Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2014-11-26

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

All times shown according to UTC.

Time Nick Message
00:07 nishanth joined #gluster-dev
00:50 shyam joined #gluster-dev
01:14 bala joined #gluster-dev
01:35 topshare joined #gluster-dev
02:08 kshlm joined #gluster-dev
02:13 bharata-rao joined #gluster-dev
02:15 soumya joined #gluster-dev
02:18 kaushal_ joined #gluster-dev
02:47 ilbot3 joined #gluster-dev
02:47 Topic for #gluster-dev is now Gluster Development Channel - http://gluster.org | For general chat go to #gluster | Patches - http://review.gluster.org/ | Channel Logs - https://botbot.me/freenode/gluster-dev/ & http://irclog.perlgeek.de/gluster-dev/
03:12 soumya joined #gluster-dev
03:26 kdhananjay joined #gluster-dev
03:43 EinstCrazy joined #gluster-dev
03:53 kshlm joined #gluster-dev
04:00 ndarshan joined #gluster-dev
04:08 itisravi joined #gluster-dev
04:14 shubhendu joined #gluster-dev
04:23 anoopcs joined #gluster-dev
04:33 jiffin joined #gluster-dev
04:35 anoopcs joined #gluster-dev
04:40 anoopcs joined #gluster-dev
04:41 rafi1 joined #gluster-dev
04:47 shubhendu joined #gluster-dev
04:51 atinmu joined #gluster-dev
04:55 Gaurav_ joined #gluster-dev
05:07 lalatenduM joined #gluster-dev
05:10 vimal joined #gluster-dev
05:16 spandit joined #gluster-dev
05:21 pranithk joined #gluster-dev
05:21 nkhare joined #gluster-dev
05:22 shubhendu joined #gluster-dev
05:25 deepakcs joined #gluster-dev
05:30 bala joined #gluster-dev
05:34 jiffin joined #gluster-dev
05:34 hagarth joined #gluster-dev
05:37 jiffin1 joined #gluster-dev
05:42 soumya joined #gluster-dev
05:48 kotresh_ joined #gluster-dev
06:00 kdhananjay joined #gluster-dev
06:01 pp joined #gluster-dev
06:03 overclk joined #gluster-dev
06:10 ppai joined #gluster-dev
06:15 hagarth pranithk: ping, can you review patches sent by Emmanuel? there was a reminder on that last week.
06:18 bala joined #gluster-dev
06:32 nishanth joined #gluster-dev
06:40 soumya joined #gluster-dev
07:00 ppai joined #gluster-dev
07:28 azar joined #gluster-dev
07:33 rgustafs joined #gluster-dev
07:35 Debloper joined #gluster-dev
07:37 raghu` joined #gluster-dev
08:09 ppai joined #gluster-dev
08:22 Debloper joined #gluster-dev
08:26 ndevos hey anoopcs, are you aware that slave23 is hanging on a patcht from you? http://build.gluster.org/job/rackspace-regression-2GB-triggered/2862/console
08:34 ppai joined #gluster-dev
08:53 shubhendu joined #gluster-dev
08:58 topshare joined #gluster-dev
09:01 hagarth xavih: ping, do you intend to keep ec under gplv3+ only? rest of the code base is dual licensed under lgplv3+ or gplv2+
09:03 xavih hagarth: I'm not sure. I used GPLv3 because this was what gluster was using when I started, however I don't what legal implications it might have, if any...
09:03 xavih hagarth: do you think it's ok to change it ?
09:03 Debloper joined #gluster-dev
09:04 hagarth xavih: yes, we have also changed licensing in the past.
09:04 hagarth a few deployments might find it difficult if parts of glusterfs are under a different licensing scheme
09:05 xavih hagarth: ok, I'll update licensing info
09:05 hagarth xavih: cool, thanks!
09:05 hagarth xavih++
09:05 glusterbot hagarth: xavih's karma is now 9
09:06 xavih hagarth: yw :)
09:06 atalur joined #gluster-dev
09:11 shubhendu joined #gluster-dev
09:39 vimal joined #gluster-dev
10:12 kdhananjay joined #gluster-dev
10:52 kdhananjay joined #gluster-dev
11:12 ppai joined #gluster-dev
11:45 pranithk left #gluster-dev
11:49 soumya_ joined #gluster-dev
11:52 kanagaraj joined #gluster-dev
12:16 ppai joined #gluster-dev
12:20 bala joined #gluster-dev
12:29 itisravi_ joined #gluster-dev
12:29 atalur joined #gluster-dev
12:45 shubhendu joined #gluster-dev
12:52 atalur joined #gluster-dev
12:56 lalatenduM kkeithley, I think we should send the proposal , we have enough material :)
12:56 kkeithley yup
12:56 kkeithley you want to do it, or shall I?
12:56 * kkeithley votes for you
12:57 lalatenduM kkeithley,  will do the write up and send it to you for review :0
12:57 lalatenduM :)
12:57 kkeithley lalatenduM++
12:57 glusterbot kkeithley: lalatenduM's karma is now 50
12:58 lalatenduM kkeithley, I am thinking of propose another one "Running a SIG in CentOS"
12:58 kkeithley I like that.
12:59 lalatenduM kkeithley, kb suggested
12:59 rgustafs joined #gluster-dev
12:59 kkeithley lol
12:59 lalatenduM kkeithley, ? (though i can guess :))
12:59 kkeithley ?
13:00 lalatenduM kkeithley, why lol
13:00 kkeithley oh, sorry, no. Just laughing that kb suggested it and it's a good idea
13:00 lalatenduM kkeithley, np :)
13:00 kkeithley seems kind of obvious actually, funny that we didn't think of it.
13:01 kkeithley sorry, I hope I didn't offend
13:01 lalatenduM kkeithley, yeah , may be it was too obvious  :)
13:01 lalatenduM kkeithley, what abt devconf ? it is happening 1 week later
13:01 kkeithley correct
13:02 kkeithley running SIG in CentOS might be a good topic for a DevConf presentation too
13:02 lalatenduM are u going?
13:02 lalatenduM kkeithley, yeah I too think so
13:02 kkeithley time's running out for submissions there too.
13:02 kkeithley I hope the CFP isn't already closed
13:03 lalatenduM kkeithley, I think it is end of Nov
13:04 * kkeithley isn't sure why the deadline isn't on the CFP page
13:05 kkeithley that would be the obvious place for it.
13:06 kkeithley Yes I'm going to both. I've submitted a lightning talk for FOSDEM (I missed the deadline for a full presentation) and two talks at DevConf.
13:06 lalatenduM kkeithley, cool
13:07 kkeithley I hope they all get selected. But I'm pretty sure I'm going. Ric didn't say no when I told him.
13:07 kkeithley got to run out for ten minutes, biab
13:08 lalatenduM kkeithley, yeah, ttyl
13:10 vimal joined #gluster-dev
13:16 topshare joined #gluster-dev
13:21 edward1 joined #gluster-dev
13:36 bala joined #gluster-dev
13:40 rgustafs joined #gluster-dev
13:42 bala joined #gluster-dev
13:45 hagarth joined #gluster-dev
13:47 tdasilva joined #gluster-dev
13:48 eljrax So, getting back to this now. Did anybody know how to get the name of the file created in the create fop callback in a translator?
14:01 kkeithley I'd suggest in the foo_create() method, instead of calling STACK_WIND, call STACK_WIND_COOKIE(frame, rfn, loc, ....); then in the foo_create_cbk the cookie param has the loc and you can get the filename from there.
14:02 kkeithley I believe that should work
14:02 eljrax Alright, that sounds like it could work. I'll give that a go.
14:02 eljrax Hold my beer, I'm going in!
14:18 nishanth joined #gluster-dev
14:21 eljrax That worked, thanks!
14:24 kkeithley eljrax++
14:24 glusterbot kkeithley: eljrax's karma is now 1
14:26 eljrax Ah, a karma system. Cheers! And right back at you
14:26 eljrax kkeithley++
14:26 glusterbot eljrax: kkeithley's karma is now 44
14:26 kkeithley lol
14:26 topshare joined #gluster-dev
14:28 hagarth eljrax: you can also use frame->local to retrieve any information that you need in the callback
14:32 eljrax What exactly does that point to? I just see void *
14:42 hagarth free to hold your data type
14:43 kkeithley see .../xlators/cluster/dht/src/dht-helper.c::dht_local_init()
14:45 eljrax Oh right, that's nifty!
14:45 kkeithley what happens when multiple xlators store something at frame->local? I've always been a bit puzzled by that (but not curious enough to figure out why)
14:45 eljrax Might stick to the cookie-solution then
14:45 hagarth kkeithley: frame->local is local to each translator (it is the frame's local like in the program layout)
14:46 kkeithley oh, okay, kinda like thread-specific-storage
14:47 hagarth in a sense yes.. STACK_WIND creates a new frame for each translator
14:47 kkeithley so every stack wind and unwind replaces frame->local
14:47 hagarth when you wind, a frame gets pushed to the STACK
14:47 hagarth when you unwind, a frame gets popped out
14:47 kkeithley ah, that explains it. thanks
14:48 kkeithley I've even seen that in the STACK_WIND macro and forgot about it until you just mentioned it.
14:49 nishanth joined #gluster-dev
14:50 eljrax So frame->local is accessible from any fop or cbk within that translator?
14:50 kkeithley yes
14:51 kkeithley both foo_create() and foo_create_cbk() have a frame parameter
14:51 kkeithley s/have/are passed/
14:52 eljrax Ah you had me confused there for a moment
14:53 kkeithley you can set frame->local in foo_create(), get it in foo_create_cbk()
14:56 kanagaraj joined #gluster-dev
15:00 atinmu joined #gluster-dev
15:05 soumya_ joined #gluster-dev
15:12 eljrax If I load this translator on two servers, the cookie is only available on the server which handles the create event. Is that by design, or am I doing something weird? It's actually good for my use-case, but I don't want to rely on that if that's not intended
15:13 eljrax So if I make a write on a client, served by server1, if I gf_log the value of the cookie, it's set. But on server2 gf_log logs (null)
15:13 eljrax I pass the cookie as strdup(loc->path)
15:15 _Bryan_ joined #gluster-dev
15:15 wushudoin joined #gluster-dev
15:17 shyam joined #gluster-dev
15:34 rafi1 joined #gluster-dev
15:42 rafi1 joined #gluster-dev
15:56 anoopcs joined #gluster-dev
15:58 rafi1 joined #gluster-dev
16:00 kkeithley eljrax: DHT (distribute) volume?  Then yes, you'll only see a file created on one server.
16:07 bala joined #gluster-dev
16:07 eljrax kkeithley: No, the file is created on both servers. They're replicated
16:07 eljrax And the translator runs on both servers. But the value of the cookie is only set on one of them
16:10 rafi1 joined #gluster-dev
16:12 kkeithley And you installed your new translator  .so on both servers and restarted the volume?
16:13 eljrax http://fpaste.org/154339/01841114/
16:13 eljrax Yeah
16:14 eljrax Line 11 outputs the file name on the server that handled the request, but (null) on the server which got the event replicated to it
16:16 jobewan joined #gluster-dev
16:16 kkeithley in a replica volume the client writes to both servers in parallel. There's no replication event. If there's a loc parameter in the foo_create() FOP on one, it should be there on the other. Off hand I don't know why there would be any difference.
16:17 kkeithley any difference between the two
16:18 eljrax Ok, let me have a good rummage around then. I'm sure it's me doing something somewhere then
16:20 eljrax That's a shame though, it'd been neat for me if that was meant to happen. I want to put a job on a queue for each file create, delete and modification. Now I'm going to have to account for each job being put on the queue $replica_count times
16:49 hagarth joined #gluster-dev
16:53 ryao joined #gluster-dev
16:59 ryao joined #gluster-dev
17:24 Gaurav_ joined #gluster-dev
17:45 lalatenduM joined #gluster-dev
17:46 Gaurav__ joined #gluster-dev
19:38 rafi1 joined #gluster-dev
19:49 shyam joined #gluster-dev
20:02 xavih joined #gluster-dev
21:12 rafi1 joined #gluster-dev
21:18 badone joined #gluster-dev
21:33 badone joined #gluster-dev
21:41 tdasilva joined #gluster-dev

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