Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2015-06-12

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

All times shown according to UTC.

Time Nick Message
02:36 kdhananjay joined #gluster-dev
02:57 kanagaraj joined #gluster-dev
03:08 overclk joined #gluster-dev
03:29 badone_ joined #gluster-dev
03:37 itisravi joined #gluster-dev
03:44 atinm joined #gluster-dev
03:49 atinm joined #gluster-dev
03:49 soumya joined #gluster-dev
03:52 nishanth joined #gluster-dev
03:58 shubhendu__ joined #gluster-dev
04:05 Yuan_ joined #gluster-dev
04:06 itisravi overclk: I'm not able to git pull either :-(
04:07 itisravi fatal: internal server error
04:07 itisravi remote: internal server error
04:07 itisravi fatal: protocol error: bad pack header
04:07 overclk itisravi, yep. An error a day keep developers away :)
04:07 itisravi overclk:  :D
04:08 itisravi overclk: You are welcome to step up as a dev-op and fix it ;)
04:09 overclk itisravi, using ff just sovled half of the problem.
04:09 itisravi hmm.
04:10 overclk itisravi, I generally cherry-pick patches to test them (apart from gerrit reviews). Now, that's what's painful :(
04:11 itisravi overclk: I actually cherry-pick even for reviews as it is easier to navigate using cscope.
04:14 overclk itisravi, we should really have a backup mechanism for reviewing patches (email, github, gerrithub..)
04:22 ashishpandey joined #gluster-dev
04:29 poornimag joined #gluster-dev
04:29 spalai joined #gluster-dev
04:30 spalai left #gluster-dev
04:33 nbalacha joined #gluster-dev
04:33 sakshi joined #gluster-dev
04:35 atalur joined #gluster-dev
04:45 ppai joined #gluster-dev
04:49 pppp joined #gluster-dev
04:53 kdhananjay joined #gluster-dev
04:54 gem joined #gluster-dev
04:55 itisravi overclk: It might be cool to add git send-email as an optional when running rfc.sh
04:55 tdasilva joined #gluster-dev
05:00 overclk itisravi, yeh. that way, patches can be cherry-picked directly from emails (+1 if using mutt ;))
05:00 itisravi overclk: spoken like a true kernel dev :)
05:03 spandit joined #gluster-dev
05:17 hgowtham joined #gluster-dev
05:20 Manikandan joined #gluster-dev
05:20 Manikandan_ joined #gluster-dev
05:22 vimal joined #gluster-dev
05:23 aravindavk joined #gluster-dev
05:26 hagarth joined #gluster-dev
05:28 ndarshan joined #gluster-dev
05:34 kshlm joined #gluster-dev
05:36 josferna joined #gluster-dev
05:40 Manikandan_ joined #gluster-dev
05:46 josferna joined #gluster-dev
05:46 hchiramm_ kshlm, any idea why its corrupted
05:47 anekkunt joined #gluster-dev
05:49 kshlm hchiramm_, no idea yet.
05:54 hagarth kshlm: running a git gc will recover usually
05:54 Gaurav__ joined #gluster-dev
05:54 hagarth kshlm: can you please point me to the specific error?
05:55 hagarth is this in gerrit's error logs?
05:56 kshlm hagarth, yes
05:56 shubhendu__ joined #gluster-dev
05:56 jiffin joined #gluster-dev
05:58 hagarth kshlm: running a git gc now
05:58 kshlm hagarth, the project-configuration page on the website had a GC button. Did you run it from there?
05:58 hagarth kshlm: no
05:59 hagarth kshlm: I usually run it from the background..
05:59 hagarth s/background/backend/
05:59 schandra joined #gluster-dev
05:59 kshlm Does this happen much?
05:59 hagarth kshlm: happens once every few months
06:00 kshlm Okay then. You know how to fix it. :)
06:01 hagarth kshlm: done
06:01 hagarth kshlm: please go ahead and check now
06:01 poornimag joined #gluster-dev
06:01 kshlm Yup. Can pull now.
06:02 kshlm hchiramm_, can you re-enable submit?
06:02 hchiramm_ kshlm, yep.. doing it
06:02 hchiramm_ hagarth++ kshlm++ thanks!
06:02 glusterbot hchiramm_: hagarth's karma is now 66
06:02 glusterbot hchiramm_: kshlm's karma is now 15
06:02 hchiramm_ I have reverted the permission
06:03 kshlm hchiramm_, can you also update the mailing-lists?
06:03 hchiramm_ sure..
06:19 pranithk joined #gluster-dev
06:26 hchiramm_ poornimag, ping, pm
06:30 spalai joined #gluster-dev
06:31 kaushal_ joined #gluster-dev
06:43 poornimag joined #gluster-dev
06:46 hchiramm joined #gluster-dev
06:50 raghu joined #gluster-dev
06:53 spalai joined #gluster-dev
06:53 deepakcs joined #gluster-dev
07:08 obnox hi guys
07:08 hchiramm Hi obnox !!!
07:08 hchiramm how  r u doing ?
07:08 obnox if I have patches that modify under doc/ in the source tree, do I need to create separate bugs for those or is there a catchall doc/ bug?
07:08 obnox hchiramm: great thanks,
07:08 obnox how are things?
07:09 hchiramm going fine.. thanks
07:09 hchiramm obnox, we are planning to keep away from gluster source doc directory
07:09 obnox hchiramm: hmm, how should I understand that?
07:09 hchiramm u may submit the Pull requests against https://github.com/gluster/glusterdocs
07:09 obnox i don't mean the docs repository
07:10 obnox i mean doc/ in the glusterfs.git
07:10 hchiramm yes, I get it ..
07:10 obnox e.g. some fixes in doc/developer-guide/translator-development.md
07:10 obnox oh, even there? ok
07:10 obnox so is the in-source-tree also going to go away?
07:10 hchiramm what I meant was , "/doc" in gluster.git is going to be deprecated
07:11 hchiramm and we will be maintaining docs in https://github.com/gluster/glusterdocs
07:11 hchiramm which is the base for gluster.readthedocs.org
07:11 hchiramm regarding the umbrella bug for doc fixes
07:11 ashiq joined #gluster-dev
07:11 hchiramm u can use https://bugzilla.redhat.co​m/show_bug.cgi?id=1206539
07:11 glusterbot Bug 1206539: high, high, ---, schandra, NEW , Tracker bug for GlusterFS documentation Improvement.
07:11 hchiramm obnox, is that clear ?
07:12 obnox hchiramm: thanks, yes
07:12 hchiramm so a PR against https://github.com/gluster/glusterdocs/blob/ma​ster/Developer-guide/translator-development.md
07:13 obnox okay, so not doing the gerrit-review process for the docs? simple pull request?
07:14 hchiramm obnox, u can add the required Cc in PR comment
07:14 obnox ok sure
07:15 hchiramm hope its clear.
07:15 hchiramm we are getting into new workflow..
07:15 hchiramm lets see how it goes.. please feel free to share your comments.
07:16 hchiramm u can respond to this thread as well. http://www.gluster.org/pipermail/g​luster-users/2015-May/022065.html
07:16 obnox yeah, clear. I assume that the doc/ will go away or at least shrink in the source tree when this has been running successfully for a time?
07:17 hchiramm http://review.gluster.org/11129 obnox
07:17 saurabh_ joined #gluster-dev
07:17 rafi joined #gluster-dev
07:17 hchiramm yeah..  thats the plan
07:19 obnox hchiramm: I see. :)
07:19 hchiramm \o/
07:49 anrao joined #gluster-dev
08:29 nbalacha joined #gluster-dev
08:36 rjoseph joined #gluster-dev
08:36 Joe_f joined #gluster-dev
08:39 Gaurav__ joined #gluster-dev
08:48 spot joined #gluster-dev
08:52 soumya joined #gluster-dev
09:04 hagarth joined #gluster-dev
09:08 nkhare joined #gluster-dev
09:19 hchiramm atinm, kshlm http://review.gluster.org/#/c/9877/
09:19 hchiramm can u please review/merge this
09:20 atinm hchiramm, I see spandit has already acked it
09:20 atinm kshlm, can u ?
09:20 hchiramm yes.. merging is pending
09:21 atinm hchiramm, netbsd vote?
09:22 hchiramm ah.. lets wait some more on it
09:26 hchiramm http://review.gluster.org/#/c/9864/ atinm can u review this
09:26 hchiramm ?
09:26 hchiramm anoop has already reviewed however need some one from glusterd  to check it
09:27 * atinm is on it
09:27 hchiramm spandit, is it something u can review  http://review.gluster.org/#/c/10473/
09:27 hchiramm atinm++ thanks!
09:27 glusterbot hchiramm: atinm's karma is now 2
09:29 atinm hchiramm, I see blank domain name in this patch
09:29 hchiramm gem, ^^^
09:30 atinm I thought gem has taken care of this in all her patches :)
09:32 spandit hchiramm, let me check
09:32 hchiramm thanks spandit++
09:32 glusterbot hchiramm: spandit's karma is now 3
09:33 spandit hchiramm, I will review that
09:33 hchiramm thanks!!!
09:35 gem joined #gluster-dev
09:36 gem atinm, Sorry, I missed that one :)
09:36 gem atinm, will send ASAP
09:36 atinm gem, np, as long as its getting fixed :)
09:54 xavih joined #gluster-dev
10:05 Manikandan_ hchiramm, ping
10:05 glusterbot Manikandan_: Please don't naked ping. http://blogs.gnome.org/mark​mc/2014/02/20/naked-pings/
10:05 hchiramm Manikandan_, pong
10:06 Manikandan_ http://review.gluster.org/#/c/9952/ patch is reviewed by kotresh..
10:06 * ndevos yells "aah!" - connect to host git.gluster.org port 22: Connection timed out
10:07 hchiramm overclk, can u merge this patch http://review.gluster.org/#/c/9952/
10:09 ndevos hagarth: I hope you are planning to update the MAINTAINERS file ;-)
10:10 ndevos hagarth: when you do, include maintainers@gluster.org in there too
10:11 overclk hchiramm, sure
10:11 hchiramm thanks overclk++
10:11 glusterbot hchiramm: overclk's karma is now 5
10:12 hchiramm ndevos, can u please merge this patch http://review.gluster.org/#/c/10670
10:13 ndevos hchiramm: no, its merged already?
10:13 hchiramm ndevos, ahhh..my bad :(
10:13 hchiramm ndevos++ hagarth++
10:13 glusterbot hchiramm: ndevos's karma is now 154
10:13 glusterbot hchiramm: hagarth's karma is now 67
10:14 * ndevos twists Humbles ear
10:14 * hchiramm hahahahahaha
10:14 ndevos not, aiaiaiaiai?
10:15 hchiramm s/h/i/ >>1  :)
10:16 Manikandan_ joined #gluster-dev
10:18 hchiramm ndevos, http://etherpad.corp.redhat.com/logging-patches
10:18 Manikandan_ overclk, ping
10:18 glusterbot Manikandan_: Please don't naked ping. http://blogs.gnome.org/mark​mc/2014/02/20/naked-pings/
10:18 hchiramm we need reviews for libglusterfs logging patches
10:19 ndevos hchiramm: interestingly, that is not a community list of patches?
10:19 hchiramm yes, it is
10:19 hchiramm the patches listed by Mohamed Ashiq are libglusterfs patches
10:19 ndevos no it is not, corp.redhat.com is not community facing
10:22 hchiramm ah.. ndevos oh.. u are talking about the etherpad
10:22 hchiramm if that is the case, I created it for tracking purpose
10:22 hchiramm yeah and its not community one
10:22 hchiramm ndevos, the libglusterfs patches , most of them need a review
10:23 ndevos yeah, how can an open source community use non-public accessible tracking methods?
10:25 hchiramm https://public.pad.fsfe.org/p/logging-patches ndevos use this url :)
10:26 obnox hchiramm: pull request for glusterdocs repo done
10:26 hchiramm obnox, reviewing
10:26 ndevos hchiramm: thanks, and please mark the other etherpad deprecated if you did not do so yet
10:26 csim ndevos++
10:26 glusterbot csim: ndevos's karma is now 155
10:27 ndevos also, I'm trying to keep up with doing reviews, but there are rather many: review.gluster.org/#/q/status:open+reviewer:ndevos
10:27 csim we need to keep openess in mind, even if that's quite easy to forget
10:27 rjoseph joined #gluster-dev
10:27 ndevos yeah, default to open!
10:27 csim ( given how many time I discuss with JustinClift on internal irc because I fogot the location of this windows :/ )
10:27 hchiramm obnox, merged :)
10:27 obnox hchiramm: cool, that was quick ;)
10:28 obnox it is good to have a process with less overhead for those docs
10:28 hchiramm indeed .. thats the aim
10:28 hchiramm obnox,
10:28 ndevos I wonder how we do review docs though?
10:28 hchiramm csim, ndevos it was an etherpad created for downstream tracking..
10:29 ndevos hchiramm: downstream tracking of upstream patches sounds wrong to me :)
10:30 hchiramm :P
10:31 csim hchiramm: yeah, that's quite tricky to split the 2 :/
10:31 csim and we do not have the luxury of having 2 differents teams
10:32 hchiramm indeed.. when downstream follow , upstream first strategy
10:32 hchiramm indeeeeeeeeeeeeeed
10:33 ndevos soumya|afk: aha! I was wondering why I did not spot the busy-loop before, seems that I did not review the patch at all - http://review.gluster.org/10342
10:35 overclk atinm, anekkunt: mind having a look at the glusterd/cli parts for http://review.gluster.org/#/c/11105/ ? (or the patch as a whole if you want).
10:40 overclk raghu, is this patch needed ? http://review.gluster.org/#/c/11127/
10:42 soumya ndevos, right..I pinged you earlier to ask if you would like submit new patch to fix that..
10:42 soumya or want to include in http://review.gluster.org/#/c/11196/ ?
10:43 ndevos soumya: I'm submitting a patch for that already, if its not posted yet...
10:43 soumya ndevos, okay...its not yet posted
10:43 ndevos soumya: ah, no,
10:43 ndevos connect to host git.gluster.org port 22: Connection timed out
10:44 * ndevos tries again
10:46 ndevos soumya: here it is: http://review.gluster.org/11198
10:50 soumya ndevos, thanks! reviewed :)
10:50 Gaurav__ joined #gluster-dev
10:50 ndevos soumya++ thank you
10:51 glusterbot ndevos: soumya's karma is now 11
10:51 soumya ndevos, its taking time to publish though ;)
10:51 ndevos soumya: no problem, regression tests will take longer (I hope)
10:55 jiffin ndevos++ for the ganesha patch review
10:55 glusterbot jiffin: ndevos's karma is now 156
11:00 hchiramm ndevos++
11:00 glusterbot hchiramm: ndevos's karma is now 157
11:05 ndevos hchiramm: I think many jenkins tests fail because of Gerrit connection timeout issues, are you looking into that?
11:06 hchiramm ndevos, I am not :(
11:06 ndevos hchiramm: is that something you or csim can check? connection timeouts when posting patches is not too bad, failing regression tests is worse
11:20 dlambrig joined #gluster-dev
11:22 soumya ndevos++ thanks for the fix and the testing
11:22 glusterbot soumya: ndevos's karma is now 158
11:23 ndevos soumya++ thanks for checking, improving and posting the patch :)
11:23 glusterbot ndevos: soumya's karma is now 12
11:23 * ndevos will have lunch now, will back later
11:24 ndevos have a good weekend all, in case you wont be here when I return :)
11:24 csim can we have a good weekend even if we are here when you return ?
11:25 ndevos yes, but I doubt your weekend starts before I return
11:26 csim I cannot redeem the "good weekend" later, there is a expirtaion date :) ?
11:26 soumya :D
11:27 soumya ndevos, in case if am not online, please remind frank to pull in today's patch also if possible
11:27 kanagaraj joined #gluster-dev
11:32 shubhendu__ joined #gluster-dev
11:35 rjoseph joined #gluster-dev
11:43 ashiq hchiramm : http://review.gluster.org/10822 & http://review.gluster.org/10825 been reviewed and regression passed
12:03 ira joined #gluster-dev
12:07 hchiramm ashiq, netbsd regression has not passed
12:09 anrao joined #gluster-dev
12:10 ashiq hchiramm, ok will re-trigger it :)
12:10 spalai left #gluster-dev
12:16 shubhendu__ joined #gluster-dev
12:19 pranithk xavih: hey! Did you get a chance to see http://review.gluster.com/11178
12:47 pranithk ndevos: which bug are you working on?
12:47 pranithk ndevos: Shall we work on that mem-accounting hang bug?
12:54 hagarth joined #gluster-dev
13:11 ndevos ... pranithk?
13:11 pranithk joined #gluster-dev
13:12 ndevos ah, there he is!
13:13 hagarth joined #gluster-dev
13:14 ndevos pranithk: I checked that system earlier, it is not in that "mem-accounting hang" state anymore...
13:14 shyam joined #gluster-dev
13:15 ndevos pranithk: Bhaskarakiran wasnt online, and I'm not sure if I can disable the other volumes so I can look into the nfs process a little better
13:16 pranithk ndevos: Ah!, He just showed me the same today in brick process as well :-)
13:17 pranithk ndevos: Do you remember the bug-id?
13:21 ndevos pranithk: 1230612
13:23 ndevos pranithk: does that only happen when a statedump is done?
13:24 ndevos pranithk: maybe http://review.gluster.org/11095 would be an improvement for it?
13:31 pranithk ndevos: Why is it hanging in normal syscalls? I thought initially it is hanging on our own GF_CALLOC/GF_FREE calls but it is hanging on CALLOC/FREE calls :-(
13:31 ndevos pranithk: it is hanging, or do they just happen to show up?
13:32 pranithk ndevos: I saw it after some time as well. It was still hung in same place :-/
13:32 ndevos pranithk: if you execute "gstack $PID" a couple of times after eachother, is it always the same calloc/free?
13:33 pranithk ndevos: I don't know about this till you told me. When it happens next time, I will check this. He re-creates it at will
13:33 ira joined #gluster-dev
13:34 hagarth pranithk: can we get a gcore once the hang is encountered?
13:34 ndevos pranithk: gstack is a trick from one of the IBM/Ganesha devs, seems easier to use than gdb for quick checking
13:35 * ndevos isnt convinced it is a real "hang", maybe a loop that just never ends
13:35 ndevos ... okay, maybe not everyone cares about the subtile difference
13:36 pranithk ndevos: http://cboard.cprogramming.com/c-pr​ogramming/33575-malloc-hanging.html
13:38 hagarth pranithk: normally malloc/calloc hanging points to a memory corruption too
13:38 pranithk hagarth: That is what I was pointing out in the link above :-)
13:38 pranithk ndevos: ^^
13:38 hagarth pranithk: ok :)
13:41 pousley joined #gluster-dev
13:41 ndevos pranithk: right, so build the upstream rpms, and pass "--with debug" to rpmbuid/mock?
13:41 pranithk ndevos: nah! this is a job for asan :-)
13:42 ndevos pranithk: asan??
13:43 csim a gcc extension to check for threads and memory corruption
13:45 ndevos hmm, never heard of that before, but we have explicit memory invalidation and checking when --enable-debug is passed to ./configure
13:47 csim https://code.google.com/p/address​-sanitizer/wiki/AddressSanitizer
13:53 ndevos so basically you only need to add -fsanitize=address to the compile options
13:53 ndevos I'm sure I've seen that option before, maybe we already have that enabled?
13:54 ndevos well, not in the gluster sources...
14:16 Manikandan joined #gluster-dev
14:16 Manikandan Humble, ping
14:16 glusterbot Manikandan: Please don't naked ping. http://blogs.gnome.org/mark​mc/2014/02/20/naked-pings/
14:16 dlambrig1 joined #gluster-dev
14:16 dlambrig1 left #gluster-dev
14:16 dlambrig1 joined #gluster-dev
14:17 hchiramm Manikandan, pong
14:37 kkeithley ndevos: Address Sanitizer came from clang. gcc added it. No, we don't normally have it turned on.
14:41 ndevos kkeithley: maybe we should turn it on for debug builds?
14:42 csim I guess a first try is to turn it and run the test suite ?
14:43 kkeithley maybe. maybe I have it on in some of the static analysis builds? need to lookj
14:51 kkeithley Hm...  I started to set up nightly runs of master built with various -fsanitize= options, e.g. address, leak, thread, and undefined+null.  It's been too long, I don't remember where I left off.
14:51 soumya joined #gluster-dev
15:00 ndevos csim: dns on build.gluster.org is still flakey, Could not resolve hostname review.gluster.org: Name or service not known
15:00 ndevos a 2nd time worked though...
15:01 csim mhhh
15:17 ndevos csim: could I get a gerrithub@gluster.org address? I would like to use that for NFS-Ganesha tests related to Gluster
15:17 csim ndevos: yes
15:17 csim ndevos: so, what time did the review failed ? ( so I can take a look at log )
15:18 csim ( the review lookup )
15:18 csim Jun 12 08:00:16 dnsmasq[17012]: query[A] review.gluster.org.gluster.org from 127.0.0.1
15:19 csim mhhh
15:19 ndevos csim: I manually did the re-posting of +1 verified as jenkins user on build.gluster.org for http://review.gluster.org/#/c/11095/ - time of the last comment
15:19 csim around 8h, ie 20 minutes ago
15:19 csim it seems there as some weird stuff
15:19 ndevos yeah, 20 minutes ago
15:21 ndevos csim: or, maybe instead of a gerrithub@gluster.org address, a nfs-ganesha@gluster.org one? makes it easier if we want to have things split up
15:21 csim ndevos: you can have both :)
15:21 ndevos csim: one would do, and I dont really care which :)
15:22 csim ndevos: ok, i can create icantdecide@gluster.org :)
15:22 ndevos :P
15:23 csim go for gerrithub
15:23 ndevos http://i.imgur.com/TURedKU.jpg is my current view btw
15:23 csim I have a similar one, in a transat, seeing the train
15:24 ndevos oh, trains are on the left side, if you really want to see those, I can make a picture of that too
15:25 csim we may make others people jealous :)
15:27 csim ndevos: alias done
15:27 ndevos csim++ thanks!
15:27 glusterbot ndevos: csim's karma is now 3
15:29 hchiramm ndevos, nice view :)
15:33 csim damn, technology hate me today
15:34 ndevos proof with a train, but I guess you understand why I'm facing the other directtion - i.imgur.com/vBHLdSo.jpg
15:35 ndevos its not technology thats bothering me today, only my calendar, I seem to be unable to get an event added
15:37 hchiramm_ joined #gluster-dev
15:43 csim http://shop.ateliers-draft.com/themes/elation-li​quid-bxl/tweaks/img/cms/cms-terrasse-jardin.jpg the view
15:50 ndevos oh, still sunny thare? its getting cloudy here :-/
15:54 ndarshan joined #gluster-dev
15:57 csim well, the picture was taken 1 year ago
16:03 ndevos ah, the weather probably has changed since then
16:12 shubhendu__ joined #gluster-dev
16:18 ndevos soumya: I fixes a compile warning in #2 of http://review.gluster.org/11198 (upcall busy-loop), can you +1 that again?
16:19 soumya ndevos, okay
16:20 ndevos dlambrig1, dlambrig: tier.t failed for a completely unrelated change, do you know about issues with it? http://build.gluster.org/job/rackspace-re​gression-2GB-triggered/10620/consoleFull
16:33 dlambrig1 ndevos: I'll take a look.
16:35 ndevos dlambrig1: thanks
16:37 dlambrig1 ndevos: suspect recent tier changes may have introduced a spurious error, thanks for letting me know.
16:39 ndevos dlambrig1: okay, as long as you're aware and have an idea how to fix it :)
16:49 dlambrig1 ndevos: I do not see this problem on my own machines. Are the build machines slower or some such so a problem is more likely to happen there?
16:49 ndevos dlambrig1: I guess they can be slower, they are VMs in "the cloud"
16:50 dlambrig1 ndevos: ok, I'll spin up a few machines and see if it recreates there
16:51 ndevos dlambrig1: yes, and if a TEST call failed, maybe you should wrap it in a EXPECT_WITHIN call or something
16:52 dlambrig1 ndevos: we do use EXPECT_WITHIN , but the timeout may need tweaking (what I suspect)
16:53 ndevos dlambrig1: oh, yes, or that :)
16:53 pranithk joined #gluster-dev
16:53 ndevos oh no, HE'S BACK
16:54 pranithk ndevos: sorry! :-(
16:54 pranithk ndevos: Had to get home real quick at that time. Just had dinner and now back
16:54 pranithk ndevos: Found something about the corruption?
16:55 ndevos pranithk: no, but I figured out who asan is
16:55 pranithk ndevos: Yeah I saw csim's response
16:55 ndevos pranithk: we could try to build with -fsanitize=... and/or ./configure --enable-debug
17:03 pranithk1 joined #gluster-dev
17:03 pranithk1 ndevos: Sorry! again
17:03 pranithk1 ndevos: Laptop froze. You know my sysadmin skills :-D. Rebooted it :-)
17:05 ndevos pranithk1: you're running tiering tests on your laptop now too?
17:05 dlambrig1 ndevos: hey I heard that!
17:06 pranithk1 ndevos: that was bitrot! Why do you always get it wrong and make him angry?
17:06 ndevos dlambrig1: :P
17:06 dlambrig1 ndevos: lol
17:06 pranithk1 dlambrig1: You should have hit him in Barcelona :-P
17:07 dlambrig1 pranithk1: hmmm exile him to the cold tier ..
17:07 pranithk1 dlambrig1: Hehe
17:08 ndevos dlambrig1: oh, right, now you can huh? got that merge permission bit?
17:08 aravindavk joined #gluster-dev
17:08 ndevos pranithk1: you want to run that ec test with a debug build?
17:09 dlambrig1 ndevos: yes you wont see my bribe check in the mail any longer
17:09 * ndevos hooray!
17:10 pranithk1 ndevos: I don't know steps to re-create
17:10 pranithk1 ndevos: I am trying different things
17:10 shaunm_ joined #gluster-dev
17:10 ndevos pranithk1: the steps to reproduce should be simple, and they should be in bug 1230612
17:10 glusterbot Bug https://bugzilla.redhat.com:​443/show_bug.cgi?id=1230612 unspecified, unspecified, ---, ndevos, NEW , Disperse volume : NFS mount hung with plain IO
17:11 ndevos hmm, that actually seems to have been assigned to me?
17:13 pranithk1 ndevos: Don't worry, once in a while you get the opportunity to learn new stuff...
17:13 ndevos pranithk1: I learn new stuff almost every day... would that be a reason to worry?
17:14 * ndevos might have a memory issue
17:14 pranithk1 ndevos: I thought you are worried that this bug is assigned to you
17:21 shyam pranithk1: can you PM me the core location, want to take a peek at it
17:24 ndevos pranithk1: if I would be worried about bugs getting assigned to me, I would surely have the wrong job
17:25 ndevos I think its the other way around, the bugs should be worried when they get assigned to me
17:28 pranithk1 shyam: No core. calloc and free hung!
17:28 pranithk1 ndevos: You spoke like one of telugu movie hero!
17:28 shyam pranithk1: the stack presented in the bug was by attaching gdb and taking a backtrace? No core saved?
17:29 Gaurav__ joined #gluster-dev
17:29 pranithk1 shyam: gah! no :-(.
17:29 shyam pranithk1: :( k
17:30 pppp joined #gluster-dev
17:35 pranithk1 ndevos: I was going through memcpy/strcpy calls in the last few commits...
17:35 * ndevos DuckDuckGo'd "telegu movie" and landed at Tollywood
17:35 ndevos pranithk1: good luck with that!
17:36 pranithk1 ndevos: http://review.gluster.org/10224 has a bug
17:36 pranithk1 ndevos: Not this one though
17:36 pranithk1 ndevos: it is upcall
17:36 pranithk1 ndevos: -        memcpy (gf_c_req->gfid, gf_up_data->gfid, 16);
17:36 pranithk1 +        gf_c_req->gfid = uuid_utoa (gf_up_data->gfid);
17:36 pranithk1 ndevos: That is a landmine
17:37 pranithk1 ndevos: uuid_utoa should never ever be used outside gf_log
17:38 ndevos pranithk1: oh, no?
17:38 * ndevos checks
17:39 ndevos oh, ew - pthread_getspecific (uuid_buf_key)
17:40 ndevos ah, but we have uuid_utoa_r() !
17:40 ndevos and gf_uuid_unparse() actually too... hmm
17:41 pranithk1 ndevos: I guess you know the fix :-)
17:42 ndevos pranithk1: just wondering how big the string would be :)
17:44 pranithk1 ndevos: 32+4
17:44 pranithk1 ndevos: 4 hiphens
17:45 ndevos pranithk1: 'man uuid_unparse' says 36 + '\0' :)
17:45 ndevos but, I was hoping we have a #define for it somewhere
17:45 xavih joined #gluster-dev
17:49 ndevos maybe I just do strdup(uuid_utoa(..)) ;-)
17:51 pranithk1 ndevos: I saw gf_uuid_parse used in same patch, check that patch once. My laptop is slow again :-(
17:51 pranithk1 xavih: ola
17:52 ndevos pranithk1: uuid_utoa is not used in upcall anymore...
17:52 pranithk1 ndevos: good!!
17:54 ndevos pranithk1: well, I guess that depends, I also do not find uuid_utoa in release-3.7...
17:55 pranithk1 ndevos: :-/ as long as it is not used in synctask it should be fine.
18:09 ndevos pranithk1: I do not really see incorrect usage of uuid_utoa that quickly
18:09 ndevos pranithk1: running with a debug build might be the easiest?
18:10 ndevos pranithk1: but I'll be leaving for the day, not sure if I have time to look into this during the weekend
18:10 ndevos you know, because of WEEKEND that you also have ;-)
18:44 craigcabrey joined #gluster-dev
19:34 kkeithley Ubuntu ppas have been updated; now with 3.7.1, 3.6.3, and 3.5.4 for most current ubuntu releases.
20:07 bfoster joined #gluster-dev
20:08 purpleidea joined #gluster-dev
20:08 purpleidea joined #gluster-dev
20:08 ndk joined #gluster-dev
21:43 dlambrig joined #gluster-dev
22:27 badone_ joined #gluster-dev

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