Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2015-05-20

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

All times shown according to UTC.

Time Nick Message
00:07 ppai joined #gluster-dev
00:08 ws2k3 joined #gluster-dev
00:20 wushudoin joined #gluster-dev
00:44 itisravi joined #gluster-dev
00:56 badone_ joined #gluster-dev
02:12 kdhananjay joined #gluster-dev
02:18 itisravi joined #gluster-dev
02:45 itisravi joined #gluster-dev
02:55 Guest72040 joined #gluster-dev
03:34 kanagaraj joined #gluster-dev
03:47 overclk joined #gluster-dev
03:48 itisravi joined #gluster-dev
03:55 ashishpandey joined #gluster-dev
03:57 suliba joined #gluster-dev
03:57 shubhendu joined #gluster-dev
04:08 sakshi joined #gluster-dev
04:11 nbalacha joined #gluster-dev
04:20 spandit joined #gluster-dev
04:32 pranithk joined #gluster-dev
04:37 pppp joined #gluster-dev
04:42 nishanth joined #gluster-dev
04:48 ashiq joined #gluster-dev
04:48 sakshi joined #gluster-dev
04:49 ashishpandey joined #gluster-dev
04:52 sakshi joined #gluster-dev
04:53 rafi joined #gluster-dev
04:54 dlambrig joined #gluster-dev
05:16 ndarshan joined #gluster-dev
05:24 atinmu joined #gluster-dev
05:24 rafi1 joined #gluster-dev
05:26 rjoseph joined #gluster-dev
05:27 Joe_f joined #gluster-dev
05:27 Manikandan joined #gluster-dev
05:28 anrao joined #gluster-dev
05:29 rafi joined #gluster-dev
05:29 hgowtham joined #gluster-dev
05:31 krishnan_p joined #gluster-dev
05:31 rafi joined #gluster-dev
05:35 kdhananjay joined #gluster-dev
05:35 poornimag joined #gluster-dev
05:38 raghu joined #gluster-dev
05:39 kdhananjay1 joined #gluster-dev
05:43 hagarth joined #gluster-dev
05:46 gem joined #gluster-dev
06:00 Gaurav_ joined #gluster-dev
06:09 itisravi hagarth: hello. Could you check if arbiter.t still fails on your laptop with http://review.gluster.org/#/c/10833/
06:10 hagarth itisravi: I think it will .. let me check and get back
06:10 itisravi hagarth: thanks!
06:10 overclk joined #gluster-dev
06:23 anekkunt joined #gluster-dev
06:26 rafi1 joined #gluster-dev
06:26 sakshi joined #gluster-dev
06:47 aravindavk joined #gluster-dev
06:48 atinm joined #gluster-dev
06:50 ira joined #gluster-dev
06:57 rafi joined #gluster-dev
06:59 tigert morning
07:00 kshlm joined #gluster-dev
07:03 krishnan_p joined #gluster-dev
07:10 ndarshan left #gluster-dev
07:11 ndarshan joined #gluster-dev
07:23 poornimag joined #gluster-dev
07:25 pranithk joined #gluster-dev
07:35 jiffin joined #gluster-dev
07:43 ashishpandey joined #gluster-dev
07:45 schandra joined #gluster-dev
07:50 atinm joined #gluster-dev
07:57 tigert hchiramm__: incoming change request for the landing page of old docs redirect
08:05 pranithk joined #gluster-dev
08:06 tigert guys, if we want to ask people to contribute to the documentation project, who would be the person to point them to?
08:06 tigert and would there be a wikipage or some information I could put on the community page?
08:07 tigert is gluster-devel a good list to discuss docs?
08:08 ndevos tigert: depends on what the target audience is and the topic of the discussions, maybe gluster-users fit better?
08:09 tigert "how to get involved in contributing to docs" I guess
08:09 tigert it would be good to have a docs team I guess
08:09 tigert who takes the responsibility of reviewing the docs for each release etc
08:10 tigert and maintaining the wiki for any leftover stale old docs etc
08:12 ndevos hchiramm and lalatenduM were marked as "doc maintainers"
08:16 tigert yeah, I guess if someone pokes their head to the users or devel list and shares docs ideas, they would notice
08:16 tigert better have info that doesnt become incorrect if people switch roles
08:17 tigert hchiramm__: another pull request incoming btw
08:17 tigert https://github.com/humblec/glusterdocs/pull/3
08:18 lalatenduM tigert: ndevos , doc maintainers as of now just maintain the docs directory , not wiki
08:18 tigert yeah
08:18 tigert we need to discuss the role of the wiki
08:18 tigert wiki needs maintainership too
08:18 ndevos lalatenduM: oh, okay, for me the wiki belongs to doc too ;-)
08:18 tigert it has a lot of the working drafts of those docs now
08:18 tigert dangerous for people to again find old crap via google
08:18 tigert :P
08:19 tigert I guess the wiki is useful unless you use etherpads etc for drafting ideas
08:20 ndevos well, there are quite some wiki pages that describe how things work, and we extend them as we hit questions or need to clarify things
08:21 ndevos "things" are processes in this case, like bug triage, bug lifecycle, maintainer guidelines, ...
08:22 tigert yeah
08:23 tigert I guess the wiki should just be cleaned up and kept in order
08:23 hchiramm__ tigert, I was afk :) .. regarding contribution to doc
08:24 hchiramm__ yeah, there will be a communication soon to the gluster MLs
08:24 hchiramm__ I am sorting the long term workflow which have to be followed
08:24 hchiramm__ once its ready we will reach to MLs
08:29 tigert ok
08:29 tigert for now I have those 2 pull requests
08:29 tigert if you can merge at least the former, I can do the redirect for old docs
08:30 tigert I don't think it has to be visible in the index even
08:30 ndevos tigert: for process/workflow kind of things, do you suggest to place them in the wiki, or should we have them on a 'real' page?
08:30 tigert it can be just a hidden page we point to (I assume it will get generated?)
08:30 tigert ndevos: we need to discuss on how we want to do them
08:31 tigert ndevos: in a way the website should likely be very small static shop window that explains how awesome gluster is
08:31 tigert and just points to different places
08:31 tigert for example the http://www.gluster.org/community/ could well be a page in the wiki
08:32 tigert but we should then make sure the wiki is kept in order
08:32 tigert and maybe structured so that a newcomer can find stuff easily
08:33 hchiramm__ tigert, we should have a note in wiki saying, it is getting deprecated..
08:33 tigert the whole wiki or just docs in the wiki?
08:33 tigert this is something we should discuss
08:34 hchiramm__ afict, whole wiki..
08:34 tigert a wiki might be useful but someone really needs to have a responsibility to maintain the contents
08:34 atalur joined #gluster-dev
08:34 tigert if it is not, then it gets full of bitrot
08:34 hchiramm__ any one can have a login in wiki and write content
08:35 hchiramm__ thats something we should restrict.
08:35 tigert you can also restrict accounts
08:35 hchiramm__ more or less, there is no point of having the documentation scattered among different places
08:35 tigert but definitely yeah
08:35 tigert it might be a place to contribute a draft of docs too
08:35 tigert but otoh
08:35 tigert you can do the same in a github repo
08:35 tigert thanks to their online editing tools
08:36 tigert but I guess its worth discussing on the list that what kind of things the wiki is used for
08:36 hchiramm__ thats why we started to migrate contents from media wiki to githu
08:36 hchiramm__ github
08:36 tigert like ndevos said, about the processes etc
08:36 hchiramm__ yeah, as I said, we will reach out to MLs soon about the plan
08:36 tigert yeah, great
08:37 hchiramm__ the current plan is this
08:37 hchiramm__ 1) developer guide will be part of gluster source
08:37 hchiramm__ 2) rest of the docs will be maintained by github project
08:37 tigert (I need to run to lunch now, but I will read the scrollback and respond when I get back)
08:37 hchiramm__ sure.
08:38 ndevos hchiramm__: things like the bug lifecycle should go to the github docs?
08:38 hchiramm__ yes. its already part of developer guide I believe
08:39 hchiramm__ http://gluster.readthedocs.org/en/late​st/Developer-guide/Developers%20Index/
08:43 ndevos hchiramm__: oh, good, but please replace the wiki contents with a refence to the right doc page
08:44 ndevos hchiramm__: otherwise I'll update the wiki page, and the contents gets out of sync again :-/
08:44 hchiramm__ ndevos, the idea is to have a note in wiki saying its deprecated.
08:44 hchiramm__ as a banner
08:44 hchiramm__ so that we can avoid editing the content..
08:44 hchiramm__ yes, there will be a note which says where is our documentaiotn
08:45 ndevos hchiramm__: oh, will the whole wiki be deprecated? in that case it should be made read-only *now*?
08:45 hchiramm__ yeah.. thats the plan
08:45 ndevos well, but at the moment we have two editable versions, thats bad
08:46 anrao joined #gluster-dev
08:46 * ndevos still needs to update the nightly builds page, but will look for doing that later
08:46 hchiramm__ ndevos, nightly build page is already part of the developer guide in github?
08:46 ashishpandey joined #gluster-dev
08:46 ndevos hchiramm__: also, the planning ad feature pages get modified on occasion
08:47 ndevos hchiramm__: I dont know? should it be? or should I keep it in the wiki?
08:47 hchiramm__ schandra, can help us to port this to github
08:47 hgowtham joined #gluster-dev
08:48 ndevos I do not have an opinion on where to keep the docs, but I want only *one* editable copy of them
08:48 schandra hchiramm__, sure
08:49 hagarth pranithk: ping, around?
08:49 pranithk hagarth: yes?
08:50 hagarth pranithk: pm
08:51 shubhendu joined #gluster-dev
08:52 ndarshan joined #gluster-dev
09:09 krishnan_p joined #gluster-dev
09:11 tigert back
09:12 ndarshan joined #gluster-dev
09:13 tigert hchiramm__: any clue as to how to make the navigation on docs collapse?
09:13 tigert hchiramm__: http://docs.phinx.org/en/latest/ < like this on the left
09:14 hchiramm__ tigert, I dont have..
09:14 hchiramm__ may be schandra can help us here :)
09:14 hchiramm__ tigert, yes, it looks cleaner
09:14 tigert yep
09:14 tigert there has to be an option since it works there
09:14 tigert but that looks like some different setup
09:15 tigert because there is no yml
09:15 schandra that is a sphinx-rst setup tigert
09:15 tigert ok
09:16 schandra so afaik mkdocs can reflect similar behaviour PROVIDED we have a custom theme
09:17 nbalacha joined #gluster-dev
09:18 spalai joined #gluster-dev
09:18 schandra I tried drop down configuration in yml but it will not affect unless we have a custom theme
09:18 tigert oh ok
09:21 shubhendu joined #gluster-dev
09:21 tigert that mkdocs author dougal works for red hat btw
09:21 schandra oh is it!!
09:21 hchiramm__ tigert, ohhhhhhhh
09:22 tigert though, I am not sure if mkdocs and readthedocs are the same
09:22 tigert but mkdocs is one way to do this
09:23 tigert well actually yes
09:23 tigert readthedocs can use mkdocs stuff
09:23 tigert readthedocs is just the service to host the stuff
09:23 schandra mkdocs and rtd are not same
09:23 tigert rtd can use mkdocs and sphinx format documentation
09:23 schandra mkdocs and sphinx help to convert .md to .html
09:23 tigert yea
09:24 schandra formats are restructuredtext and markdown
09:24 schandra rst uses sphinx and md uses mkdocs
09:24 tigert d0ugal is the author of mkdocs
09:24 tigert he is asleep still though it seems
09:25 tigert idle     : 0 days 9 hours 30 mins 28 secs
09:26 schandra yea
09:35 tigert hm
09:36 tigert isnt the repo structured a bit strangely, looks like mkdocs expects the content to be under docs/ ?
09:38 tigert hm, maybe not
09:38 * tigert RTFM's
09:38 schandra tigert, if there is docs or doc directory then it will be starting point
09:38 tigert so how do you build this?
09:40 schandra a repo url will be configured
09:42 * tigert wonders why there is no "docs" folder in the git repo
09:42 tigert do you move things by hand?
09:42 schandra because repo is docs only repo
09:43 schandra making repo itself the docs folder
09:43 tigert aah
09:43 tigert so the documentation content will be a git submodule or something?
09:43 tigert and the tooling is separate?
09:43 schandra tigert, I did not understand "moving things by hand?"
09:44 schandra I think hchiramm__ has a more convincing answer
09:44 tigert I mean, I want to use mkdocs to test that my changes will work, before making a pull request
09:44 tigert and I am trying to run mkdocs
09:44 tigert so I am wondering how should I do this
09:44 schandra okay! now I get
09:44 hchiramm__ tigert, A simple option would be to configure ur own project in readthedocs
09:44 hchiramm__ point your repo
09:44 tigert yeah, I have that too
09:45 schandra so if you want to test using mkdocs
09:45 tigert but I wish to try things without making comments
09:45 tigert commits
09:45 tigert I have mkdocs installed locally
09:45 schandra u can use mkdocs serve
09:45 schandra tigert,
09:45 nishanth joined #gluster-dev
09:46 tigert yeah, but it expects a docs/ folder
09:46 schandra is it
09:47 tigert ERROR - file not found: docs/index.md
09:47 schandra yes tigert , it expects, my bad
09:47 schandra I have mkdocs installed locally too
09:48 * tigert thinks the docs git repo should be something we can use directly to build
09:48 tigert and just have a .gitignore for any built files
09:48 schandra I have folder "my project" which has docs/ and theme/ folders along with mkdocs.yml
09:49 tigert would it make sense to have all this in the docs git repo?
09:49 schandra serving with mkdocs and building with readthedocs is different
09:50 tigert so that you can git clone, cd glusterdocs; mkdocs serve ?
09:50 tigert oh ok
09:50 tigert so readthedocs needs a different structure?
09:50 schandra tigert, git repo in the sense gluster/glusterfs?
09:50 * tigert thinks he is not understanding something and just wastes time, sorry
09:51 ndevos tigert: time is not wasted if you learn something from it ;-)
09:52 tigert yeah :)
09:53 tigert so I could just mkdir docs
09:53 tigert and move all the markdown files there
09:54 schandra yes.
09:54 schandra (just to fulfill the requirement of testing using mkdocs without actually commiting)
09:55 tigert yeah
09:55 tigert is there a reason we don't want to have the correct structure in the glusterdocs git repo?
09:55 tigert for clarity?
10:02 schandra hmmm let me put it like this>> http://fpaste.org/223711/
10:03 tigert yeah, I got it to work
10:03 tigert thanks
10:03 tigert good points
10:03 tigert it makes sense
10:04 tigert I added one more pull request
10:04 schandra so omitting a docs/ folder we reduce the hierarchy to 2 level.
10:04 tigert true, makes sense now
10:07 tigert hchiramm__: around? can you poke at the pull req?
10:09 hchiramm__ tigert, /me checking
10:13 tigert simply adding a line seems to be required to add the doc into the build
10:14 tigert without link title it seems to be just unlisted
10:14 tigert which is fine with this
10:23 sks joined #gluster-dev
10:23 csaba joined #gluster-dev
10:34 schandra tigert, away?
10:38 tigert no
11:04 nishanth joined #gluster-dev
11:06 spandit joined #gluster-dev
11:08 overclk joined #gluster-dev
11:13 spandit joined #gluster-dev
11:43 nishanth joined #gluster-dev
11:48 anrao joined #gluster-dev
11:49 hgowtham joined #gluster-dev
11:57 rgustafs joined #gluster-dev
11:59 kkeithley ndevos: AFAIK there's no bug for that.
12:00 ndevos no bug, no fix
12:02 hagarth1 joined #gluster-dev
12:02 badone_ joined #gluster-dev
12:03 * ndevos thought kkeithley would immediately file a bug and request for a fix
12:05 ndevos REMINDER: Gluster Community Meeting *now* happening in #gluster-meeting
12:08 ashiq joined #gluster-dev
12:10 rraja joined #gluster-dev
12:13 * kkeithley curses debian packaging again
12:15 ndevos kkeithley: like "[Gluster-users] Packaging error"?
12:25 kkeithley yeah, like that
12:31 itisravi joined #gluster-dev
12:37 dlambrig joined #gluster-dev
12:52 kkeithley ndevos: BZ 1223385, 1223390   ;-)
12:52 anoopcs tigert, planet gluster can take in blogs which are not related to gluster, right?
12:53 tigert well, that wouldnt make sense
12:53 tigert Planet Internet is kinda boring :)
12:53 tigert or has a lot of stuff
12:53 tigert you should filter by tag and tag posts that are related to gluster
12:53 tigert most blog tools can have tag / category specific feeds
12:54 tigert like many of the existing ones in planet
12:54 tigert http://raobharata.wordpress.com/tag/gluster/feed/ for example
12:54 tigert wow that is pretty old content there btw
12:54 ndevos kkeithley: oh, thank you!
12:55 tigert feel free to change those too if you think something is stale and no longer related to gluster
12:55 ndevos tigert: where should we place our avatar?
12:55 anoopcs tigert, I could see many other posts on planet.gluster.org, that's why
12:55 * tigert updates instructions
12:56 tigert yeah those feeds should be fixed
12:56 anoopcs tigert, Ok. Just need to confirm.. Otherwise planet gluster wouldn't make sense
12:56 tigert yeah
12:57 tigert sure
12:57 ndevos kkeithley: you assigned those bugs to yourself, does that mean I only need to review the patch and -1 the backport if you did not follow the DOCUMENTED guidelines?
12:57 tigert so for unrelated posts you should check the feed whether the author has tags
12:57 tigert and change the feed accordingly
12:57 kkeithley yup
12:57 tigert if you push to git, and the yml is valid, it should update automatically to the planet within a hour I think
12:58 ndevos tigert: who can merge pull requests?
12:58 kkeithley but if you wanted to jump in an fix them, and follow the guidelines, feel free
12:59 ndevos kkeithley: whatever you like, I can find something else to do too :)
12:59 tigert ndevos: me, and its in the gluster org afaik
13:01 ndevos tigert: oh, okay, maybe you want to find some more people, ask for volunteers somewhere?
13:01 tigert I think we need to setup teams on github a bit better
13:01 tigert for web and docs admins
13:01 tigert gotta think yeah
13:01 tigert now I got to run, ttyl!
13:02 ndevos cya tigert!
13:02 kshlm ndevos, who's doing the forward porting of mt-shd?
13:03 kshlm Has it been decided yet?
13:03 ndevos kshlm: I dont know, I guess the afr guys would look into that
13:04 kshlm I was thinking of taking a shot at it.
13:04 nkhare joined #gluster-dev
13:04 ndevos kshlm: I'm sure they would appreciate that :)
13:05 kshlm I think I'll do it. How hard could it be? :)
13:06 ndevos indeed, it can not be *that* hard, but you might need to fix the coding style too, that can be time consuming
13:06 hchiramm__ kkeithley++
13:06 glusterbot hchiramm__: kkeithley's karma is now 69
13:15 sks joined #gluster-dev
13:15 firemanxbr joined #gluster-dev
13:20 shyam joined #gluster-dev
13:22 rgustafs joined #gluster-dev
13:25 badone_ joined #gluster-dev
13:27 vimal joined #gluster-dev
13:33 wushudoin joined #gluster-dev
13:37 pranithk xavih: Did you start working on ec_get_size_version?
13:38 xavih pranithk: I've just sent the patch for cooperative locking (http://review.gluster.org/10845/). I can start working on it right now if you want.
13:38 pranithk xavih: Isn't it a big change?
13:38 pranithk xavih: I saw your patch.
13:38 pranithk xavih: reasonably bigone...
13:39 xavih pranithk: I think it could be done in an easier way. My idea was to remove ec_get_size_version() call from managers, and include a flag in ec_lock().
13:39 xavih pranithk: after a successful lock, an xattrop whould be sent if the flag is set
13:40 xavih pranithk: fop's manager will continue once the lock and, optionally, the ec_get_size_version, have completed
13:40 xavih pranithk: on unlock it's a similar process
13:41 xavih pranithk: do you think it's a too big change ?
13:41 pranithk xavih: How much time do you think it will take for you... I am on a timeline for this change :-). If you think you can complete it by tomorrow afternoon India time, please feel free to otherwise I can take it up.
13:42 pranithk xavih: You just guide me if you think the timeline is too much :-)
13:42 xavih pranithk: I'll do
13:42 xavih pranithk: are you ok with the approach ?
13:43 pranithk xavih: let me check the code once...
13:45 pranithk xavih: If I understand it correctly ec_lock takes one lock after other, won't the latency increase for rename this way? We could wind parallel get_size_versions if we do it outside ec_lock?
13:46 pranithk xavih: Or in ec_lock() after all the locks are issued you will call xattrops in parallel? that can be fine as well I guess...
13:46 xavih pranithk: yes, I can't take both locks at the same time because this could cause a deadlock
13:47 xavih pranithk: to reduce latency I plan to immediately call xattrop just after each successful lock to avoid adding even more latency
13:47 pranithk xavih: yes. But xattrops can be wound in parallel. Okay that step is done. Next is how do we make sure versions are gathered data part when entry lock is taken, metadata part when inode lock is taken for directories...
13:47 xavih pranithk: I plan to call xattrop after a successful lock and in parallel with the next lock
13:48 xavih pranithk: no unless I wait until all locks are taken
13:49 xavih pranithk: we can check lock type and fop id for this
13:49 pranithk xavih: not fop->id,  inode->ia_type
13:49 pranithk xavih: setattr can come on directories as well...
13:49 xavih pranithk: right
13:49 pranithk xavih: well any metadata fop...
13:50 pranithk xavih: so lock type and inode->ia_type
13:50 xavih pranithk: yes. I think this should work
13:50 pranithk xavih: yes, seems okay I guess :-)
13:50 pranithk xavih: Wait
13:50 pranithk xavih: You were saying you want to do this for prepare_update?
13:51 xavih pranithk: in fact I was planning to merge ec_prepare_update and ec_get_size_version because they basically do the same with some small differences
13:52 xavih pranithk: and make all fops to use the same function with small differences depending on an 'update' flag
13:52 pranithk xavih: If we do this lock + xattrop then prepare_update will give problem... because it will wind xattrop with +1.
13:52 pranithk xavih: but we use prepare_update only on write and truncate so there is only one lock. So it won't be a problem...
13:53 xavih pranithk: no, no, prepare_update will also be removed from fop managers
13:53 pranithk xavih: okay let me tell the problem...
13:53 xavih pranithk: it will be used internally from ec_lock() using the flags I mentioned
13:53 pranithk xavih: If rename were to do prepare_update then on first lock we would have wound xattrop +1 for dirty but lock for next entry could fail leading to un-necessary dirty flag on soruce directory...
13:54 pranithk xavih: if it is just get_size_version, then there is no problem.
13:54 pranithk xavih: because it doesn't really modify anything...
13:55 pranithk xavih: Am I making any sense?
13:55 xavih pranithk: a rename modifies both source and destination directories. Unless we are not considering dirty flag for directories, it should be incremented
13:56 pranithk xavih: We do dirty flag modification at the end of transaction not in prepare_update for directory operations...
13:56 xavih pranithk: what do you mean with: "but lock for next entry could fail"
13:57 pranithk xavih: But if we were to do and lock succeeds only on source and then we wind xattrop +1 for dirty even before we attempt the lock for destination. Lock on destination could fail which will lead to failure of rename but 'dirty' flag on source will remain...
13:58 xavih pranithk: but we need to mark both directories dirty. There will be always some combination that causes this effect because we cannot do both operations atomically
13:59 pranithk xavih: Hmm.. you are right!
13:59 pranithk xavih: I guess this change should be fine :-)
13:59 xavih pranithk: on unlock of the second lock (after a failure of the first one), we'll decrease the dirty flag
14:00 xavih pranithk: if a crash happens in the middle, self-heal will take care of that (in fact there's no modification so self-heal will basically clear the dirty flag)
14:00 pranithk xavih: yes, don't do that decreasing part, for it we need to tell unlock where the fop failed etc.
14:01 xavih pranithk: ok. I'll start working on this. As soon as I've something I'll upload it so that you can see the changes and verify that everything is ok :)
14:02 pranithk xavih: Yeah, in case you get pulled into something send me the patch, since we kind of know the solution, I can finish the work...
14:02 xavih pranithk: ok, thanks :)
14:03 pranithk xavih: I am still not as fast as you for changing ec code :-( will take 1 or 2 releases I guess :-)
14:08 pranithk xavih: one more question
14:08 pranithk xavih: this change you are going to do could conflict with co-operative locking you just sent right?
14:12 xavih pranithk: most probably. I'm doing it on top of the cooperative locking change to avoid unnecessary rebases later
14:12 pranithk xavih: brilliant!
14:13 pranithk xavih: Are there any things you planned for ec which are big but we are not in a hurry?
14:13 pranithk xavih: It would be nice if I touch ec transaction a lot more so that I get better at ec...
14:15 pranithk xavih: I guess you already started coding :-). Okay we shall talk about it later..
14:16 purpleidea joined #gluster-dev
14:16 pranithk purpleidea: long time no see!
14:20 * pranithk thinks everybody is busy except me. That's it I'm off to testing my locks patch
14:20 hagarth pranithk: are you so bored?
14:20 * hagarth is never busy ;)
14:21 pranithk hagarth: not bored :-).
14:21 xavih pranithk: I've some ideas to improve ec, but this will need some more talk :P
14:21 pranithk xavih: Please let me know. You are churning out code so fast I feel like I am doing nothing in ec :'-(
14:21 xavih pranithk: do you think you are not doing anything ?!?!?!
14:22 xavih pranithk: I think self-heal daemon integration has been a lot of work...
14:22 pranithk xavih: we didn't get a chance to get good testing there... So once these bugs are fixed lot of pounding will happen there!
14:23 xavih pranithk: anyway it's a lot of work
14:23 pranithk xavih: yeah, it didn't get your complete review, so I don't completely know if data-self-heal I wrote is as good :-/
14:24 pranithk xavih: After you send this patch, please spend some time reviewing it. Tell me mistakes I did, and I will send patches to fix them.
14:24 xavih pranithk: sure :)
14:30 kkeithley hchiramm++
14:30 glusterbot kkeithley: hchiramm's karma is now 36
14:30 kkeithley hchiramm__++
14:30 glusterbot kkeithley: hchiramm__'s karma is now 2
15:07 kshlm I've forward ported the multi-threaded shd patch contributed by fb, and posted https://review.gluster.org/10851 for initial review.
15:07 kshlm The change works, but there are still a lot of changes, especially related to style, to be done.
15:09 kshlm pranithk, since you're not busy you could review https://review.gluster.org/10851
15:18 lpabon joined #gluster-dev
15:19 Gaurav_ joined #gluster-dev
15:19 ndevos kshlm: nice!
15:19 kkeithley left #gluster-dev
15:20 kkeithley joined #gluster-dev
15:37 purpleidea pranithk: heyy!
15:38 purpleidea pranithk: i've been around, but i guess hacking on different things. Oh-My-Vagrant and some other stuff mostly
15:40 ndevos hmm, why am *I* supposed to review https://review.gerrithub.io/233820 ?
15:40 ndevos kkeithley: maybe you want to have a look at ^ too?
15:58 dlambrig joined #gluster-dev
16:12 hagarth joined #gluster-dev
16:16 pranithk joined #gluster-dev
16:23 kanagaraj joined #gluster-dev
16:44 lpabon joined #gluster-dev
16:58 ndevos hagarth: I remember the item for the meeting I wanted to ask about!
16:58 ndevos I need someone that likes to code in Java, so that we can push https://github.com/jenkinsc​i/ssh-slaves-plugin/pull/16
16:59 ndevos it will allow us to shutdown/poweron Jenkins slaves automatically :)
16:59 Gaurav_ joined #gluster-dev
16:59 ndevos that should save us quite a bit in out Rackspace account
17:00 ndevos Gaurav_: welcome! are you a Java developer?
17:07 hagarth ndevos: we can ask some of the ovirt-gluster developers to help
17:10 Gaurav_ ndevos: hiii i am c/c++ developer
17:10 glusterbot Gaurav_: c/c's karma is now 5
17:10 Gaurav_ ndevos: is there any thing that need to be done using java
17:11 ndevos Gaurav_: just some merging/updating of https://github.com/jenkinsc​i/ssh-slaves-plugin/pull/16
17:11 ndevos hagarth: if those devs care and would like to have a go at it, that would help, otherwise I can do it too
17:12 * ndevos does not mind Java too much, rather Java than Perl
17:12 ndevos perl--
17:12 glusterbot ndevos: perl's karma is now -1
17:17 ndevos Gaurav++
17:17 glusterbot ndevos: Gaurav's karma is now 1
17:17 ndevos Gaurav_++ ?
17:17 glusterbot ndevos: Gaurav_'s karma is now 2
17:17 ndevos hagarth: solved! :D
17:19 hagarth ndevos++
17:19 glusterbot hagarth: ndevos's karma is now 127
17:22 kshlm joined #gluster-dev
18:14 hagarth joined #gluster-dev
18:51 dlambrig joined #gluster-dev
18:57 shyam Did http://review.gluster.org just go down?
19:13 pranithk1 joined #gluster-dev
19:14 kshlm shyam, the filesystem just got full!
19:14 kshlm We'd cleared some junk out a little while ago, but we didn't expect it to get full so soon.
19:14 shyam Yup, that is the error I got as well when I tried to commit something, when I went to take a look at the web interface...
19:15 shyam kshlm: FWIW "fatal: Unpack error, check server log error: unpack failed: error No space left on device To ssh://Shyam@git.gluster.org/glusterfs.git  ! [remote rejected] HEAD -> refs/for/master/bug-1222126 (n/a (unpacker error))"
19:16 kshlm I'm checking what filled the disk up quickly.
19:30 dlambrig joined #gluster-dev
21:02 ppai joined #gluster-dev
21:13 badone_ joined #gluster-dev
21:48 ndevos hahaha, I just got this badge: https://badges.fedoraproject​.org/badge/secretary-general
21:48 ndevos seems that zodbot in #gluster-meeting needs the correct username in the fedora account to give badges :)
21:49 * ndevos was asked to set his correct username last week
22:32 ppai joined #gluster-dev
22:43 dlambrig joined #gluster-dev

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