Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2015-05-10

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

All times shown according to UTC.

Time Nick Message
00:06 dlambrig_ joined #gluster-dev
00:09 shyam joined #gluster-dev
01:45 dlambrig_ left #gluster-dev
01:53 badone_ joined #gluster-dev
02:16 shyam left #gluster-dev
02:35 aravindavk joined #gluster-dev
02:44 rjoseph|afk joined #gluster-dev
03:01 krishnan_p joined #gluster-dev
03:02 krishnan_p hagarth, Good morning! Are we still merging backports to 3.7 branch?
03:03 hagarth krishnan_p: yes, for 4 hrs more
03:03 krishnan_p hagarth, thanks. I will continue merging backport patches until then.
03:04 hagarth krishnan_p: cool, thanks.
03:05 hagarth Gaurav_: the bitrot patches are failing compilation .. are you working on those?
03:05 krishnan_p hagarth, slides et al. ready for Gluster Summit '15?
03:06 hagarth krishnan_p: yet to start ;)
03:06 krishnan_p hagarth, I don't see "Publish and submit" button with the gerrit upgrade + old screen. Does gerrit auto merge things with Code-Review, Verified (2, 2)?
03:07 hagarth krishnan_p: you will get a submit button once you provide +2
03:07 hagarth CR +2, Publish and Submit cannot be done in one shot.
03:07 kotreshhr joined #gluster-dev
03:08 krishnan_p hagarth, thanks. I am positively UI challenged :(
03:09 hagarth krishnan_p: are your slides ready?
03:09 kotreshhr hagarth: Could you pull in these three patches?
03:09 kotreshhr http://review.gluster.org/#/c/10640/
03:09 kotreshhr http://review.gluster.org/#/c/10698/
03:09 kotreshhr http://review.gluster.org/#/c/10699/
03:10 hagarth kotreshhr: all passed regression now?
03:10 kotreshhr hagarth: Yes
03:11 hagarth kotreshhr: will do so in a bit
03:11 kotreshhr hagarth: Thanks!
03:14 krishnan_p hagarth, we have intial set of slides ready. still some distance to go :)
03:14 Gaurav_ hagarth: ya i am looking into it
03:14 krishnan_p hagarth, just defining the scope of the presentation can be tricky, especially with 4.0.
03:15 krishnan_p hagarth, we aim to keep the slides focussed on few issues during presentation and go berserk during Q and A.
03:15 krishnan_p hagarth, does that make sense?
03:16 krishnan_p review.gluster.com 503 error anybody?
03:20 hagarth krishnan_p: sounds like a good idea
03:20 hagarth krishnan_p: if it is a config error, use logout instead of login in the URL and it will work
03:20 hagarth krishnan_p: service unavailable error to me too now
03:20 krishnan_p hagarth, this is not configuration error. Aravinda's trick worked for me
03:21 hagarth krishnan_p: yes, logging into r.g.o to check if gerrit service is running.
03:23 hagarth seems to be back again
03:25 atinmu joined #gluster-dev
03:26 krishnan_p hagarth, thanks.
03:32 * krishnan_p is going to rebase #10339 - failed to merge due to merge conflict. Any objections?
03:35 atinmu hagarth, did we backport Jeff's patch?
03:35 hagarth atinmu: yes
03:35 atinmu hagarth, ok, was just cross checking as it was on me
03:38 atinmu krishnan_p, http://review.gluster.org/#/c/10664/4 ?
03:38 atinmu krishnan_p, I've given +1
03:39 atinmu backport of 10664 is also +1ed from me
03:41 krishnan_p atinmu, it has a merge conflict :(
03:42 atinmu Gaurav_, could u check??
03:42 hagarth Gaurav_: would you be able to resolve the bitrot compilation errors soon?
03:42 Gaurav_ hagarth: yes i am looking into it only
03:43 hagarth Gaurav_: please do and let us know if you need help. bitrot is one of the prime reasons why we moved the deadline.
03:43 Gaurav_ hagarth: ya sure i will update you soon
03:45 atinmu hagarth : is the compilation error persisted in master ?
03:45 hagarth atinmu: this is a new compilation error due to bitrot patches in review (not yet merged)
03:45 atinmu hagarth, ohh ok!!
03:46 Joe_f joined #gluster-dev
03:46 Joe_f hagarth: Hi hagarth
03:47 Joe_f hagarth: I have resolved the merge conflict http://review.gluster.org/10609, could you please merge . As it is important for 3.7
03:48 hagarth Joe_f: I thought you mentioned otherwise on Friday :)
03:48 rajesh joined #gluster-dev
03:48 hagarth Joe_f: let regression complete and I will merge that
03:49 Joe_f hagarth: yes but i checked the code in 3.7 it doesnt have the necessary check for tiering client pid with rebalancer pid ...
03:49 Joe_f hagarth: My bad
03:50 hagarth Joe_f: ok
03:50 atinmu did anyone retrigger http://review.gluster.org/#/c/10449/ ?
03:50 atinmu if not I am doing that
03:51 Joe_f hagarth: Thanks a ton! :)
03:51 hagarth Joe_f: np
03:55 atinmu hagarth, http://review.gluster.org/#/c/10647 can be merged
03:58 krishnan_p joined #gluster-dev
03:59 hagarth atinmu: ok
03:59 krishnan_p hagarth, power just came back
04:00 atinmu http://review.gluster.org/#/c/10339/ is been rebased
04:00 krishnan_p hagarth, I could do the honours for 10339. Thanks
04:00 atinmu krishnan_p, ^^
04:00 hagarth krishnan_p: great!
04:01 * hagarth bbiab
04:01 krishnan_p What a cinematic come back from 'power outage' :)
04:01 kotreshhr joined #gluster-dev
04:01 atinmu krishnan_p, can u merge 10647?
04:01 krishnan_p atinmu, sure
04:02 atinmu krishnan_p, http://review.gluster.org/#/c/10651 is also rebased
04:02 atinmu krishnan_p, sorry wrong link
04:02 * krishnan_p thinks that we should have glusterbot expand review #s into its permalink URL
04:03 atinmu http://review.gluster.org/#/c/10651 is also ready for merge
04:04 aravindavk once build is available only 3.7 bugs to be marked ON_QA? or master bugs also can be marked ON_QA?
04:05 atinmu aravindavk, master bugs can be marked ON_QA as per the nightly builds we release
04:06 atinmu aravindavk, hagarth can correct me if I am wrong
04:07 aravindavk atinmu: thanks
04:07 krishnan_p atinmu, could you run tier.t locally on the rebased 10339?
04:08 atinmu krishnan_p, is it failing?
04:08 atinmu krishnan_p, let me check
04:08 krishnan_p atinmu, I see that attach-tier implementation has changed since patchset 9.
04:08 krishnan_p atinmu, no, not that I know of. But it would be good to know if its only incidental and not an effect of auto-merge
04:09 krishnan_p atinmu, I want this checked since we are merging this without rerunning regression.
04:09 atinmu krishnan_p, sure
04:09 krishnan_p atinmu, I confirmed that the delta affects only attach-tier, which should be caught by tier.t
04:09 krishnan_p atinmu, thanks.
04:10 rafi joined #gluster-dev
04:11 krishnan_p Is anyone observing the load on r.g.o? I am sure it is being pounded like never before :)
04:11 nbalacha joined #gluster-dev
04:14 atinmu tier.t is taking ages to finish locally
04:14 krishnan_p atinmu, #10651 done.
04:15 atinmu krishnan_p, cool
04:17 krishnan_p rafi or Joe_f, could you rebase 10647. There is a merge conflict
04:17 Joe_f krishnan_p: sure
04:17 atinmu krishnan_p, tier.t passed for 10339
04:19 krishnan_p rafi, did you change attach-tier to use "hot-type" as the key in a recent change?
04:20 rafi krishnan_p: yes
04:20 krishnan_p rafi, I am asking this in context of #10339.
04:20 krishnan_p rafi, Ah that makes me comfortable with 10339, without a regression rerun.
04:20 krishnan_p rafi, thanks
04:21 dlambrig_ joined #gluster-dev
04:21 rafi krishnan_p: 10647 is dependent on 10675
04:21 krishnan_p rafi, OK. Did I merge it out of order?
04:22 rafi krishnan_p: nop ,there will be a merge conflict
04:22 dlambrig_ are we ok with 10647
04:22 krishnan_p and how about 10339. It still fails with merge conflict
04:23 krishnan_p dlambrig_, thanks.
04:23 rafi krishnan_p: first 10675 ,10647. 10677, 10678
04:24 raghu joined #gluster-dev
04:24 raghu Gaurav_: there?
04:24 Gaurav_ raghu: yes
04:24 atinmu the new interface is not that great to explain the dependency
04:24 raghu Gaurav_: was there a 2nd patchset sent to backport bitrot patches from master to 3.7?
04:24 atinmu the earlier was pretty easy
04:24 krishnan_p atinmu, I wasn't looking closely either :(
04:25 atinmu even me too
04:25 Gaurav_ raghu, yes, that scrubber frequency patch
04:25 raghu Gaurav_: Well the 2nd patch set is not even compiling properly
04:26 rafi krishnan_p: should i rebase 10647 ?
04:26 Gaurav_ raghu, 1st patch set was also not compiling properly
04:26 Gaurav_ raghu, could you check in your machine in 1st patch set branch. whether they are compiling or not
04:26 krishnan_p rafi, wait. let me try with the dependencies in mind.
04:27 rafi krishnan_p: ok, I'm in bus stop, with my laptop connected via mobile :)
04:27 Gaurav_ raghu, i saw on rackspace regression that all of your patch for 3.7 was hung yesterday night due to compilation error. i
04:27 rafi krishnan_p: I was travelling
04:28 rafi krishnan_p: I will come online after 30 mnt
04:28 rafi krishnan_p: is that fine ?
04:28 krishnan_p rafi, no issues.
04:28 rafi krishnan_p: thanks
04:29 krishnan_p dlambrig_, could you rebase 10647?
04:29 Gaurav_ raghu, all of that happen before sending my patch on top of your patches. so i thought regression was hung because of another reason so thougt of making my patch dependend of all of 3.7 bitrot patch so that regression can retrigger. now i saw that there is compilation error
04:29 dlambrig_ krishnan_p: yes will try now.
04:29 gem joined #gluster-dev
04:29 krishnan_p dlambrig_, thanks. it must be really late for you.
04:30 raghu Gaurav_: atleast it is compiling properly on my setup
04:30 dlambrig_ 12:30 here
04:30 raghu Gaurav_: let me give one more try
04:30 krishnan_p dlambrig_, that's late for me!
04:30 dlambrig_ :)
04:31 jiffin joined #gluster-dev
04:32 krishnan_p dlambrig_, don't upload 10339 after rebase yet.
04:33 krishnan_p dlambrig_, I am trying to understand the dependency, but r.g.o is taking forever. I might have tried merging patches out of order.
04:33 Gaurav_ raghu: can you checkout your change means http://review.gluster.org/#/c/10722/ checkout this patch from release3.7 branch and do make
04:33 dlambrig_ ok
04:33 Gaurav_ raghu: it will fail
04:33 krishnan_p dlambrig_, hmm. Looks like I did exactly that :(
04:34 raghu Gaurav_: Why that patch alone? Since it is a dependent patch on other patches....I will apply all those patches and check...
04:34 krishnan_p dlambrig_, 10328 ... (series) will be merged once 10328 passes regression.
04:34 jiffin joined #gluster-dev
04:34 dlambrig_ krishnan_p: has 10647 been merged?
04:34 * krishnan_p will be back after brunch.
04:35 krishnan_p dlambrig_, nope. merge conflict.
04:35 krishnan_p dlambrig_, 10647 isn't explicitly dependent on any of the patches. So, it must be a genuine merge conflict.
04:35 krishnan_p dlambrig_, are you aware of any patch that I should merge before 10647?
04:36 dlambrig_ to fix the conflict, should I do : git rebase origin/master? That brings up a conflict in run_tests.sh, which 10647 does not touch.
04:36 Gaurav_ raghu: i told about checkout not alone that patch and if you do checkout then all dependent patch will come.
04:36 krishnan_p dlambrig_, yeah that should do the trick
04:37 atinmu dlambrig_, probably u guys can drop a note about dependency tree to krishnan_p & hagarth
04:37 krishnan_p dlambrig_, it is likely that you are being affected by other merges
04:37 krishnan_p dlambrig_, only if the dependency is not visible in gerrit UI. Otherwise, it would be my own folly :)
04:37 * krishnan_p now needs to be at the table. will be back.
04:39 kotreshhr joined #gluster-dev
04:41 raghu Gaurav_: In 2nd patchset, patch which includes changes done in stub is not at all there
04:42 Gaurav_ raghu: could you tell me patch number for stub
04:42 raghu Gaurav_: 10722
04:44 Gaurav_ raghu: this patch is seperate patch . i didn't make this patch as dependent i told you to checkout all patch below this patch. because this patch doesn't have any dependency
04:44 Joe_f joined #gluster-dev
04:45 Gaurav_ raghu: you are checkouting all patch below this patch so this patch will not be there. even you checkout from this patch also then there will be compilation error
04:45 Joe_f hagarth: what happens if there are bugs in 3.7 .. how will the fixes go in ...?
04:45 Gaurav_ raghu: if you checkout from 10722 then also there will be error
04:46 raghu Gaurav_: why did u make it a separate patch? I sent all of them as dependent patches because backporting becomes easier...otherwise now they have to be accepted separately
04:46 raghu thats the reason I sent all the patches that are there in master as dependent patches. I could have applied and sent one by one. But we dont have time now.
04:47 raghu and when I applied those patches here on my setup it did not fail. It compiled properly
04:47 Gaurav_ raghu: i make my patch dependent of your top 2nd patch so 10722 is coming top of my patch
04:48 Gaurav_ raghu: see patch review.gluster.org/#/c/10721/
04:48 raghu Gaurav_: Well its not coming in 2nd patchset
04:48 hagarth Joe_f: we will keep fixing them in release-3.7
04:48 Gaurav_ raghu: and see right side on window
04:48 Gaurav_ raghu: you can see the dependency
04:48 Joe_f hagarth: :) super!
04:48 hagarth Gaurav_, raghu: can you please work on the minimal set of patches that we need for bitrot? we are about 120 minutes away from the cutoff.
04:49 Joe_f hagarth,: will there be a deadline for that also ?
04:49 hagarth Joe_f: yes, 3.7.1 will have its own schedule
04:49 raghu Gaurav_: yeah. I can see it now. Dont know why it was not there before.
04:49 Gaurav_ hagarth: ya sure
04:49 raghu hagarth: yeah, sure.
04:50 Joe_f hagarth: oh u mean we will keep fixing them in "release-3.7.1"
04:50 Joe_f hagarth: got it
04:50 hagarth Joe_f: release-3.7 is the branch for all 3.7.x releases
04:51 Joe_f hagarth: Okie Dokie .. .
04:52 nbalacha joined #gluster-dev
04:52 hagarth atinmu: http://review.gluster.org/#/c/10664/ is good to go?
04:52 dlambrig_ left #gluster-dev
04:52 Gaurav_ raghu, hagarth: i got the error now.
04:53 Gaurav_ raghu, hagrath: i think error is in patch http://review.gluster.org/#/c/10380/ which has merge in master having file tw.c and tw.h
04:53 Gaurav_ and patch review.gluster.org/#/c/10716/ for release 3.7 branch dont have these file
04:53 Gaurav_ raghu: can you look :)
04:53 atinmu hagarth, it looks good to me, but does that introduce compilation error?
04:54 atinmu Gaurav_, which patch introduces compilation error?
04:54 hagarth atinmu: it is very confusing atm :)
04:54 atinmu hagarth, :)
04:55 raghu Gaurav_: that patch is part of this patchset
04:56 atinmu I seriously feel from next release onwards we should introduce merge windows to reduce rebase efforts
04:56 raghu hagarth: gerrit applies all the patches below before running test on a patch right (in case of dependent patches)
04:56 hagarth raghu: right
04:56 Gaurav_ raghu: ya that patch is parth of this patch set. but i dont know why you forgot to include tw.c and tw.h file in  http://review.gluster.org/#/c/10716/
04:56 Gaurav_ can you open http://review.gluster.org/#/c/10716/ patch set
04:56 Gaurav_ and see the diff 1..2
04:56 hagarth atinmu: what we are now having is a merge window.. not sure how we can avoid rebasing
04:56 Gaurav_ there will be no difference
04:57 Gaurav_ but if you open same patch which merge in master http://review.gluster.org/#/c/10380/
04:57 atinmu hagarth, merge window for every features?
04:57 Gaurav_ having tw.c and tw.h file
04:57 hagarth atinmu: that will serialize & slow down things .. we can't do that.
04:57 Gaurav_ so i just copy tw.c and tw.h file
04:57 atinmu hagarth, my intension is multiple features should not race at merge
04:57 Gaurav_ and compile again
04:57 Gaurav_ and error was gone
04:58 Gaurav_ atinmu, all the bitrot patch for release 3.7 causing error.
04:58 hagarth raghu: I don't see a backport of your anon fd handling patch in release-3.7 ?
04:58 raghu hagarth: I have sent it (patchid- 10722)
04:58 Gaurav_ raghu: did you get the error now
04:59 hagarth raghu: ok
04:59 raghu Gaurav_: I got your point. I remember git adding all the untracked files and later adding them in makefile as well. Dont know how it missed. Can you do git add of those files to that patch and send the patches again?
05:00 Gaurav_ raghu: ya sure :)
05:05 Gaurav_ raghu: done.
05:05 raghu Gaurav_: if u need any help let me know
05:05 Gaurav_ raghu: its done
05:05 raghu Gaurav_: ohh. cool :)
05:05 raghu Gaurav_++
05:05 glusterbot raghu: Gaurav_'s karma is now 1
05:06 Gaurav_ raghu: i have included tw.c and tw.h file in my patch it self.
05:07 raghu Gaurav_: I think you have missed to add 10722 (i.e. the patch for the stub changes)
05:07 Gaurav_ raghu, oh ya
05:07 raghu Gaurav_: Please add it at the top of all the changes
05:07 Gaurav_ i m sending once more.
05:07 raghu Gaurav_: sure.
05:09 * atinmu is logging off
05:09 * atinmu feels he is done with his part
05:10 atinmu is anything awaiting on me?
05:12 * krishnan_p is back
05:12 rafi joined #gluster-dev
05:14 krishnan_p rafi, could you rebase  10647?
05:14 rafi krishnan_p: I will do
05:15 krishnan_p rafi, thanks.
05:16 rafi krishnan_p: what about 10677 and 10678
05:17 Gaurav_ raghu: could you check wheter 10722  need tw.c and tw.h file . while making dependent and doing ./rfc.sh its showing that no change in 10722 hence not proceeding further
05:17 krishnan_p rafi, are they independent?
05:18 krishnan_p 10677 hasn't passed netbsd yet.
05:18 rafi 10677 follows 10678, they are dependent
05:18 rafi krishnan_p: should i rebase it ?
05:20 Gaurav_ krishnan_p, ping   http://review.gluster.org/#/c/10680/      and http://review.gluster.org/#/c/10664/ need to merge. could you look into them
05:20 Gaurav_ krishnan_p, glustserd crash patch
05:21 rafi joined #gluster-dev
05:22 krishnan_p rafi, I am not sure if there is a way to (re)trigger a netbsd build without rebasing
05:22 rafi krishnan_p: ok i will rebase it
05:22 krishnan_p hagarth, is there a (simple) way to find if a particular patch ran netbsd regression via jenkins?
05:22 krishnan_p rafi, hold on a bit.
05:22 rafi krishnan_p: 10647 is completely out
05:22 krishnan_p rafi, if hagarth has a simpler way to know if netbsd ran regression  but just didn't update gerrit ...
05:23 krishnan_p rafi, ok, rebase away :)
05:23 hagarth krishnan_p: I don't think it is very easy
05:23 krishnan_p hagarth, OK.
05:23 rafi krishnan_p: I will try to rebase it
05:24 rafi joined #gluster-dev
05:24 krishnan_p Gaurav_, 10680 netbsd regression not complete. Ping me when all regressions have passed.
05:25 AndChat|589056 joined #gluster-dev
05:25 rafi krishnan_p: seems like a complete re work required
05:25 rafi krishnan_p: 10647
05:25 Gaurav_ krishnan_p, ya sure
05:25 hagarth AndChat|589056: neat way to get on IRC :)
05:25 krishnan_p rafi, ouch.
05:25 krishnan_p hagarth, what's that?
05:26 hagarth krishnan_p: raghu has got on here through AndroidChat IRC client
05:26 krishnan_p hagarth, I get it now. Android client
05:26 krishnan_p raghu, hagarth, :)
05:29 Joe_f nbalacha:hi nithya this is the one for 3.7
05:29 Gaurav_ krishnan_p, 10680 net bsd test hang because of i think net bsd machine problem. i have retrigger net bsd regression test for 10680
05:29 Joe_f http://review.gluster.org/#/c/10742/
05:30 krishnan_p Gaurav_, hagarth http://review.gluster.org/#/c/10664/ is not getting merged. I initiated a merge. It has been a while, I don't see the effect. Any ideas what could have gone wrong?
05:30 krishnan_p Gaurav_, OK. ping me when that runs successfully
05:30 Gaurav_ krishnan_p, ok sure
05:31 Gaurav_ krishnan_p, http://review.gluster.org/#/c/10664/ showing merge status
05:31 Gaurav_ and in All->merged window its showing merge
05:32 Gaurav_ krishnan_p, it seems some issue with gerrit itself
05:34 nbalacha Joe_f , ok will take a look
05:35 nbalacha Joe_f, you might want to abandon the other one
05:35 Joe_f hagarth: there was a merge conflict again in http://review.gluster.org/10609 i have resolved it  + there is another patch http://review.gluster.org/10742
05:35 krishnan_p Gaurav_, 10664 merged
05:35 Joe_f sure
05:35 Joe_f i will go that
05:36 nbalacha Joe_f, done.
05:36 Joe_f nbalacha: thanks :)
05:36 Gaurav_ krishnan_p, yes it merged
05:41 nbalacha Joe_f, take a while to get used to it :)
05:42 hagarth Joe_f: will you abandon http://review.gluster.org/#/c/10741/ ?
05:44 jiffin joined #gluster-dev
05:48 krishnan_p Gaurav_, raghu please confirm if these are the only patches - http://review.gluster.com/#/q/status:open+cha​nge:+10512+OR+change:+10602+OR+change:+10680
05:48 Joe_f nbalacha: yes i wil
05:48 krishnan_p ... needed for bitrot
05:49 krishnan_p Joe_f, rafi - could you confirm that these are the only patches for tiering/glusterd - http://review.gluster.com/#/q/status​:open++change:+10328+OR+change:+1033​9+OR+change:+10449+OR+change:+10647
05:49 glusterbot krishnan_p: http://review.gluster.com/#/q/status:open's karma is now 1
05:50 Joe_f krishnan_p: yes confirm
05:50 Gaurav_ krishnan_p, there are more patches in 3.7 branch also
05:51 Gaurav_ krishnan_p, that are waiting for regression
05:51 hagarth a new landing page in documentation for developers - http://gluster.readthedocs.org/en/late​st/Developer-guide/Developers%20Index/
05:53 Gaurav_ hagarth: in release 3.7 spurious failure is going on
05:53 Gaurav_ Test Summary Report
05:53 Gaurav_ -------------------
05:53 Gaurav_ ./tests/basic/ec/ec-5-1.t (Wstat: 0 Tests: 289 Failed: 37)
05:53 Gaurav_ Failed tests:  73, 91, 105, 147-164, 166-167, 171-172
05:53 Gaurav_ 176-177, 181-182, 186-187, 198, 207, 216
05:53 glusterbot Gaurav_: -----------------'s karma is now -2
05:53 Gaurav_ 225, 234, 243
05:54 hagarth Gaurav_: have retrigerred that test
05:54 Gaurav_ hagarth: ok
05:54 hagarth from my discussion with Pranith .. any test in ec can fail
05:54 Gaurav_ we can make them as a bed test
05:55 Gaurav_ for release 3.7
05:57 hagarth Gaurav_: please mark both 5-1.t and 12-4.t as bad tests and send a patch
05:57 hagarth Gaurav_: please send to both master and release-3.7
05:58 Gaurav_ hagrath: ok
05:59 hagarth Gaurav_: ec-3-1.t as well
06:00 krishnan_p joined #gluster-dev
06:01 krishnan_p Gaurav_, if there are other patches to be merged, please update https://public.pad.fsfe.org/p​/pending_gluster_3.7_patches
06:01 Gaurav_ hagarth, krishnan_p, ok sure
06:01 krishnan_p It would help if you pasted the change no. of all the patches space separated.
06:03 krishnan_p gerrit search term for viewing a list of patches - "status:open change:<id1> OR change: <id2> ..."
06:09 krishnan_p Gaurav_, I am only merging glusterd related patches
06:09 krishnan_p in fact 10602 isn't a glusterd patch
06:09 Gaurav_ krishnan_p, ok
06:09 Gaurav_ krishnan_p, yes
06:10 krishnan_p hagarth, can we remove all striked through patches from the etherpad (it's anyway version controlled or not?)
06:10 krishnan_p hagarth, mining through the etherpad is becoming difficult.
06:10 hagarth krishnan_p: we can.. etherpad has a timeline and we should be able to revert back
06:11 krishnan_p hagarth, may I?
06:11 hagarth Gaurav_: please backport bad tests to release-3.7
06:12 hagarth krishnan_p: sure, go ahead
06:12 Gaurav_ hagarth: ok doing
06:12 hagarth yay! one of rafi's patches passed regression
06:14 Gaurav_ hagarth: done
06:14 hagarth Joe_f: can you confirm if this looks good to you in gerrit - http://review.gluster.org/10328 ?
06:14 hagarth Gaurav_: thanks, merged
06:15 gem joined #gluster-dev
06:16 krishnan_p hagarth, how does the etherpad look now? B-)
06:17 hagarth krishnan_p: lean and mean! :)
06:18 Joe_f joined #gluster-dev
06:19 krishnan_p hagarth, logging patches seem to have not passed netbsd :(
06:20 hagarth krishnan_p: worst case we can defer logging to 3.7.1
06:21 krishnan_p hagarth, OK.
06:23 hagarth raghu: around?
06:29 krishnan_p ndevos, you have commented that this should not be taken in for 3.7 - http://review.gluster.org/#/c/10731/
06:29 krishnan_p ndevos, is that your stance at this point? If so, Gaurav_ could you please remove it from the etherpad for the time being?
06:30 Gaurav_ krishnan_p, ya will remove it
06:31 krishnan_p Gaurav_, thans.
06:33 krishnan_p hagarth, patch pending scoreboard: tiering - 3; bitrot - 4
06:34 krishnan_p hagarth, is that what you see too?
06:34 krishnan_p hagarth, netbsd hasn't run for http://review.gluster.com/#/c/10680/
06:36 raghu hagarth: yeah
06:38 hagarth raghu: mind reviewing the uss patch from rafi?
06:38 hagarth krishnan_p: right .. bitrot has a few more from raghu that hasn't been updated in the etherpad
06:40 Gaurav_ now bitrot have 10 patch on ether pad
06:42 gaurav|away nick /Gaurav_|away
06:42 jiffin1 joined #gluster-dev
06:43 raghu hagarth: sure
06:46 rafi joined #gluster-dev
06:47 hagarth raghu: also need reviews for http://review.gluster.org/#/c/10602/
06:47 raghu hagarth: reviewed it. Have provided a comment
06:48 hagarth raghu: ok
06:48 hagarth raghu: in that case, should we push this uss patch to 3.7.1?
06:52 rafi joined #gluster-dev
06:54 raghu hagarth: Its more of a coding comment instead of functionality comment. In uss if statfs comes on ".snaps", it is sent to the regular volume with path being set to root i.e. "/". There he was doing loc.inode = inode->table->root. Then he was doing loc.inode->ia_type = IA_IFDIR; I have told in the comment that is not necessary as the root inode can be refed in loc.inode and later doing loc_wipe after STACK_WIND will
06:54 raghu unref it.
06:55 raghu But still I am ok with pushing it for 3.7.1
06:55 raghu I have given +1 to 10602. But you also take a look before merging it.
06:56 rafi hagarth: How many more time left ?
07:00 kripper joined #gluster-dev
07:04 jiffin joined #gluster-dev
07:06 rafi hagarth: http://review.gluster.org/#/c/10678 can you merge this ?
07:20 hagarth rafi: not much, maybe another 40 minutes or so
07:21 hagarth rafi: need to wait a regression run to complete for 10678
07:27 rafi1 joined #gluster-dev
07:35 Gaurav_ joined #gluster-dev
07:42 anrao joined #gluster-dev
07:57 Joe_f joined #gluster-dev
08:00 * ndevos waves from the breakfast table _o/
08:14 rafi joined #gluster-dev
08:21 hagarth current state:
08:22 hagarth 1. awaiting bitrot & tiering patches to complete regression on 3.7
08:22 hagarth 2. consolidated information about patches here - https://public.pad.fsfe.org/p​/pending_gluster_3.7_patches
08:22 jiffin joined #gluster-dev
08:36 Gaurav_ hagarth: could you look into http://review.gluster.org/#/c/10680/  . its already merge in master branch. and its passed nedbsd regression test also but not showing in patch http://build.gluster.org/job/rackspace-n​etbsd7-regression-triggered/5234/console
08:40 ndevos Gaurav_, hagarth: I'll fix that
08:40 Gaurav_ ndevos, thnx
08:40 ndevos Gaurav_: the NetBSD vote is there too now :)
08:41 * ndevos leaves the merging to hagarth
08:41 Gaurav_ ndevos, its surprising delay
08:41 Gaurav_ ndevos, its passed regression quite before :)
08:44 ndevos Gaurav_: I pushed the vote again, it failed to send the vote before
08:45 Gaurav_ ndevos, cool, thank you :)
08:50 jiffin joined #gluster-dev
09:03 kotreshhr left #gluster-dev
09:06 jiffin joined #gluster-dev
09:15 gem joined #gluster-dev
09:18 gem joined #gluster-dev
09:19 hagarth Gaurav_: can you please avoid refreshing entire patchsets?
09:20 hagarth all patches would restart regression runs and we would need to await longer
09:20 Gaurav_ hagarth: there are dependency
09:20 Gaurav_ hagarth: all regressing test was failing
09:20 hagarth Gaurav_: even in that case, it would be better to wait
09:20 hagarth Gaurav_: for instance, regression run was not complete for this patch http://review.gluster.org/#/c/10716/
09:21 hagarth and it was in a fairly mature state when I checked last
09:21 hagarth now it will start from scratch
09:21 Gaurav_ hagarth: this is because tw.c and tw.h file i used in http://review.gluster.org/#/c/10739/ patch but it actually needed by review.gluster.org/#/c/10716 patch. it was failing compilation
09:22 hagarth Gaurav_: wasn't this sorted out a few hours back?
09:23 Gaurav_ hagarth: it was shorted out. but tw.c and tw.h file should place in review.gluster.org/#/c/10716 patch
09:23 Gaurav_ previously i put in top of dependency patch
09:24 Gaurav_ but gerrit was considering individually
09:24 hagarth Gaurav_: right, that cannot work
09:24 Gaurav_ now every thing is fine. i have put dependency file in right patch. menas parent of dependency patch. previously it was at child of depency patch
09:25 hagarth Gaurav_: we expect each patch to compile fine on top of dependencies.. so anything needed by a later patch in your dependency chain would have to be part of your earlier patches.
09:25 hagarth Gaurav_: but we have gone past the deadline :-/
09:26 Gaurav_ hagarth: i should have observed that earlier. just one hour ago none of test case was running
09:27 Gaurav_ then i retrigger all test case
09:27 Gaurav_ i saw comilation failure
09:31 hagarth Gaurav_:  this is unfortunate, we will possibly need more than 2 hours to determine whether bitrot patches can be merged.
09:33 Gaurav_ hagarth: yes we need to wait for them. this time all of the regression test should pass.
09:55 soumya joined #gluster-dev
09:56 jiffin joined #gluster-dev
10:08 kripper joined #gluster-dev
10:25 anrao joined #gluster-dev
10:48 * hagarth will bbiab .. state transition
11:40 jiffin joined #gluster-dev
11:55 raghu joined #gluster-dev
11:57 dlambrig_ joined #gluster-dev
11:57 dlambrig_ left #gluster-dev
12:04 Gaurav_ raghu: 10716,10717,10718,10719,10720,10721, passed regression test. but few regression failing in net bsd due to i think infrastructure problem. that net bsd regression already passed on master branch. these patch can be merge now.
12:04 raghu Gaurav_: what about the last patch? i.e. 10722? It has failed regression right?
12:05 raghu because of compilation failure on one of the dependent patches
12:05 Gaurav_ raghu: that patch need to rebase once that patch merge. see the niels comments on that patch
12:05 raghu there is no need of rebasing I think. Because the code changes are in a different file
12:07 raghu Gaurav_: Can regression be triggered on that patch alone? I hope it does not run regression on the dependent patches
12:07 Gaurav_ raghu, its showing  error >>>   make[3]: *** No rule to make target `tw.c', needed by `libglusterfs_la-tw.lo'.  Stop.
12:08 raghu How can it show if you have addressed that compilation error??
12:08 Gaurav_ raghu, now that file is tw.c and tw.h file available once above mentioned patch merged
12:08 raghu Patch need not be merged for them to be available. You just have to apply those patches properly before 10722.
12:10 raghu In this case I assume gerrit is still applying patches of older patchset (where tw.c and tw.h were not there) when it has to run regression on 10722
12:11 Gaurav_ raghu: ya its seems gerrit is applying older patches
12:12 Gaurav_ raghu: should we merge above mention patches now
12:12 raghu Gaurav_: When you had sent your latest patchset, was 10722 part of that?
12:12 raghu is hagarth around?
12:12 Gaurav_ raghu: no, because gerrit was showing error
12:12 raghu what? what error?
12:12 raghu so u skipped that patch and sent the remaining patches?
12:13 Gaurav_ raghu: gerrit was showing that there is no actual dependecny bw 10722 and rest of your dependent patches. so 10722 need to skiped
12:13 raghu Gaurav_: where was it showing that error?
12:13 Gaurav_ raghu: i also discuss the same problem with niels yesterday night
12:14 Gaurav_ ndevos:
12:14 Gaurav_ raghu: so atlast we need to remove 10722 patch and need to send remaining patches. thats what i did.
12:15 raghu Gaurav_: the only reason I sent all those patches at once is to get them merged as soon as possible. Not because there is a code dependency. I wanted to avoid this situation. Thats why I had sent it along with other patches, so that I dont have to keep on waiting for each patch to get merged and then send the next patch.
12:15 Gaurav_ raghu: while submitting patches. i have pasted output to ndevos also
12:16 ndevos Gaurav_, raghu: which patch fixes that tw.c compilation issue? we might be able to merge that, rebase 10722 after that, and test only that one?
12:16 Gaurav_ ndevos, 10716
12:17 nbalacha joined #gluster-dev
12:17 ira joined #gluster-dev
12:17 ndevos raghu, Gaurav_: can you review those changes too, a +2 would help so that I can merge it
12:18 raghu ndevos: yeah we can do that. I am not sure if we have that much time.
12:18 ndevos raghu: I also dont know, hagarth seems to be offline, so I guess there is still some time ;-)
12:19 ndevos raghu: could merging only 10716 cause any problems?
12:19 raghu ndevos: hmm. Then probably we can merge the remaining patches and then resend 10722 independently????
12:19 Gaurav_ raghu, ndevos, i dont think regression test will fail now. because all of patches have merge in master with passing regression test. we just backported
12:19 ndevos raghu: that was what I was thinking
12:20 raghu ndevos: yeah, we can do that. The only reason I want to see 10722 to be in is it handles some issues that bitrot has with nfs
12:21 ndevos raghu: sounds like a safe approach to me then, and if 10722 is missing from the release, we can make a know-issue-note for it
12:21 raghu ndevos: ok. Lets merge the other patches and I will send 10722 right away
12:22 ndevos raghu: okay, I'm just waiting for those +2 reviews now
12:23 raghu Gaurav_: can you +2 those patches? it seems the patches are in my name. So a +2 will help to get it merged
12:24 Gaurav_ raghu, ok
12:25 ndevos raghu, Gaurav_: what is the correct order of the patches, start with 10716 and then just +1 ?
12:25 Gaurav_ ndevos, yes
12:25 raghu ndevos: yeah. 10716-10721
12:25 ndevos Gaurav_: ok
12:25 Gaurav_ raghu, check once
12:26 Gaurav_ i mean order check.
12:26 Gaurav_ raghu, gave +2
12:28 ndevos raghu: 10716 has been merged, you can rebase 10722 now
12:28 shyam joined #gluster-dev
12:28 Gaurav_ ndevos, raghu, rest of the patches should be merge
12:29 ndevos Gaurav_: yes, I'm going through them now
12:29 raghu ndevos: ok. Please merge all the remaining patches as well (till 10721)
12:30 ndevos raghu, Gaurav_: done - only 10722 is now waiting on you and the regression tests
12:30 raghu I am sending now
12:30 ndevos raghu, Gaurav_: what about 10739 ?
12:30 Gaurav_ ndevos, that can also be merge
12:31 Gaurav_ raghu, can you give +1 to 10739
12:31 Gaurav_ raghu, its merged in master branch
12:31 ndevos raghu: make that a +2 please :)
12:32 ndevos Gaurav_: why is libglusterfs/src/tw.c part of 10739 in release-3.7, but not in the master branch?
12:32 raghu I am sending 10722 now
12:32 * ndevos wonders if things are broken now...
12:33 Gaurav_ ndevos, i should have remove tw.c file. because of previous compilation failure i add these file in this patch
12:33 Gaurav_ ndevos, can it be replace if you merge this patch
12:34 ndevos Gaurav_: I dont know how merging would handle it, I'll try it out locally
12:34 Gaurav_ ndevos, ok
12:41 raghu ndevos: I have rebased and sent 10722. I hope it passes all the tests soon and makes it to 3.7
12:41 raghu fingers crossed
12:42 kdhananjay joined #gluster-dev
12:43 Gaurav_ raghu: could you ack 10739
12:45 ndevos raghu: yes, 10739 needs a +2 before I can merge it, local merging is fine, the tw.{c,h} changes do not conflict
12:55 ndevos Gaurav_: 10722 needs a review too :)
12:55 Gaurav_ ndevos, i will review it
12:57 nbalacha joined #gluster-dev
13:03 hagarth joined #gluster-dev
13:04 hagarth raghu: ping, around?
13:06 ndevos Gaurav_: just merged 10739 too, the backport looks good to me
13:06 Gaurav_ ndevos, i saw, thank you
13:06 hagarth Gaurav_: what is pending now?
13:07 ndevos hagarth: I've been merging the bitrot changes, and have updated the etherpad with all current statusses
13:07 Gaurav_ 10722
13:07 Gaurav_ hagarth: only 10722.
13:07 hagarth ndevos: great thanks, my state transition is now complete and can be online for the next 35 minutes or so :)
13:08 ndevos hagarth: hah, ok! but a regression test takes a little longer than 35 minutes :)
13:08 hagarth ndevos: unfortunately that happens to be the case
13:08 hagarth ndevos: would you mind creating a 3.7.0beta2 tag once 10722 is in?
13:09 ndevos hagarth: sure, I can do that
13:09 hagarth ndevos++ thanks!
13:09 glusterbot hagarth: ndevos's karma is now 123
13:09 ndevos hagarth: np!
13:09 ndevos Gaurav_: why a +1 on 10722, and not a +2?
13:10 Gaurav_ ndevos, ok giving +2
13:10 hagarth :)
13:10 ndevos Gaurav_: ok, thanks!
13:10 hagarth what about the snapshot patches? move them to 3.7.1?
13:11 ndevos hagarth: I do not think anyone reviewed those yet, but regression tests are running for them
13:12 hagarth ndevos: I saw a few +1s to one of the patches (statfs in uss) .. but again it is required primarily for samba
13:13 ndevos hagarth: I +1'd one, but I dont know much about snapshot internals to give it a +2
13:14 hagarth ndevos: raghu felt it was ok .. he suggested a minor change which can come by subsequently too
13:14 ndevos hagarth: yeah, that is 10358 for the master branch
13:15 hagarth ndevos: let us get those in. Avra will also feel happier :).
13:15 ndevos hagarth: I think it should be fine to include it, regressions passed, so it should not break anything too much
13:16 hagarth ndevos: right, the fact that we haven't heard many problems from uss indicates that it is safe to be adventurous there ;)
13:16 ndevos oh, and I rebased 10740 just a while back, regressions failed, but should be fixed with the latest changes
13:17 ndevos hagarth: hehe, yeah, that is a nice way to express it :D
13:18 kkeithley1 joined #gluster-dev
13:20 ndevos hagarth: are you checking 10358 and you want to merge it?
13:20 kkeithley1 joined #gluster-dev
13:21 hagarth ndevos: checking
13:23 shyam joined #gluster-dev
13:26 hagarth ndevos: merged
13:27 ndevos hagarth: ok!
13:28 ndevos hagarth: I'll let the regressions for 10740 and 10722 finish, and will merge them when done
13:28 hagarth ndevos: ok
13:33 ndevos hagarth: other than those 2 patches, nothing needs to get in, right?
13:34 rafi joined #gluster-dev
13:34 hagarth ndevos: that should be it.
13:34 ndevos hagarth: ok, very good
13:35 hagarth finally, the end is in sight ;)
13:36 ndevos only a little later than planned...
13:37 hagarth yeah .. there is some room for improvement
13:50 badone_ joined #gluster-dev
13:54 rafi ndevos: How is going ?
13:56 ndevos rafi: I'm doing good, how are you>
13:56 ndevos ?
13:57 ndevos rafi: regression tests are proceeding, patches 10740 and 10722 are the ones that need to pass before we're done with beta2 (= release?)
13:58 rafi ndevos: I'm good
13:58 rafi ndevos: ok
13:58 rafi ndevos: when are leaving for Barcelona ?
14:00 ndevos rafi: tomorrow morning, my flight only takes a little over 2 hours
14:01 ndevos 2 hours and 15 minutes, to be precise
14:01 rafi ndevos: great
14:02 rafi ndevos: What about our release ? Are we taking any new ?
14:03 ndevos rafi: no, we're waiting for those two patches to pass regression testing, then I'll tag the release
14:05 rafi ndevos: if you could take http://review.gluster.org/#/c/10677/ , that would be great :)
14:06 ndevos rafi: it really should have been reviewed... and it is labelled as "Cannot Merge" for some reason
14:08 rafi ndevos: i resolved the conflict, and send the new patch set
14:09 ndevos rafi: its a little late for that, how critical it this?
14:10 rafi ndevos: very important,
14:10 ndevos rafi: hah! that is what everyone says :)
14:10 rafi ndevos: if there is two or more nodes, and if there is tiered volume , then there will be a cksum, which result in peer reject state
14:11 rafi ndevos: that is right :)
14:13 rafi ndevos: it is actually a minor change, and it is also passed the regression last time :)
14:13 ndevos rafi: hmm, could you get someone to review that change, and I'll make my mind up on including it later
14:14 rafi ndevos: great , <3
14:15 rafi ndevos: I have asked to Dan to review it :)
14:15 ndevos rafi: okay, lets see - regression should pass in about 2 hours from now
14:17 ndevos rafi: I'm stepping out for a while, and will check back on it later
14:18 rafi ndevos: ok, thanks a lot for your kind
14:52 Gaurav_ ndevos: 10722 is ready for merge
14:53 raghu ndevos: yes. It has passed regression tests and has got +2
14:56 rafi joined #gluster-dev
15:14 ndevos Gaurav_: yes, thanks :)
15:16 badone_ joined #gluster-dev
15:24 wushudoin joined #gluster-dev
16:04 dlambrig_ joined #gluster-dev
16:45 kripper joined #gluster-dev
16:58 kripper left #gluster-dev
17:03 dlambrig_ left #gluster-dev
17:08 rafi joined #gluster-dev
17:40 rafi joined #gluster-dev
17:50 dlambrig_ joined #gluster-dev
18:19 gem joined #gluster-dev
18:25 dlambrig_ joined #gluster-dev
19:16 dlambrig_ joined #gluster-dev
19:47 hagarth joined #gluster-dev
19:47 dlambrig_ joined #gluster-dev
20:49 rafi joined #gluster-dev
21:16 wushudoin joined #gluster-dev
21:17 wushudoin joined #gluster-dev
21:22 badone_ joined #gluster-dev
21:38 vimal joined #gluster-dev
21:40 vimal joined #gluster-dev
22:03 vimal joined #gluster-dev
22:08 vimal joined #gluster-dev
22:24 soumya joined #gluster-dev
22:40 nishanth joined #gluster-dev

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