 nigelb It does not have permission to see them.
06:37 kshlm That should be good.
06:37 kshlm The check-bug smoke job should fail for private bugs.
06:37 nigelb excellent.
06:39 rastar joined #gluster-dev
06:44 rafi joined #gluster-dev
06:45 s-kania joined #gluster-dev
06:46 devyani7 joined #gluster-dev
06:54 spalai joined #gluster-dev
06:56 poornima_ joined #gluster-dev
07:10 Muthu_ joined #gluster-dev
07:15 pur joined #gluster-dev
07:27 sanoj joined #gluster-dev
07:36 kdhananjay joined #gluster-dev
07:36 itisravi_ joined #gluster-dev
07:54 devyani7 joined #gluster-dev
08:00 hchiramm joined #gluster-dev
08:13 misc nigelb: so, the reboot worked fine ?
08:19 nigelb misc: I ended up not needing downtime at all.
08:19 nigelb So I didn't reboot.
08:19 nigelb I'll do an announce and do the reboot tomorrow.
08:32 itisravi joined #gluster-dev
08:34 aspandey joined #gluster-dev
08:34 kdhananjay joined #gluster-dev
08:37 nigelb ndevos: Would you find it useful to have the log from mock for the devrpms?
08:38 nigelb especially when the job fails and someone wants to debug it.
08:38 ndevos nigelb: yes, they are helpful, I think those are kept as artifacts already?
08:39 nigelb No :(
08:39 rastar atinm: do we need to update the op-version manually on a upgrade from 3.7.x to 3.8.x?
08:39 nigelb Or at least I think not?
08:39 nigelb Point me to where you see them?
08:40 ndevos nigelb: like https://build.gluster.org/job/devrpm-el7​/838/artifact/RPMS/el7/x86_64/build.log
08:40 nigelb aha
08:40 atinm rastar, op-version bump up always need a manual intervention, isn't it?
08:40 ndevos nigelb: from https://build.gluster.org/job/devrpm-el7/838/ -> expand and then somewhere between the rpms
08:40 nigelb I hadn't seen that until now :)
08:41 nigelb ndevos: Anyway, so I'm doing with converting the jobs on our CI to JJB. Next I'd like to tackle our jobs on Centos-CI.
08:41 nigelb This will break your workflow from clicking things.
08:41 nigelb Are you okay with that?
08:41 rastar atinm: oh ok, I thought if everything in the storage pool updated at the same time, we don't need
08:42 atinm rastar, nopes, that's not the case for gluster
08:42 rastar atinm++
08:42 glusterbot rastar: atinm's karma is now 65
08:42 ndevos nigelb: yeah, that should be fine, although I'm restructuring a few jobs, but I guess those are all under the nfs-ganesha project
08:43 nigelb okay. I'll post diffs of the generated xml for you to review.
08:43 nigelb There will be some change, because of defaults on the JJB end.
08:43 nigelb usually, they're sane defaults.
08:47 atalur joined #gluster-dev
08:58 kdhananjay joined #gluster-dev
08:58 itisravi joined #gluster-dev
09:08 atinm joined #gluster-dev
09:19 karthik_ joined #gluster-dev
09:21 atalur joined #gluster-dev
09:45 penguinRaider joined #gluster-dev
09:49 kdhananjay atinm: could you clarify what you mean by backport guidelines in http://review.gluster.org/15221 ?
09:49 ndevos kshlm, atinm, samikshan: could one of you write a user understandable note describing the problem of bug 1366813 ?
09:49 glusterbot Bug https://bugzilla.redhat.com:​443/show_bug.cgi?id=1366813 urgent, urgent, ---, sbairagy, MODIFIED , Second gluster volume is offline after daemon restart or server reboot
09:50 ndevos kshlm, atinm, samikshan: I plan to do an emergency release of 3.8.3 today, unless you guys do not think that is needed
09:51 atinm kdhananjay, https://gluster.readthedocs.io/en/lates​t/Developer-guide/Backport-Guidelines/ (specifically point 6)
09:51 kshlm ndevos, Okay.
09:51 atinm ndevos, I am in favor of 3.8.3, no objection from my end
09:52 ndevos atinm: heh :) I'm accepting backports where the old git-meta-tags have been removed completely, that is better than keeping them just-as-is
09:52 samikshan ndevos: I could work on the user understandable note
09:53 ndevos samikshan: great, many thanks! just put it in an etherpad, or post it to termbin.com or such
09:53 ndevos samikshan++
09:53 glusterbot ndevos: samikshan's karma is now 1
09:54 atinm ndevos, to me we should just stick to one approach
09:54 post-factum wow, worker ant!
09:54 post-factum nigelb++
09:54 glusterbot post-factum: nigelb's karma is now 27
09:55 ndevos atinm: oh, yes, I would really like it when everyone uses the same format, but some minor differences can be acceptible
09:55 samikshan ndevos: Is the commit message as found on https://bugzilla.redhat.com​/show_bug.cgi?id=1366813#c4 qualify as an user-understandable note describing the problem?
09:55 glusterbot Bug 1366813: urgent, urgent, ---, sbairagy, MODIFIED , Second gluster volume is offline after daemon restart or server reboot
09:55 post-factum atinm: regarding 15221. did you meant incorrect tags?
09:55 post-factum *mean
09:56 samikshan The last line from the commit message from that commit message can be excluded I guess
09:56 kdhananjay atinm: ndevos yeah, in fact that's what i've been following all these years. a pointer to the master patch is mentioned in the commit message of course ..
09:57 post-factum kdhananjay: putting right tags is quite hard before merging master patch
09:57 kdhananjay post-factum: Agree with you.
09:57 post-factum kdhananjay: but now it is merged :)
09:58 atinm post-factum, kdhananjay : I never felt an urgency of sending backports until the mainline patch is merged
09:58 ndevos samikshan: yeah, that is very close to what I need, but a little more details (only affects 2+ volumes), and maybe that 'force start
09:58 ndevos is a workaround
09:58 post-factum atinm: agree
09:58 samikshan ndevos: Ah sure.
10:00 ndevos samikshan: I want readers (sysadmins) of the note to think "oh, yes, I understand that they needed to do a new release already" and "wow, am I affected? how do I check/fix my env?"
10:00 kdhananjay fine, whatever it takes to get the patches merged i guess!
10:01 samikshan ndevos: Right. Got it :)
10:02 post-factum kdhananjay: i hope updating commit message won't retrigger regression tests now...
10:03 kdhananjay post-factum: i think it will.
10:03 kdhananjay :-/
10:03 post-factum nigelb: ^^ ?
10:03 kshlm ndevos, https://bugzilla.redhat.com​/show_bug.cgi?id=1366813#c6 , does this work?
10:03 glusterbot Bug 1366813: urgent, urgent, ---, sbairagy, MODIFIED , Second gluster volume is offline after daemon restart or server reboot
10:03 post-factum kdhananjay: the only test should be retriggered is some part of smoke that compares bug id and other stuff
10:05 post-factum kdhananjay: it seems we are late
10:06 post-factum kdhananjay: pranithk merged 15222
10:06 s-kania joined #gluster-dev
10:07 kdhananjay post-factum: to make matters worse, there doesn;'t seem to be an 'edit' button on  gerrit to change commit message from UI.
10:08 post-factum kdhananjay: git review + ./rfc.sh will do the trick ;)
10:09 samikshan ndevos: kshlm has already added a nice note describing the problem: https://bugzilla.redhat.com​/show_bug.cgi?id=1366813#c6. Does that work?
10:09 glusterbot Bug 1366813: urgent, urgent, ---, sbairagy, MODIFIED , Second gluster volume is offline after daemon restart or server reboot
10:09 post-factum i don't see edit button as well
10:09 post-factum kdhananjay: no
10:10 post-factum kdhananjay: i see it!
10:10 post-factum kdhananjay: you click on commit message in file list and have edit button
10:10 kdhananjay post-factum: ah, let me try
10:10 kdhananjay post-factum: ooh! nice. thanks :) post-factum++
10:10 glusterbot kdhananjay: post-factum's karma is now 29
10:11 post-factum yay
10:28 ndevos samikshan++ kshlm++ looks good, thanks!
10:28 glusterbot ndevos: samikshan's karma is now 2
10:28 glusterbot ndevos: kshlm's karma is now 103
10:30 msvbhat joined #gluster-dev
10:31 ndevos poornima_: if you have other gfapi or upcall related TODOs, please file bugs for them so that other maintainers/developers can take a look at them
10:51 pranithk1 joined #gluster-dev
10:51 suliba joined #gluster-dev
10:54 ashiq joined #gluster-dev
10:59 nigelb post-factum: It shouldn't lose the regression results
11:06 atinm joined #gluster-dev
11:10 post-factum nigelb: yep, everything is ok
11:10 spalai left #gluster-dev
11:13 post-factum pranithk1: you probably would like to merge 15173 as well
11:13 post-factum also, 14835 and 15167 need someone's attention
11:18 pranithk1 post-factum: I merged the first one. I am not well versed with gfapi, it is probably better ndevos|shyam takes a look at them...
11:18 post-factum pranithk1: thanks!
11:18 pranithk1 post-factum: They are the maintainers I think
11:18 post-factum pranithk1: anyway, second statement wasn't for you ;)
11:18 pranithk1 post-factum: cool
11:19 Bhaskarakiran joined #gluster-dev
11:23 post-factum pranithk1: should i backport 15070 to 3.8?
11:24 itisravi joined #gluster-dev
11:26 pranithk1 post-factum: sure why not :-)
11:26 post-factum pranithk1: dunno, maybe you want to do that very much ;)
11:26 post-factum pranithk1: will do
11:29 shyam joined #gluster-dev
11:35 nishanth joined #gluster-dev
11:41 ndevos kshlm, atinm, samikshan: how does http://review.gluster.org/#/c​/15234/1/doc/release-notes/3.8.3.md look to you?
11:42 Bhaskarakiran joined #gluster-dev
11:43 pranithk1 left #gluster-dev
11:44 rraja joined #gluster-dev
11:44 kshlm ndevos, One comment.
11:46 ndevos kshlm: yes?
11:47 kshlm Pasting my comment from the review, "Since you've swapped this and the paragraph below, the 'Also' doesn't make sense here. It needs to be on the next line."
11:48 ndevos kshlm: hmm, I prefer to have the problem statements together, and the workaround last
11:49 ndevos kshlm: shall I move that sentence before the "This is a regression ..." ?
11:49 kshlm Okay. It's just the 'Also' that is nagging me.
11:50 ndevos dropping the "Also, this" and replacing it with "It"
11:50 ndevos and moving it a little up
11:50 kshlm Sounds good.
11:52 ndevos kshlm: like http://termbin.com/uypr ?
11:59 kshlm ndevos, http://termbin.com/n1u7
12:00 kshlm The structure I'm trying to keep is 'Cause Problem Solution'
12:00 kshlm The cause is we aren't starting all bricks
12:00 kshlm 2 problems are that volumes aren't started after either reboot or restart of glusterd
12:00 kshlm Temporary solution is to force start.
12:00 ndevos kshlm: works for me too, but I'm not very fond of 2x "this" after eachother :)
12:03 ndevos kshlm++ thanks, I've taken your termbin.com version and pushed it
12:03 glusterbot ndevos: kshlm's karma is now 104
12:03 * kkeithley thinks cause, problem, solution is in the wrong order.
12:04 kkeithley The problem is that volumes aren't started after reboot
12:04 kkeithley the cause is a bug introduced in patch xyz
12:04 kkeithley the solution is ...
12:04 kkeithley IMO
12:05 ndevos kkeithley: sorry, too late!
12:07 mchangir joined #gluster-dev
12:09 kkeithley srsly?
12:10 kkeithley we can delete that tarball and you can run another release. ;-)
12:10 kkeithley but it's not that big a deal
12:10 pur joined #gluster-dev
12:11 ndevos nigelb: lets see the from-address of https://build.gluster.org/job/release/167/console when the mail arrives
12:11 kkeithley jenkins@build.gluster.org
12:12 ndevos still? I thought nigelb re-configured it to something like noreply@gluster.org that just /dev/null's any mail
12:12 * ndevos is still waiting for the email... it has to cross oceans and takes a while
12:14 kkeithley bleah.  79 files have unused variable warnings when #pragma GCC diagnostic ignored "-Wunused-variable" isn't leaking from the rpc/xdr headers.
12:14 ndevos oh, good cleanup!
12:15 kkeithley get ready for a flood of patches
12:35 post-factum kkeithley: one per warning?
12:35 nigelb ndevos: I didn't change the email ID. Just converted it to JJB. Is all well?
12:38 kkeithley 79 files. Some files have multiple warnings
12:40 kkeithley some of them I will bundle into one gerrit submission, e.g. xlators/storage/posix/src/{foo,bar,baz}
12:43 ndevos nigelb: oh, yes, all is well - I was just expecting an other from-address, but I understand that was not planned :)
12:45 ndevos kkeithley: it's not a strfmt warning, but maybe add it to https://github.com/gluster/glusterfs​-patch-acceptance-tests/blob/master/​centos-ci/gluster_strfmt/run-test.sh ?
12:49 kkeithley ndevos: yes, could do that. But once the pragmas don't leak out of the rpc/xdr headers then people should see the warnings and fix them, even on x86_64/amd64 builds.
12:49 kkeithley it's not like strfmts that nobody sees on x86_64
12:52 ndevos kkeithley: oh, yes, I guess this would be low-priority for the testing then
12:56 dlambrig joined #gluster-dev
12:57 msvbhat joined #gluster-dev
13:05 shyam joined #gluster-dev
13:08 nbalacha joined #gluster-dev
13:18 julim joined #gluster-dev
13:20 kkeithley I wouldn't say it's low priority.  That's not the same as the need for a separate test though.
13:23 ndevos kkeithley: have you ever thought about making the inode-ctx a void* instead of an 'unsigned long long int' or whatever it is?
13:24 kkeithley no, I haven't thought about that. Where is that?
13:26 nigelb ndevos: couldn't we do this as part of our normal smoke test? https://github.com/gluster/glusterfs​-patch-acceptance-tests/blob/master/​centos-ci/gluster_strfmt/run-test.sh
13:26 kkeithley ugh, we're not going to pass the smoke test with the unused variable warnings.
13:26 kkeithley https://build.gluster.org/job/smoke/29970/console
13:28 kkeithley merde
13:29 ndevos kkeithley: the functions from https://github.com/gluster/glusterfs/bl​ob/master/libglusterfs/src/inode.h#L179
13:30 ndevos kkeithley: I think it would be good if you send an email to the maintainers list and ask all of them to remove unused variables
13:31 kkeithley is there more context for this?  What about sizeof(void*) on a 32-bit machine.  Don't we need 64-bit?
13:31 ndevos kkeithley: "git review -d 14085" should get them a tree that they can use for compile-testing
13:32 ndevos kkeithley: yeah, some usage of it complains about type-casting pointers with incorrect size, its on 32-bit of course
13:32 hgowtham joined #gluster-dev
13:34 kkeithley and then wait....   grow a beard...   get old...    I already have fixes for the unused variables.  I was hoping to send them as follow-up patches instead of one nuclear patch.
13:35 ndevos kkeithley: you can easily change the order of the patches, please do not send them all in one!
13:35 nigelb Oh, I have a terrible idea I want to run past everyone :D
13:35 kkeithley I guess I can send the small patches and let http://review.gluster.org/14085 grow moss for a while longer
13:35 nigelb Instead of https://github.com/gluster/glusterfs-​patch-acceptance-tests/pull/47/files where someone will have to take a lot of energy to merge the netbsd diff
13:35 ndevos or that
13:35 nigelb I'm going to have different files
13:36 nigelb and slowly move the changes into one file.
13:36 nigelb in smaller, piecemeal reviews which doesn't break things.
13:36 post-factum and each patch will trigger bunch of regression tests?..
13:37 nigelb so we can't currently easily apply changes to the regression script on netbsd machines easily.
13:37 post-factum i'd call stop_machine() and apply one patch bomb at once ;)
13:38 nigelb post-factum: each patch will still trigger the same amount of regression tests.
13:38 nigelb Just that right now, regression.sh is what we *think* runs.
13:38 nigelb BUt that's not what actually runs for netbsd.
13:38 nigelb there are differences
13:39 msvbhat joined #gluster-dev
13:39 post-factum i mean, one huge patch instead
13:40 nigelb oh, I haven't had much review help going that way.
13:40 nigelb Plus, it's also easy to make mistakes that way.
13:40 post-factum ^(
13:40 nigelb Like, I could totally make a mistake that'll slip through review.
13:40 nigelb the netbsd scripts diverged in 2012.
13:41 nigelb phew, freebsd only has small changes.
13:47 mchangir joined #gluster-dev
13:48 nigelb kkeithley: If you want netbsd7-smoke, all you have to do is ask :)
13:48 nigelb Though, the person to ask might be Emmanuel.
13:49 kkeithley well, consider it asked for.
13:49 kkeithley my wondering was semi-rhetorical
13:49 nigelb I assumed there was a reason, honestly.
13:50 nigelb Do regression tests also run smoke?
13:50 nigelb Or rather, does a successful regression imply a succesful smoke?
13:58 shyam joined #gluster-dev
13:58 atinm joined #gluster-dev
14:02 kkeithley when you submit a patch several tests get run, including one called smoke.
14:02 mchangir ndevos, can you take a peek at the revised http://review.gluster.org/15184
14:03 nigelb I know the order of what gets triggered.
14:03 nigelb I'm asking about what actually runs in it.
14:03 kkeithley Independent of that you can set Verified+1 (after verifying, whatever that consists of) and then the regression tests are run.
14:03 nigelb does regression test also run posix compliance?
14:03 kkeithley no, the {CentOS,NetBSD}-regressions don't rerun the smoke tests.
14:04 kkeithley posix tests are part of the smoke tests
14:04 kkeithley IIRC
14:06 nigelb Right, so we should probably run them on netbsd7 as well.
14:14 hagarth joined #gluster-dev
14:26 kkeithley well, yes. I wonder why we're still bothering with netbsd6- anything. But if Emmanuel is doing the work to maintain it and wants it, then fine.  I'm just not sure why we do netbsd6-smoke and netbsd7-regression
14:26 kkeithley why not netbsd7 for both.
14:27 kkeithley and do I dare ask what release of freebsd the freebsd-smoke is running?
14:30 ndevos mchangir: looks good, except for the 'pacakage' typo
14:33 nigelb kkeithley: 10.1-RELEASE-p36
14:33 kkeithley ah, good, at least that's current
14:33 misc 11 is not out yet, no ?
14:33 nigelb no
14:34 nigelb production version is 10.3
14:34 kkeithley 11rc1 was last week or so
14:34 nigelb 11 release is in sep
14:36 msvbhat joined #gluster-dev
14:47 mchangir ndevos, rats!
14:56 shyam joined #gluster-dev
14:58 dlambrig joined #gluster-dev
15:07 wushudoin joined #gluster-dev
15:07 spalai joined #gluster-dev
15:16 dlambrig joined #gluster-dev
15:29 hchiramm joined #gluster-dev
15:30 hagarth joined #gluster-dev
15:33 dlambrig joined #gluster-dev
15:40 msvbhat joined #gluster-dev
15:41 Bhaskarakiran joined #gluster-dev
15:45 kotreshhr joined #gluster-dev
15:49 shyam joined #gluster-dev
15:50 rraja joined #gluster-dev
15:57 dlambrig joined #gluster-dev
16:04 obnox joined #gluster-dev
16:04 dlambrig joined #gluster-dev
16:08 dlambrig joined #gluster-dev
16:16 rafi joined #gluster-dev
16:18 kkeithley here comes the patch MIRV
16:19 rjoseph joined #gluster-dev
16:20 lalatenduM joined #gluster-dev
16:23 sac joined #gluster-dev
16:31 shyam joined #gluster-dev
16:31 pkalever joined #gluster-dev
16:35 rraja joined #gluster-dev
16:45 mchangir joined #gluster-dev
16:47 hagarth joined #gluster-dev
17:06 atalur joined #gluster-dev
17:07 atalur nigelb++ for the Worker Ant :)
17:07 glusterbot atalur: nigelb's karma is now 28
17:24 kkeithley here comes the second wave of MIRV patches
17:24 poornima joined #gluster-dev
17:43 shyam joined #gluster-dev
17:45 msvbhat joined #gluster-dev
17:49 ilbot3 joined #gluster-dev
17:49 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/
17:56 dlambrig joined #gluster-dev
18:03 rafi joined #gluster-dev
18:18 rafi joined #gluster-dev
18:44 ndevos kkeithley: ai, all the unused variable warning patches have component "core" in the subject :-/
18:44 kkeithley you don't say
18:44 rafi joined #gluster-dev
18:45 suliba joined #gluster-dev
18:56 dlambrig joined #gluster-dev
19:36 kotreshhr left #gluster-dev
19:46 ashiq joined #gluster-dev
19:59 dlambrig joined #gluster-dev
20:02 dlambrig joined #gluster-dev
20:12 shyam joined #gluster-dev
20:40 pranithk1 joined #gluster-dev
22:15 jobewan joined #gluster-dev

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