fedora-qa-20090211

--- Log opened Wed Feb 11 11:03:46 2009
pingoulet's see if it works this time11:03
pingouBeer time !!11:04
wwoodsjlaska, viking_ice, f13, adamw, leitz: ping11:04
wwoodspingou: ha11:04
wwoodsFREE BEER11:04
bristothey !11:04
pingouit works :)11:04
wwoodsnow that I have your attention.. who's here for QA?11:04
wwoodsbristot: hiya11:04
-!- drago01 [n=linux@chello062178124130.3.13.univie.teleweb.at] has quit [Read error: 110 (Connection timed out)]11:04
* wwoods has a bit of a cold 11:05
bristotwwoods, i'm here to QA meeting.11:06
wwoodsno real agenda for today, just status updates from me11:06
wwoodsbristot: well, welcome!11:06
wwoodsbristot: and thanks for helping with the Fedora 11 alpha testing11:06
wwoodsthanks to anyone who helped on that11:06
-!- kirkz [n=kirk202@67-41-232-111.slkc.qwest.net] has joined #fedora-meeting11:07
bristotwonderer_, thanks, it's funny :)11:07
bristotops, wwoods11:07
wwoodsso we've got a test day coming up tomorrow, for iSCSI11:08
-!- hanthana_ [n=hanthana@124.43.60.172] has joined #fedora-meeting11:08
-!- spoleeba [n=one@fedora/Jef] has joined #fedora-meeting11:08
-!- kirkz [n=kirk202@67-41-232-111.slkc.qwest.net] has quit [Client Quit]11:08
wwoodshttps://www.redhat.com/archives/fedora-test-list/2009-February/msg00401.html11:08
wwoodshttps://fedoraproject.org/wiki/QA/Test_Days/2009-02-1211:08
-!- kital [n=Joerg_Si@fedora/kital] has joined #fedora-meeting11:09
wwoodsI'm really inexperienced with iSCSI so I'm interested to find out how you set it up etc.11:09
wwoodsbut yeah, if you want more info, check the wiki page11:09
rayvdis it for the anaconda iscsi components specifically?11:09
-!- kirkz [n=kirk202@67-41-232-111.slkc.qwest.net] has joined #fedora-meeting11:09
wwoodsrayvd: no, reading the test plan it looks like general iSCSI testing11:10
wwoodsI think the specific test cases listed are anaconda-based11:10
rayvdk11:11
wwoodsrayvd: hm, I'm wrong11:11
wwoodsthe testing is centered around anaconda's iSCSI support11:11
rayvdahh, gotcha.11:11
wwoodsI mean, there's some general info on setting up iSCSI stuff11:12
wwoodsbut the testing is focused on making sure anaconda works11:12
wwoodswe do a lot of anaconda-specific testing because it's a weird, specialized environment11:12
wwoodsso we have to treat it specially 11:12
-!- Zool^ [n=kaland@213.225.96.164] has quit [Remote closed the connection]11:14
wwoodsso, yes11:14
-!- wwoods changed the topic of #fedora-meeting to: Fedora QA Meeting | autoqa11:14
wwoodsanother thing we're working on is autoqa - automated test scripts that run when certain events happen11:15
wwoodsright now the one we're wiring up is post-repo-update - tests that run when a repo updates11:15
wwoodsso we can check for things like repoclosure11:15
wwoodswe have a git repo here: http://git.fedorahosted.org/git/?p=autoqa.git;a=summary11:16
-!- nicubunu [n=nicubunu@mail.apsro.com] has joined #fedora-meeting11:17
wwoodswe're also using the rhtslib (soon to be renamed beaker-lib) stuff from here: 11:17
wwoodshttps://fedorahosted.org/beaker11:17
-!- nicubunu [n=nicubunu@mail.apsro.com] has quit [Client Quit]11:17
wwoodswe'll make a more public announcement once we get some of the details worked out11:18
wwoodsbut, basically, we'll have a way to run whatever tests you can think of, written in python/perl/bash/whatever11:18
f13sorry I'm late, but I'm here now.11:18
wwoods(subject to review, etc)11:18
-!- kital [n=Joerg_Si@fedora/kital] has quit [Remote closed the connection]11:19
wwoodswhenever a new package is built, or repos get updated, or new release candidates are built, or whatever.11:19
wwoodsanyway, it's still a very young project but we're moving along nicely.11:20
wwoodsany questions heree?11:20
wwoodsokay. another thing I want to talk about briefly is nitrate11:20
-!- wwoods changed the topic of #fedora-meeting to: Fedora QA Meeting | nitrate11:20
wwoodssome of you might remember the Testopia website we had a while back11:21
wwoodsnitrate is a replacement for that.11:21
wwoodsif I understand it right, it's a web service to track test cases, test plans, and keep track of who's running what tests (and when and how)11:21
wwoodsso in the near future - if we're lucky, when we get to F11Beta - 11:22
wwoodswe'll have a nice web-ui that you can log into and click buttons on to report test results 11:22
wwoodsrather than having to edit a wiki page.11:22
-!- warren [n=warren@redhat/wombat/warren] has joined #fedora-meeting11:22
wwoodsmostly right now all that's in the repo is a bunch of design mockups from whiteboard sessions with UI designers and stuff11:23
wwoodsbut they're trying to get the code freed as quick as they can.11:24
-!- hanthana [n=hanthana@124.43.61.203] has quit [Read error: 113 (No route to host)]11:24
wwoodsany questions about nitrate?11:24
-!- Parafeed [n=clive@abbey1.objectsoft-systems.ltd.uk] has quit [Read error: 104 (Connection reset by peer)]11:24
f13want now11:25
wwoodsindeed11:25
* jlaska struggling w/ wireless11:26
-!- wwoods changed the topic of #fedora-meeting to: Fedora QA Meeting | etc.11:26
wwoodsanything else we should discuss?11:26
wwoodsas I said, I've got a bit of a cold so my ability to keep focused is.. off11:27
-!- mcepl [n=mcepl@49-117-207-85.strcechy.adsl-llu.static.bluetone.cz] has left #fedora-meeting []11:27
wwoodsso I'm trying to keep this short11:27
wwoodsif there's nothing else that we need to discuss for the benefit of the group, let's call it done11:28
f13should we use this meeting, or some other time to discuss the tests and watchers?11:28
wwoodsoh, we can certainly talk about that11:28
f13I've got the repo watcher which would let us run post-repo-update test for all of our updates repos and rawhide11:28
wwoodsthat conversation should also hit the lists, though11:28
f13also hopefully mmcgrath is around today to allocate a box for us11:29
wwoodsnice11:29
wwoodsI was thinking about what info we need to be able to pass from watchers to tests11:29
wwoodsprobably we just want a bunch of key=value pairs? or do we need to be able to handle lists and such too?11:30
-!- jcheung [n=root@221.223.122.50] has quit [Remote closed the connection]11:30
f13dunno, I think that squarely lives in the test development arena11:31
f13and I'm trying to stay out of that so that I can get my other deliverables done11:31
-!- Parafeed [n=clive@abbey1.objectsoft-systems.ltd.uk] has joined #fedora-meeting11:31
wwoodswell, yeah, but you're writing the watcher11:31
wwoodsso you're gonna need to pass some info to the tests. so I should probably tell you how that works11:31
f13exactly (:11:32
* viking_ice joins late in the game11:32
wwoodsso: key=val key=val ... sufficient for most things?11:33
-!- hanthana_ is now known as hanthana11:33
* viking_ice plays catchup 11:33
f13hard to say, and each bucket may have it's own syntax11:33
wwoodswell yeah11:33
f13I don't think we should worry too much about it now, until we actually have more than one bucket active11:33
f13and some tests being written.11:33
wwoodsbut we've got one generic harness, so the harness needs to deal with whatever the watcher gives it, and pass that along to the tests11:34
f13everything is pretty fluid right now and can be made to adapt.  The most important thing in my mind is that we just get some tests going as soon as possible.11:34
wwoodsfair 'nuff11:34
f13wwoods: that's pretty easily handled with argv[1:]11:34
wwoodsI just want to make sure we avoid locking ourselves into shortsighted design11:34
f13or argv[2:]11:34
f13argv0 is autoqa, argv1 is the bucket, argv[2:] is what gets sent along to the bucket11:34
wwoodsright, but ordered parameters.. that's just asking for trouble11:35
f13not really11:35
f13take away the master harnesses need/desire to parse the cli beyond the first argument11:35
wwoodsanyway you can put "key=value" into argv[n] and we can split it up and do whatever11:35
wwoodsso it's probably not a problem11:35
f13we shouldn't /have/ to split it, as you just pass everything else along to the bucket11:35
f13let the bucket parse the arguments11:35
wwoodsOK, so each bucket's test.sh can have a parse_args() function that does whatever parsing may be necessary11:37
viking_icewwoods: Regarding nitrate will the test cases be fetched ( or info ) from the wiki last time I checked nitrade was just a photographs of whiteboard with some ideas ( and the writing on the whiteboard was as encrypted as mine )11:37
f13anyway, I'll wrangle up a host to start some test deployment.11:37
wwoodsalso a choose_tests() function to filter the test list depending on those args11:37
f13wwoods: sounds reasonable.  Does it have to be .sh?  'cause shell sucks mmm-key?11:38
wwoodsviking_ice: there's a partially-complete version inside red hat - they're trying to get it released11:38
wwoodsf13: no, the point of having the harness is to allow the tests to be whatever language you like11:38
f13ok so "test.sh" was just an example11:38
wwoodsso long as you emit logs/errors to stdout/stderr and return 0 for pass and >= 1 for fail11:38
f13right-o11:39
wwoodsyeah I might put some other stuff in the selftest dir just to prove the point11:39
wwoodshello.py, hello.pl, hello.c11:39
viking_icewwoods: curios if we need to rewrite the test cases or the testing again for nitrate or the wiki pages can somehow be used..11:39
wwoodsviking_ice: probably we can do some API glue to pull tests from the wiki and push them into nitrate11:39
viking_icegreat11:40
wwoodsbut I haven't examined the APIs for creating new test cases in nitrate11:40
f13and that may only need to be a one-time affair11:40
wwoodsright, I'm assuming that nitrate will use the same auth stuff (FAS) as the wiki11:41
wwoodsso there's no real reason anyone who can edit the wiki couldn't just put stuff into nitrate instead11:42
wwoodsbut we'll see11:42
f13nod11:42
f13gotta have a bucket before you can talk about how to get people access to the bucket11:42
wwoodsyup.11:43
wwoodsonce we have an example test working11:43
wwoodsand I've shown it to the RHTS guys to make sure it's not completely stupid11:43
wwoodswe should bring this to f-d-l/f-t-l 11:44
wwoodsand see if people have any good ideas for hooks/tests/etc11:44
viking_icewwoods: so when nitrate is online we could move pages from the wiki to nitrate and just continue work there11:44
f13wwoods: we've got a nice pile to start with.11:45
f13wwoods: I'd like to wait until we have some actual hooks being ran and test results being delivered before we go shopping for more11:45
viking_icesounds like a good idea11:45
wwoodsfair 'nuff11:46
wwoodstrying to walk the line between transparency and bikeshed-painting11:47
-!- Gaaruto [n=Gaaruto@fedora/Gaaruto] has joined #fedora-meeting11:47
-!- neverho0d [n=psv@vpn-pool-78-139-211-191.tomtel.ru] has quit [Read error: 60 (Operation timed out)]11:47
f13we did have the fudcon meeting, and I posted the videos to the test-list11:47
f13and we have the wiki page where as things get checked in we'll link to them11:48
wwoodsyeah but we had to have messed up somewhere11:48
wwoodsyou didn't know the git repo existed11:48
f13I just don't like asking people to dream up things for an unknown entity11:48
wwoodsheh11:48
-!- Nirmal [n=npathak@nat/redhat-in/x-424fb0995fefb295] has joined #fedora-meeting11:48
f13wwoods: I did, but forgot.  The git repo was created during fudcon, and well things slipped my mind11:48
wwoodsalright then11:49
wwoodsonce we've got some actual factual working code11:49
wwoodswe'll bring it to the lists11:49
wwoodsuntil then, JFDI11:49
f13worksforme11:49
wwoods'kay11:51
wwoodsanything else?11:51
* wwoods takes that as a no11:53
f13beta, steamrolling ahead11:53
wwoodsah yes. how's our rawhide?11:53
f13it installed yesterday11:55
f13both nfs, and http11:55
f13also I think I found a way to shave another 19M~ off the x86(_64) CD1/DVD11:56
wwoodsnice. Yeah I've got a http x86_64 install going11:56
nirikhas anyone tried any live media compose/testing?11:58
wwoodsI promised a few folks I'd try to set up a "does rawhide work" website11:58
bristotI'm trying install in a KVM vm, but the installation be freezing ..11:58
wwoodsso this is my reminder to myself to do that.11:58
wwoodsbristot: jlaska has been reporting problems with KVM installs as well11:58
bristotwwoods, i'll find about..11:59
* nirik notes he's hitting the ext4 bug with normal livecd compose, switching to ext3 last night got a image composed, but it blows up on boot with a 'libcap.so.2' error. ;( 11:59
f13nirik: I haven't seen the blow up, but I've seen the ext4 lockup11:59
f13nirik: sandeen is working on the ext4 lockup (but not with much luck :/)11:59
f13nirik: are you composing on an ext3 system?12:00
nirikyeah, it pretty much happens everytime for me. I think it might be related to multi cpu systems.12:00
nirikyes.12:00
nirikdoes it not happen on a ext4 system?12:00
f13not sure12:00
f13I haven't tried that at the place I was getting lockups12:00
nirikanyhow, I can keep poking at it... I have changes I want to make, but I kinda want to test them before blindly changing the ks file. ;)12:01
-!- J5 [n=quintice@nat/redhat/x-9acffde9e2677f56] has joined #fedora-meeting12:01
-!- MEP [n=MEP@83.225.238.112] has left #fedora-meeting []12:01
f13heh12:01
wwoodsalright then12:02
wwoodsanything else?12:02
jlaskaiSCSI test day tomorrow12:03
jlaskawe've got a good test plan and feature page in place ... so please join to help shake some fun out of iSCSI (https://www.redhat.com/archives/fedora-test-list/2009-February/msg00401.html)12:04
-!- Zool^ [n=kaland@19.81-166-29.customer.lyse.net] has joined #Fedora-meeting12:04
jlaskabristot: I've got a few KVM issues I've been tracking (with lots of help from markmc) ... we can catch up after the meeting for the issues you are experiencing12:05
wwoodsalright then, I think we're good12:05
-!- nphilipp [n=nils@nat/redhat/x-7b0680a0475af395] has quit ["Leaving"]12:06
wwoodsthanks all - we'll be in #fedora-qa (myself after a bit of lunch) for any further discussion12:06
-!- valente [n=vali@89.121.200.106] has quit ["Leaving."]12:06
--- Log closed Wed Feb 11 12:06:25 2009

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!