Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2013-12-05

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

All times shown according to UTC.

Time Nick Message
00:53 Technicool joined #gluster-dev
01:26 itisravi joined #gluster-dev
01:31 lpabon joined #gluster-dev
01:36 awheeler joined #gluster-dev
01:46 awheeler joined #gluster-dev
02:36 awheeler joined #gluster-dev
02:55 itisravi joined #gluster-dev
03:03 awheeler joined #gluster-dev
03:13 kshlm joined #gluster-dev
03:21 awheeler joined #gluster-dev
03:37 shubhendu joined #gluster-dev
03:50 itisravi joined #gluster-dev
03:50 bharata-rao joined #gluster-dev
03:51 itisravi_ joined #gluster-dev
03:57 mohankumar joined #gluster-dev
03:59 kanagaraj joined #gluster-dev
04:10 ndarshan joined #gluster-dev
04:17 awheeler joined #gluster-dev
04:22 _BryanHm_ joined #gluster-dev
04:35 raghu joined #gluster-dev
04:36 ppai joined #gluster-dev
04:38 shyam joined #gluster-dev
04:58 hagarth joined #gluster-dev
05:01 bala joined #gluster-dev
05:12 Technicool joined #gluster-dev
05:20 awheeler joined #gluster-dev
05:30 bulde joined #gluster-dev
05:30 Eco_ joined #gluster-dev
05:37 spandit joined #gluster-dev
05:42 aravindavk joined #gluster-dev
05:48 Eco_ joined #gluster-dev
05:48 badone joined #gluster-dev
05:50 ababu joined #gluster-dev
05:54 shubhendu joined #gluster-dev
05:57 ndarshan joined #gluster-dev
05:58 bala joined #gluster-dev
06:01 kanagaraj joined #gluster-dev
06:10 hagarth joined #gluster-dev
06:19 lalatenduM joined #gluster-dev
06:22 spandit joined #gluster-dev
06:31 awheeler joined #gluster-dev
06:50 foster joined #gluster-dev
06:51 kanagaraj joined #gluster-dev
06:52 ndarshan joined #gluster-dev
06:52 bala joined #gluster-dev
06:53 aravindavk joined #gluster-dev
06:53 hagarth joined #gluster-dev
07:51 hagarth joined #gluster-dev
07:56 shubhendu joined #gluster-dev
08:19 shubhendu joined #gluster-dev
08:24 shyam joined #gluster-dev
08:27 vshankar joined #gluster-dev
08:55 bharata-rao hagarth, ping
09:06 ndarshan joined #gluster-dev
09:20 ndarshan joined #gluster-dev
09:30 hagarth bharata-rao: pong
09:30 bharata-rao hagarth, Saw that glfs_fini closes the log file (ctx->log.logfile) which can be a problem with qemu
09:31 bharata-rao hagarth, qemu uses stderr for it's own logging as well as sets glusterfs logfile to stderr
09:31 bharata-rao hagarth, if glfs_fini closes stderr, then subsequent qemu error messages are lost
09:31 bharata-rao hagarth, I wonder if glfs_fini should close the logfile only if it not stderr
09:33 hagarth bharata-rao: it might be not possible for glfs api to determine if it is stderr or not
09:34 bharata-rao hagarth, if (ctx->log.logfile && ctx->log.logfile != stderr) fclose(ctx->log.logfile)
09:35 hagarth bharata-rao: with daemons, the regular descriptors usually won't refer to std{in,out,err}
09:36 bharata-rao hagarth, I am talking about the client side logging enabled by qemu using glfs_set_logging()
09:36 bharata-rao hagarth, qemu uses this...
09:37 bharata-rao glfs_set_logging(glfs, "-", 4);
09:37 bharata-rao hagarth, and libgfapi sets the logfile to stderr in response to that
09:39 bharata-rao hagarth, check libglusterfs/src/logging.c:375
09:43 hagarth bharata-rao: agree, when you pass "-" glfs_api just goes ahead and uses the value it gets (2 in this case)
09:43 bharata-rao hagarth, yes and it closes stderr when glfs_fini is called which causes problems to qemu
09:48 hagarth bharata-rao: since logfile is a char *, we would have to perform a strncmp .. with "-", "/dev/stderr" etc.
09:49 bharata-rao hagarth, I think ctx->log.logfile is FILE *
09:51 bharata-rao hagarth, The code I pasted first (change in glfs_fini) should do the trick, I am just wondering if this is indeed the correct way to address the problem
09:53 ndarshan joined #gluster-dev
10:07 bharata-rao hagarth, shall a send a patch to fix this ?
10:28 shyam joined #gluster-dev
10:31 ppai joined #gluster-dev
10:55 hagarth bharata-rao: ctx->log.logfile is a char *
10:56 hagarth bharata-rao: oops filename is a char *
10:56 bharata-rao hagarth, typedef struct gf_log_handle_ {   FILE            *logfile; } :)
10:59 hagarth bharata-rao: yeah, give me a moment
10:59 bharata-rao hagarth, sure
11:09 ndarshan joined #gluster-dev
11:11 ppai joined #gluster-dev
11:56 ira joined #gluster-dev
12:03 kkeithley1 joined #gluster-dev
12:16 itisravi joined #gluster-dev
12:18 ppai joined #gluster-dev
12:31 edward2 joined #gluster-dev
12:31 edward2 joined #gluster-dev
12:55 hagarth1 joined #gluster-dev
13:35 lalatenduM joined #gluster-dev
14:06 davidbierce joined #gluster-dev
14:23 awheeler joined #gluster-dev
14:29 ababu joined #gluster-dev
14:50 bala joined #gluster-dev
15:04 lalatenduM joined #gluster-dev
15:18 bala1 joined #gluster-dev
15:20 lpabon joined #gluster-dev
15:27 bala joined #gluster-dev
15:33 wushudoin joined #gluster-dev
16:15 shyam joined #gluster-dev
16:19 ababu joined #gluster-dev
16:42 Eco_ joined #gluster-dev
16:44 shyam joined #gluster-dev
17:00 semiosis anyone have an opinion on bug 1038391? https://bugzilla.redhat.co​m/show_bug.cgi?id=1038391
17:00 glusterbot Bug 1038391: unspecified, unspecified, ---, amarts, NEW , gf-error-codes.h is missing from source tarball
17:00 glusterbot Bug http://goo.gl/LrzFhH unspecified, unspecified, ---, amarts, NEW , gf-error-codes.h is missing from source tarball
17:00 semiosis synopsis, need to run autogen.sh before configure to generate gf-error-codes.h in release (qa) source tarball
17:00 semiosis never before have we had to run autogen.sh on a release tarball
17:02 ndk joined #gluster-dev
17:22 hagarth joined #gluster-dev
17:23 awheeler joined #gluster-dev
17:23 johnmark kkeithley_: ping
17:23 johnmark a2: ping
17:23 johnmark ^^^ any idea if there's *any* documentation for gfapi?
17:24 johnmark even if in some stage of incompleteness
17:24 johnmark hagarth: ^^^
17:25 hagarth johnmark: https://github.com/gluster/glust​erfs/blob/master/api/src/glfs.h
17:25 hagarth this is the best that we have got
17:26 johnmark hagarth: ok :)
17:26 johnmark I thought jdarcy was supposed to be working on that :/
17:26 johnmark hagarth: and I don't know if the docs team ever tackled it for the sake of RHS releases
17:27 hagarth johnmark: i don't think the docs team has written anything for RHS either on libgfapi
17:32 johnmark hagarth: ok
17:32 johnmark just wanted to check
17:40 semiosis johnmark: hahahaha
17:40 semiosis i thought there were no docs because it's 99% the standard posix api and anyone using it is supposed to "just know" it
17:41 semiosis i referred to the system man pages when i needed help
17:41 semiosis and have managed OK so far with that
17:41 semiosis working on the java filesystem
17:44 semiosis for example, the stat(2) man page covers stat, fstat, and lstat, which correspond to the libgfapi functions glfs_stat, glfs_lstat, glfs_fstat
17:59 johnmark semiosis: oh sweet - good to know :)
18:00 johnmark semiosis: for the uninitiated, there will, however, be demand for "how the hell do I use this?" docs
18:00 johnmark not everyone knows what they're doing :)
18:00 johnmark s/not/most/
18:00 johnmark er wait
18:00 johnmark s/not everyone/almost no one/g
18:00 johnmark heh
18:00 semiosis yeah it was painful
18:01 semiosis i had no prior experience with the vfs api
18:01 johnmark semiosis: speaking of... how's the java filesystem work going these days?
18:01 johnmark I want to interview you about that
18:01 semiosis slow but progressing
18:01 johnmark (as well as some of hte other projects there)
18:02 johnmark purpleidea: ^^^
18:02 semiosis you know, joe brockmeier interviewed me for some redhat community blog, but afaik it was never published :/
18:02 johnmark semiosis: hrm... let me check with him
18:03 semiosis w/e you want to do, i'm all for it
18:03 johnmark aight - thanks
18:03 johnmark and I meant on a google hangout, so we could publish on youtube, etc.
18:03 johnmark but it would be good to get it all at the same time
18:04 johnmark semiosis: ask and ye shall receive:
18:04 johnmark < jbrooks> Joe told me he's shooting for early next week
18:04 johnmark so there's that
18:04 semiosis neat!
18:04 johnmark semiosis: speaking of, was it jzb or jbrooks interviewing you for a case study?
18:05 semiosis uhhh, i dont remember.... i think i dropped the ball on that
18:05 johnmark semiosis: ah, ok
18:06 semiosis i wrote some stuff down after the summit following a skeleton document david nally gave me.  let me see if i can find it
18:06 johnmark semiosis: ah, nice. ok
18:06 semiosis wrote it on the plane ride back & then promptly forgot about it :(
18:10 johnmark ha! ok
18:12 semiosis found the doc.  going to give it a refresh & send it to you tonight
18:17 kkeithley_ johnmark: pong
18:26 johnmark kkeithley_: know if jdarcy has worked up any docs on gfapi?
18:27 kkeithley_ Not that I know of
18:28 johnmark ok
18:28 kkeithley_ the gfapi header has javadoc-style comments for each of the entry points.
18:30 johnmark kkeithley_: ok
18:30 * johnmark takes a lookie-loo
18:34 semiosis [12:25] <hagarth> johnmark: https://github.com/gluster/glust​erfs/blob/master/api/src/glfs.h
18:35 johnmark yes, I know
21:23 Eco_ joined #gluster-dev
22:13 purpleidea johnmark: o hai
22:29 badone joined #gluster-dev
23:02 awheeler joined #gluster-dev
23:24 _BryanHm_ joined #gluster-dev
23:38 bala joined #gluster-dev

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