Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2015-07-31

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

All times shown according to UTC.

Time Nick Message
01:45 pranithk joined #gluster-dev
01:51 pranithk joined #gluster-dev
02:40 morph- joined #gluster-dev
02:48 dlambrig_ joined #gluster-dev
03:13 kdhananjay joined #gluster-dev
03:28 shubhendu joined #gluster-dev
03:32 kdhananjay joined #gluster-dev
03:46 Gaurav__ joined #gluster-dev
03:51 overclk joined #gluster-dev
03:51 atinm joined #gluster-dev
03:56 byreddy_ joined #gluster-dev
04:07 sakshi joined #gluster-dev
04:10 kanagaraj joined #gluster-dev
04:14 nishanth joined #gluster-dev
04:18 overclk joined #gluster-dev
04:31 ppai joined #gluster-dev
04:35 nbalacha joined #gluster-dev
04:39 dlambrig_ joined #gluster-dev
04:41 kshlm joined #gluster-dev
04:50 ndarshan joined #gluster-dev
04:50 overclk_ joined #gluster-dev
04:58 deepakcs joined #gluster-dev
05:01 ndarshan joined #gluster-dev
05:03 pppp joined #gluster-dev
05:06 apandey joined #gluster-dev
05:08 kotreshhr joined #gluster-dev
05:08 vimal joined #gluster-dev
05:08 kotreshhr joined #gluster-dev
05:13 gem_ joined #gluster-dev
05:22 rafi joined #gluster-dev
05:24 Saravana_ joined #gluster-dev
05:27 kdhananjay joined #gluster-dev
05:29 aravindavk joined #gluster-dev
05:30 primusinterpares joined #gluster-dev
05:35 Bhaskarakiran joined #gluster-dev
05:40 Gaurav__ joined #gluster-dev
05:45 anrao joined #gluster-dev
05:49 anekkunt joined #gluster-dev
05:55 Manikandan joined #gluster-dev
05:55 hgowtham joined #gluster-dev
05:56 kdhananjay joined #gluster-dev
05:56 rjoseph joined #gluster-dev
05:57 ashiq joined #gluster-dev
05:59 hagarth joined #gluster-dev
06:04 skoduri joined #gluster-dev
06:06 pranithk joined #gluster-dev
06:09 apandey joined #gluster-dev
06:13 kdhananjay joined #gluster-dev
06:14 aravindavk joined #gluster-dev
06:23 vmallika joined #gluster-dev
06:28 ppai joined #gluster-dev
06:33 nbalachandran_ joined #gluster-dev
06:34 atalur joined #gluster-dev
06:40 spalai joined #gluster-dev
06:41 overclk joined #gluster-dev
06:51 overclk joined #gluster-dev
06:51 pranithk ndevos: Could you take a look at http://review.gluster.org/#/c/11796/
07:05 Manikandan joined #gluster-dev
07:13 overclk_ joined #gluster-dev
07:21 anrao joined #gluster-dev
07:30 ndevos pranithk: looks good, but the bug it is filed against is already closed?
07:48 sakshi joined #gluster-dev
07:54 Manikandan joined #gluster-dev
07:55 overclk joined #gluster-dev
08:03 pranithk ndevos: damn!, I will raise a new one
08:04 nbalachandran_ joined #gluster-dev
08:30 kdhananjay joined #gluster-dev
08:45 ppai joined #gluster-dev
08:48 overclk joined #gluster-dev
08:55 spalai joined #gluster-dev
08:59 anrao joined #gluster-dev
09:03 csim so, after installing everything blocking the configure script of gluster in smoke test, it give this result :
09:03 csim http://fpaste.org/250086/38333384/
09:03 csim are we ok with having a few "no" on a builder ?
09:07 ppai joined #gluster-dev
09:08 ndevos csim: yes, that should be fine
09:08 ndevos csim: I;m not sure if we have regression tests that use aio, maybe install the -devel package for that
09:09 ndevos that would be libaio-devel
09:10 ndevos and if you add ncurses-devel and readline-devel, one more "no" should be gone
09:11 csim ndevos: then, should we force the option in the configure script, to make sure we detect regression in installation ?
09:12 ndevos csim: no, not really, the regression tests do not need to get run for (normal) building
09:13 csim ndevos: oh, i do not want to have it in configure.ac, more on the ./configure somewhere in the jenkins build
09:13 csim also, the script fail with :
09:13 csim /usr/bin/mkdir: cannot create directory ‘/usr/lib/python2.7/site-packages/gluster’: Permission denied
09:14 ndevos csim: ah, right, there is a build.sh script somewhere, you mean to add --enable-<blah> to that?
09:16 csim ndevos: yep
09:17 ndevos csim: I do not see a real issue with that, but some glusterfs versions add more ./configure options, the build.sh is version independent :-/
09:17 csim ndevos: great, you just ruined my hope :p
09:18 nishanth joined #gluster-dev
09:18 ndevos csim: sorry, but you may almost start your weekend :)
09:19 overclk joined #gluster-dev
09:19 ndevos csim: also, this should have a copy/paste ready list of packages: http://gluster.readthedocs.org/en/lates​t/Developer-guide/Building%20GlusterFS/
09:20 csim ndevos: thanks, I will see which one are missing, as i was on try/error/install for now :)
09:20 ndarshan joined #gluster-dev
09:21 ndevos csim: that works and is surely more fun, but you may be able to save some time
09:22 shubhendu joined #gluster-dev
09:23 csim ndevos: it keep my skills fresh :p
09:23 csim as a packager, i had to do that quite often
09:24 ndevos csim: want to help out with the glusterfs packaging? ;-)
09:25 ndevos pranithk: could you poke someone that knows about afr to check out http://thread.gmane.org/gmane.comp.file​-systems.gluster.user/21588/focus=21649 ?
09:27 csim ndevos: well, I can give guidance in following best practice, or do package review if needed for fedora
09:28 ndevos csim: I meant more for the maintaining of the package, kkeithley mostly does the fedora updates now, but I'm sure he like to share the work
09:29 csim ndevos: I can hardly do jenkins stuff, so I do not think I will be able to sustain any help :/
09:30 ndevos csim: this is fedora packaging, and centos storage-sig, it has nothing to do with jenkins directly :)
09:30 csim ndevos: well, more "I do not have much free time" :)
09:31 ndevos csim: yeah, that one I was expecting :D
09:34 kdhananjay ndevos: I had a chat with the user (Geoffrey) last evening. He says he was able to manually resolve the split-brains.
09:35 ndevos kdhananjay++ ok, thanks! could you send a short reply to that email, so that we can all see that there was some assistance?
09:35 glusterbot ndevos: kdhananjay's karma is now 7
09:35 kdhananjay Sure ndevos.
09:36 ndevos kdhananjay: great :)
09:43 kaushal_ joined #gluster-dev
09:48 sakshi joined #gluster-dev
09:53 csim so, I do not understand how the smoke test is supposed to work
09:54 csim because + make install 'CFLAGS=-g -O0 -Wall -Werror' -j 4
09:54 csim if run as non root is not gonna work
09:54 csim or... this is a actual bug ?
10:04 overclk joined #gluster-dev
10:06 ppai joined #gluster-dev
10:08 anrao joined #gluster-dev
10:19 schandra joined #gluster-dev
10:20 sakshi joined #gluster-dev
10:22 pranithk left #gluster-dev
10:31 rafi1 joined #gluster-dev
10:36 shubhendu joined #gluster-dev
10:40 skoduri joined #gluster-dev
10:42 Manikandan joined #gluster-dev
10:49 sakshi joined #gluster-dev
10:56 josferna joined #gluster-dev
11:00 gem joined #gluster-dev
11:02 spalai joined #gluster-dev
11:11 arao joined #gluster-dev
11:13 sakshi joined #gluster-dev
11:16 Manikandan joined #gluster-dev
11:19 kotreshhr left #gluster-dev
11:24 dlambrig_ joined #gluster-dev
11:26 firemanxbr joined #gluster-dev
11:33 overclk joined #gluster-dev
11:35 arao joined #gluster-dev
11:35 aspandey joined #gluster-dev
11:49 rafi joined #gluster-dev
11:51 gem joined #gluster-dev
11:54 overclk joined #gluster-dev
11:54 skoduri joined #gluster-dev
11:57 Manikandan joined #gluster-dev
12:02 sakshi joined #gluster-dev
12:11 ndarshan joined #gluster-dev
12:14 hagarth joined #gluster-dev
12:16 arao joined #gluster-dev
12:21 Manikandan joined #gluster-dev
12:35 kotreshhr joined #gluster-dev
12:40 kotreshhr left #gluster-dev
12:46 arao joined #gluster-dev
12:54 spalai joined #gluster-dev
13:00 overclk joined #gluster-dev
13:04 overclk_ joined #gluster-dev
13:15 spalai joined #gluster-dev
13:30 spalai joined #gluster-dev
13:35 rjoseph joined #gluster-dev
13:37 spalai joined #gluster-dev
13:44 arao joined #gluster-dev
13:44 overclk joined #gluster-dev
14:01 spalai joined #gluster-dev
14:27 arao joined #gluster-dev
14:41 ndevos csim: uh, yeah, that CFLAGS looks strange, why would the "-j 4" be outside of the ''?
14:42 ndevos also, -j4 is a make option, not sure what it has to do in the CFLAGS
14:42 ndevos ah! its even stranger, that CFLAGS should get expanded, not included in full in ''
14:43 csim ndevos: I am preparing a patch
14:43 ndevos csim: you can pass variables on the 'make' command line, I think that is done here, but the '' are placed awkward
14:43 csim because make install as jenkins fail
14:44 ndevos ok :)
14:48 anekkunt joined #gluster-dev
14:49 ndevos kkeithley++ thanks for the compiler/linking corrections and additions :)
14:49 glusterbot ndevos: kkeithley's karma is now 88
14:52 arao joined #gluster-dev
14:55 nbalachandran_ joined #gluster-dev
14:58 kkeithley yw
15:04 arao joined #gluster-dev
15:13 pousley joined #gluster-dev
15:26 samikshan joined #gluster-dev
15:26 tdasilva joined #gluster-dev
15:43 skoduri joined #gluster-dev
15:55 ndevos kkeithley: do you have a preference for solving (most of) the linking/loading issues?
15:55 ndevos at the moment, I tend to think that C is the most suitable
16:13 arao joined #gluster-dev
16:15 cholcombe joined #gluster-dev
16:24 kkeithley By C you mean namespace/prefix?
16:26 kkeithley That's going to touch a lot of files
16:26 kkeithley and might be prone to mistakes
16:27 kkeithley although any solution has the possibility of mistakes
16:29 cholcombe kkeithley: i managed to get the rpc command working :)
16:30 kkeithley cholcombe++
16:30 glusterbot kkeithley: cholcombe's karma is now 1
16:30 cholcombe :)
16:30 cholcombe i'm confused about the rpc interface though.  there's a generic rpc_call function that passes an opaque dict around.  Why go with that instead of having 1 call per command?
16:31 cholcombe cause it leads me to having to parse and understand the dict contents
16:31 kkeithley I suspect it's been like that since the beginning.
16:37 kkeithley but that's a good point. You might want to raise that as an issue to the the glusterd-2.0 people.
16:45 cholcombe will do
16:45 cholcombe i wanted to use the rpc system so i didn't have to do so much regex lol.
16:45 cholcombe turns out that wasn't the case.  You have to regex the dict results
16:46 cholcombe kkeithley: should I email the devel list about it?  Think they would be responsive to changing it?
16:49 kkeithley yes, send mail to -devel. Certainly they will consider it, and I would think would not reject it out of hand. (But I can't promise that they would change it.)
16:50 cholcombe sure.  yeah i understand
16:50 cholcombe i just didn't know if they would just dismiss suggestions out of hand :)
16:51 kkeithley No, they won't. They're pretty good that way.
16:51 cholcombe sweet
16:51 cholcombe are they rewriting it in Go or did that get shelved?
16:51 ndevos kkeithley: C) Another possiblity is to link with --version-script
16:52 kkeithley No, Go hasn't been shelved.
16:52 cholcombe wow interesting
16:53 kkeithley ndevos: haha. Option C, not "C" language. Yes, I'm leaning toward that as well.
16:53 ndevos kkeithley: lol, yes *that* C
16:58 csim glusterfs-patch-acceptance-tests need to go to gerrit, or that's github review ?
16:58 ndevos csim: github only I think
16:59 kkeithley ndevos: better than --version-script is libtool -export-symbols.   On linux it just does --version-script. On Mac maybe it will do something useful
16:59 ndevos kkeithley: well, sure, I only copy/pasted the first few words of C
17:00 kkeithley yeah sorry, that's just me being pedantic (or anal ;-))
17:02 overclk joined #gluster-dev
17:07 * ndevos starts his weekend, cya!
17:09 csim is someone in charge of merging stuff on github : https://github.com/gluster/gluste​rfs-patch-acceptance-tests/pulls ?
17:12 cholcombe kkeithley: is there a spot in the docs or code where it says what params are packed into the dict's for rpc calls?  I couldn't find anything other than break pointing the code and tracing the source
17:15 kkeithley unfortunately we don't have good devel docs. I don't know off hand of any docs or the code for this. Other than getting up early on a day other than Friday and asking here, when the people in India are online
17:15 arao joined #gluster-dev
17:15 cholcombe ok
17:15 cholcombe yeah i figured :-/
17:16 cholcombe kkeithley: well i sent an email to the list asking for additional more explicit rpc calls :)
17:17 cholcombe http://www.gluster.org/pipermail/g​luster-devel/2015-July/046304.html
17:24 Gaurav__ joined #gluster-dev
17:25 aravindavk joined #gluster-dev
17:30 kkeithley hey, you even got a reply already
17:30 kkeithley a nice reply too
17:40 csim so, just to be 100% sure
17:40 csim just running "glusterd" will not start anything ?
17:41 csim ie, I need to give a path ?
17:41 csim or there is some case where it will not return immediatly to the shell ?
17:43 cholcombe kkeithley: yeah ! :)  There's an xdr library for Go so using the existing sun rpc shouldn't be a problem
17:46 csim (ok, found the issue, there is no configuration at the place it is looking for)
19:02 jobewan joined #gluster-dev
20:11 arao joined #gluster-dev
20:37 ndevos cholcombe: most of the rpc stuff should be parsed by wireshark, any reason you're not using that?
20:44 cholcombe ndevos, because it uses unix sockets to communicate with the local glusterd daemon
20:44 cholcombe how do you point the cli at another server?
20:45 cholcombe ndevos: if i could switch the cli to tcp that'd save me a lot of effort!
20:45 jbautista- joined #gluster-dev
20:46 badone joined #gluster-dev
20:51 jbautista- joined #gluster-dev
20:51 _Bryan_ joined #gluster-dev
20:55 ndevos cholcombe: you can use "gluster --remote-host=$OTHER_HOST ..."
20:56 cholcombe ndevos: you da man :D
20:56 ndevos cholcombe: I mostly capture tcpdumps with something like: tcpdump -s0 -iany -w/var/tmp/some-action.pcap tcp
20:57 ndevos oh, add " and not port 22"
20:57 cholcombe right
20:57 cholcombe don't want your ssh crap mucking things up
20:58 ndevos you could just capture port 24007 too, thats glusterd, I guess you're not interested in the rest
20:59 cholcombe yeah ideally i'd like just the interaction of one command at a time
21:00 cholcombe that makes it easy to tease out the dict parameters
21:02 ndevos cholcombe: I'm not sure how complete the cli <-> glusterd support in wireshark is, in case something is missing, send me the .pcap :)
21:03 cholcombe haha will do
22:17 kotreshhr joined #gluster-dev
22:20 kotreshhr left #gluster-dev
22:20 kotreshhr joined #gluster-dev
22:57 kotreshhr left #gluster-dev

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