Camelia, the Perl 6 bug

IRC log for #gluster-dev, 2013-04-05

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

All times shown according to UTC.

Time Nick Message
00:02 kkeithley1 uh, yeah, that's where you saw it. gotcha
01:17 bala1 joined #gluster-dev
01:41 bala1 joined #gluster-dev
02:03 portante joined #gluster-dev
02:07 yinyin joined #gluster-dev
03:26 yinyin joined #gluster-dev
03:49 sgowda joined #gluster-dev
04:02 lalatenduM joined #gluster-dev
04:03 lala_ joined #gluster-dev
04:31 aravindavk joined #gluster-dev
04:31 mohankumar joined #gluster-dev
04:32 sripathi joined #gluster-dev
04:38 deepakcs joined #gluster-dev
04:41 yinyin joined #gluster-dev
04:55 bala joined #gluster-dev
04:56 hagarth joined #gluster-dev
05:16 sripathi1 joined #gluster-dev
05:22 bala joined #gluster-dev
05:40 raghu joined #gluster-dev
05:43 bala joined #gluster-dev
06:01 vshankar joined #gluster-dev
06:17 sripathi joined #gluster-dev
06:36 rastar joined #gluster-dev
07:10 bala joined #gluster-dev
07:30 bala joined #gluster-dev
08:37 puebele1 joined #gluster-dev
08:55 puebele1 joined #gluster-dev
09:26 sripathi joined #gluster-dev
10:15 hagarth joined #gluster-dev
11:19 hagarth joined #gluster-dev
11:20 aravindavk joined #gluster-dev
11:28 puebele2 joined #gluster-dev
11:30 puebele3 joined #gluster-dev
11:49 puebele1 joined #gluster-dev
12:18 aravindavk joined #gluster-dev
13:28 mohankumar joined #gluster-dev
13:41 ollivera joined #gluster-dev
14:06 awheeler_ portante: ping
14:08 awheeler_ kkeithley: ping
14:08 kkeithley hi
14:09 awheeler_ Hey Kaleb.  looking at 4781, the test fails as files are missing from the data directory that would be built by gluster-swift-gen-builders.
14:09 awheeler_ So, should those pre-generated files be added to the repo, or generated by the test, or ?
14:10 awheeler_ Here's the Jenkins console showing the failures: http://build.gluster.org/job/ufounit/1141/console
14:11 kkeithley portante is better suited to answer that
14:12 awheeler_ Makes sense.  When I create the files, the test works.  But, seems like I shouldn't have to do that to make the test work.  I noted you +1 the code review, so figured things looked different for you.
14:13 wushudoin joined #gluster-dev
14:15 kkeithley maybe a good catch on your part.
14:16 kkeithley Let's see what portante says
14:16 awheeler_ I assume he had a plan, as this was solved previously by creating a fake ring, which he replaced.
14:16 awheeler_ Sounds good.  Is te portante|afk?
14:17 kkeithley he's in the building
14:17 kkeithley maybe hasn't made it to his desk yet
14:26 lpabon joined #gluster-dev
14:35 lpabon joined #gluster-dev
14:50 bala joined #gluster-dev
15:38 kkeithley grrr, another regression test is stalled
15:52 hagarth joined #gluster-dev
16:00 copec left #gluster-dev
16:10 bala joined #gluster-dev
16:26 awheeler_ portante: Ping
16:26 jbrooks joined #gluster-dev
16:27 hagarth joined #gluster-dev
16:29 gbrand_ joined #gluster-dev
16:42 rastar joined #gluster-dev
16:56 lpabon joined #gluster-dev
17:04 portante awheeler_: here
17:04 awheeler_ Just reviewed your code change -- the ring.gz files are missing, and need to be manually generated for the test to pass.
17:05 portante missing? hmm, I think because the .gitignore hides them by default, I'll add them in my tree and resubmit
17:06 awheeler_ Ah, that makes sense.  Why did you decide not to use the fake ring -- get more of the swift ring code to be tested?
17:08 hagarth joined #gluster-dev
17:11 portante because the Ring() class initialization reads /etc/swift by default, so since it is reading a directory anyways, just seemed simpler to create the ring files the way we want and use em
17:11 jbrooks joined #gluster-dev
17:11 awheeler_ Ah, got it.
17:12 portante they unit tests fail if you run them on a system that does not have a /etc/swift prior to this change
17:12 portante have you tried running the unit tests with tox?
17:12 awheeler_ Oh, ok, makes sense.
17:13 awheeler_ yes, and it failed just like the build, until I ran the ring builder script in the data directory, then they ran fine -- after I installed python2.7.
17:13 portante it should run with 2.6 or 2.7
17:14 awheeler_ It is setup to test both 2.6 and 2.7, and w/o 2.7 it complains about its lack.
17:14 portante I think we should probably just test against 2.7, but not sure the rest of glusterfs is ready for just going to 2.7
17:14 awheeler_ Well, RHEL6 and CentOS 6 are standardized on 2.6, so seems like that has to be part of the testing.
17:15 awheeler_ There are no official 2.7 RPMs for RHEL 6/CentOS 6
17:15 portante okay, just updated the change to use grizzly 1.8.0, and fixed the missing .ring.gz files
17:20 kkeithley as I understand it, tox will use 2.6 on RHEL/CentOS and 2.7 on things that have 2.7
17:21 kkeithley so far Swift seems to be agnostic for the most part on python 2.6 or 2.7.
17:21 kkeithley Hopefully things stay that way
17:23 portante tox will use what is configured in tox.ini
17:23 portante right now it attempts to find a 2.6 and 2.7 compiler to run.
17:24 kkeithley oh, you told me that once already. :-/
17:25 awheeler_ 2.6 passed.  What's with the red flag on tests that take longer than x to run?
17:25 portante it is a configuration option in tox.ini for what color to code test names that complete in a certan time
17:26 awheeler_ So, for now it's arbitrary?  I'm getting red on some of the tests.
17:26 portante Yes, it just means some test took "too long"
17:27 awheeler_ That seems to have fixed it.  Review updated.
17:27 portante for example, I am modifying the tox.ini file to try out py33 and pypy
17:28 portante we might want to also adopt pep8, and then we can use the pep8 tool that way
17:29 portante looks like we cannot use python three because eventlet is not there yet
17:30 hagarth joined #gluster-dev
19:30 puebele1 joined #gluster-dev
20:38 jbrooks joined #gluster-dev
22:32 jbrooks joined #gluster-dev
22:53 jbrooks_ joined #gluster-dev

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