Wednesday, 2014-03-19

« Tuesday, 2014-03-18 Index Thursday, 2014-03-20 »
*** matsuhashi has joined #openstack-trove00:20
*** ViswaV has quit IRC00:32
*** nosnos has joined #openstack-trove00:51
*** nosnos_ has joined #openstack-trove00:53
*** nosnos has quit IRC00:57
*** tanisdl has quit IRC01:01
*** eguz_ has quit IRC01:15
openstackgerritDan Nguyen proposed a change to openstack/trove: Added route for Admin API to support guest upgrade  https://review.openstack.org/8141001:20
*** esp has joined #openstack-trove01:20
*** khyati_ has quit IRC01:21
*** mattgriffin has quit IRC01:34
*** mattgriffin has joined #openstack-trove01:36
*** wrongfan has joined #openstack-trove01:43
*** datsun180b has quit IRC01:51
wrongfanhi,all.Is there anyone working on the Scheduled Task Service?01:59
*** ViswaV has joined #openstack-trove02:01
wrongfancore developers in rackspace or mirantis or others?02:02
*** ViswaV_ has joined #openstack-trove02:03
*** ViswaV has quit IRC02:06
*** amytron has joined #openstack-trove02:09
*** khyati_ has joined #openstack-trove02:13
*** amytron has quit IRC02:15
*** grapex has joined #openstack-trove02:22
*** michael-yu has joined #openstack-trove02:25
*** michael-yu has quit IRC02:32
*** amcrn has quit IRC02:49
*** eghobo has joined #openstack-trove03:00
*** matsuhashi has quit IRC03:13
*** matsuhashi has joined #openstack-trove03:20
*** harlowja is now known as harlowja_away03:23
*** matsuhashi has quit IRC03:25
*** grapex has quit IRC03:31
*** global_ has quit IRC03:36
*** nosnos_ has quit IRC03:41
*** michael-yu has joined #openstack-trove04:05
*** michael-yu has quit IRC04:11
openstackgerrityangxurong proposed a change to openstack/trove: Use built-in print() instead of print statement  https://review.openstack.org/8046804:12
openstackgerrityangxurong proposed a change to openstack/trove: Use built-in print() instead of print statement  https://review.openstack.org/8046804:13
openstackgerrityangxurong proposed a change to openstack/trove: Use built-in print() instead of print statement  https://review.openstack.org/8046804:14
*** matsuhashi has joined #openstack-trove04:23
*** nosnos has joined #openstack-trove04:35
*** wrongfan has quit IRC04:47
*** amcrn has joined #openstack-trove04:53
*** wrongfan has joined #openstack-trove05:01
*** mattgriffin has quit IRC05:05
cp16netSlickNik: you there?05:22
*** shivam has joined #openstack-trove05:24
*** jesusaurus has quit IRC05:26
SlickNikcp16net: here now, sup?05:29
cp16netso i saw something you approved that worrys me a little05:36
cp16nethttps://review.openstack.org/#/c/80010/05:36
cp16netwhat do you think about my comment?05:36
*** amytron has joined #openstack-trove05:36
amcrncp16net: the original patch-set actually had int-tests (see up until iteration #4)05:43
SlickNikcp16net: Looking at the patch again. +1 on the comment on the integration test.05:44
amcrnSlickNik: getting int-tests is going to be tricky, because it'll require a trove-integration merge first (and adding a non-mysql-5.5 version will likely break a bunch of other hardcoded tests)05:44
amcrnthen merge trove05:45
cp16netamcrn: i see that point but isnt redis already there without backups currently?05:45
cp16netgranted i know that wont be forever... but its true right now05:46
amcrnthere's been no official release with redis, has there?05:46
amcrn(i.e., it wasn't there by havana label cut, was it?)05:46
SlickNikcp16net: afaik we're not building a redis image for the int-tests.05:46
cp16nethavana no...05:47
cp16netoh...05:47
cp16nethmmm05:47
cp16neti see the delima05:48
cp16netwell i think its something that we should definatly add as a test case05:48
amcrndefinitely agreed int-tests need to be added, it's just a question of whether this patch gets backed out until that's true05:49
SlickNikcp16net: I fully agree. I think we're currently in a situation where we're not really testing other datastores as part of the int-tests.05:50
cp16netyeah thats true05:50
SlickNikThis is something that we can hopefully rectify soon after icehouse.05:50
cp16netyea05:50
cp16netspeaking of tests...05:50
cp16neti love the fact that there is some heat related tests05:51
cp16netbut i am really scared of how they are implemented05:51
cp16netchanging configs and restarting the service in the tests05:51
* amcrn chuckles05:51
cp16neti wish we could run like another suite in parallel for heat05:52
SlickNikcp16net: I think that idea is actually worth exploring.05:55
cp16netoh... and no wonder why the the gate passes after adding them... it doesnt run them!05:55
SlickNikcp16net: Even if it means that the heat suite runs only a subset of the tests that the integration gate runs (for example, it wouldn't re-run the API/flavor tests over again).05:56
shivam_Hi SlickNIk05:57
SlickNikhi shivam_05:58
shivam_can you please review this05:58
shivam_https://review.openstack.org/#/c/66499/05:58
cp16neti mean if it _was_ another box running the heat suite of tests then it could run everything05:58
*** saju_m has joined #openstack-trove05:58
cp16netshivam_: i just reviewed it05:59
shivam_this is being on delay for a long time05:59
*** matsuhashi has quit IRC05:59
cp16neti was just talking about it actaully05:59
shivam_ok thanks yes just saw it06:00
*** matsuhashi has joined #openstack-trove06:01
cp16netyeah i just noticed that it wasnt in the output of the rdjenkins output06:01
cp16netalright i am out06:01
*** shalini has joined #openstack-trove06:01
cp16neti think its a little later here..06:02
cp16netlate*06:02
amcrnpeace out cp16net06:02
SlickNikamcrn: Once juno is open, let's chat about how we can add these non-default datastores to the integration gate, and start testing some of the scenarios that cp16net mentioned.06:02
SlickNikcp16net: later.06:02
amcrnSlickNik: agreed, we need to find a way to get multiple versions and types doing some form of basic checks06:02
cp16netSlickNik: yeah that would be grrrrr-ate! ;-)06:02
cp16netBender: My life, and by extension everyone else's is meaningless.06:03
SlickNikheh06:03
shaliniHi SlickNik , cp16net06:08
SlickNikhi shalini06:09
shaliniCan we have a discussion on my patch https://review.openstack.org/#/c/66499/06:09
shaliniwhat are we planing to add heat int-tests06:09
shalinias you guys are already discussing the same06:09
SlickNikshalini: looks like your patch needs rebasing.06:11
shaliniOh..!!06:11
shaliniI'll do it06:11
openstackgerritA change was merged to openstack/trove: Added Backup/Restore validations  https://review.openstack.org/8001006:11
SlickNikshalini: Also cp16net mentioned that although new tests have been added, none of these are actually being added to the main integration gate, so they're not running as part of the int-tests.06:11
shaliniBut as per comments like we need to add tests in gate06:11
shaliniwhat should i do06:12
SlickNikTo add tests to the int-tests gate, they need to be added to the master group as well (cp16net left a comment on the patch).06:14
SlickNikWe were also discussing the possibility of adding the heat tests as a separate gate job.06:14
shaliniI did not find any such group in int-test.py . can u specify what master group mean?06:16
shaliniOk. So to make tests run we'll be having reddwarf kind of thing which execute whole int-tests for heat instance also?06:17
shaliniOk06:19
shaliniDo he mean to add heat tests also in existing groups of black-box?06:19
shaliniBut i think that will disturb the infra..06:20
*** achampion has quit IRC06:32
*** matsuhashi has quit IRC06:47
*** ViswaV_ has quit IRC06:49
*** matsuhashi has joined #openstack-trove06:50
*** amytron has quit IRC06:55
*** khyati_ has quit IRC07:02
*** flaper87|afk is now known as flaper8707:14
*** eghobo has quit IRC07:26
*** matsuhashi has quit IRC07:30
*** matsuhashi has joined #openstack-trove07:31
*** shivam has quit IRC07:55
*** tanisdl has joined #openstack-trove08:24
*** amcrn has quit IRC08:42
*** iartarisi has joined #openstack-trove08:45
*** cweidenkeller has quit IRC09:01
*** dukhlov has quit IRC09:33
*** dukhlov has joined #openstack-trove09:33
*** SushilKM has joined #openstack-trove09:34
*** dukhlov has quit IRC09:35
*** SnowDust has joined #openstack-trove09:43
*** dukhlov has joined #openstack-trove09:46
*** SushilKM has quit IRC10:24
*** key4 has joined #openstack-trove10:26
*** key4 is now known as setho210:26
*** setho2 has left #openstack-trove10:26
*** setho has joined #openstack-trove10:26
*** SnowDust has quit IRC10:33
*** tattabbum has joined #openstack-trove10:37
*** tanisdl has quit IRC10:53
*** flaper87 is now known as flaper87|afk11:17
amrithso ... I know I'm beginning to sound like a stuck gramaphone player (oh crap, some people don't know what that is) but I'm thinking I must have dozed off at the key signing party because no one (other than juice and abramley) have signed my key. I thought there were others at the party?11:18
amrithhub_cap: have I perchance missed a step or something? kevinconway, did you sign my key? if yes it didn't get pushed ...11:19
*** matsuhashi has quit IRC11:22
*** matsuhashi has joined #openstack-trove11:24
*** demorris has joined #openstack-trove11:25
*** matsuhashi has quit IRC11:37
*** achampion has joined #openstack-trove11:41
*** demorris has quit IRC11:59
*** tattabbum has quit IRC12:03
*** flaper87|afk is now known as flaper8712:08
*** pdmars has joined #openstack-trove12:12
*** achampion has quit IRC12:18
*** grapex has joined #openstack-trove12:39
*** casanch1 has joined #openstack-trove12:39
*** grapex has quit IRC12:43
*** robertmyers has quit IRC12:45
*** grapex has joined #openstack-trove12:48
*** radez_g0n3 is now known as radez12:53
*** grapex has quit IRC13:02
*** grapex has joined #openstack-trove13:02
*** grapex has quit IRC13:03
*** grapex has joined #openstack-trove13:03
*** wrongfan has quit IRC13:14
*** rhodgin has joined #openstack-trove13:21
*** robertmyers has joined #openstack-trove13:23
*** nosnos has quit IRC13:24
*** robertmy_ has joined #openstack-trove13:25
*** robertmyers has quit IRC13:25
*** mattgriffin has joined #openstack-trove13:31
*** achampion has joined #openstack-trove13:48
openstackgerritA change was merged to openstack/trove: Fixes insecure update of /etc/fstab file  https://review.openstack.org/6912513:53
*** casanch1_ has joined #openstack-trove14:01
*** casanch1 has quit IRC14:02
*** casanch1_ is now known as casanch114:03
*** demorris has joined #openstack-trove14:05
*** rramirez has joined #openstack-trove14:05
*** tattabbum has joined #openstack-trove14:07
openstackgerritAndreas Jaeger proposed a change to openstack/trove: Setup trove for translation  https://review.openstack.org/8153214:07
*** jcru has joined #openstack-trove14:08
*** kevinconway has joined #openstack-trove14:15
*** shivamshukla has joined #openstack-trove14:17
*** shivamshukla has quit IRC14:17
*** freyes has joined #openstack-trove14:20
*** freyes has quit IRC14:25
*** shivamshukla has joined #openstack-trove14:25
*** amytron has joined #openstack-trove14:29
*** rramirez has quit IRC14:36
*** datsun180b has joined #openstack-trove14:38
*** thedodd has joined #openstack-trove14:49
*** jmontemayor has joined #openstack-trove14:53
openstackgerritshivam shukla proposed a change to openstack/trove: Tests for heat based instance workflow  https://review.openstack.org/6649914:59
*** jmontemayor has quit IRC15:02
*** shalini_ has joined #openstack-trove15:02
*** jmontemayor has joined #openstack-trove15:05
*** flaper87 is now known as flaper87|afk15:11
*** ViswaV has joined #openstack-trove15:15
*** ViswaV_ has joined #openstack-trove15:16
*** flaper87|afk is now known as flaper8715:18
*** shivamshukla has quit IRC15:18
*** ViswaV has quit IRC15:19
hub_capamrith: i need to generate a new key15:20
hub_capmy chromebook kjust got torqued and i lost the hdd15:20
hub_capcomplments of chromeos's on screen "recovery" feature my wife accidentally enabled15:21
hub_capso dont worry about my key heh15:21
* hub_cap is so sad15:21
*** demorris has quit IRC15:25
* hub_cap lost a day to setting up his laptop15:28
openstackgerritDan Nguyen proposed a change to openstack/trove: Added separate rate limit setting for mgmt POST  https://review.openstack.org/8155715:32
*** ViswaV_ has quit IRC15:35
*** ViswaV has joined #openstack-trove15:35
*** shivam has joined #openstack-trove15:36
grapexAnyone runnng OSX seeing this problem when trying to run tox? https://gist.github.com/TimSimpsonR/964499216:03
*** eghobo has joined #openstack-trove16:06
hub_capgrapex: that looks like its runnig your system python setup.py16:06
hub_capand my guess is you have an older pip in system python16:07
grapexOh, duh16:07
grapexI thought virtualenv would magically update pip, but I guess it uses pip to update other things16:07
grapexthanks16:07
hub_capnpnp16:08
*** eghobo has quit IRC16:08
grapexI guess virtualenv could use pip to update itself, but that would be a bit presumptuous of it.16:09
*** flaper87 is now known as flaper87|afk16:11
*** eghobo has joined #openstack-trove16:11
vgnbkrSpeaking of pip updating itself, does anyone know where that happens in the kick-start image?  I couldn't find it, and had to do it manually since python has pip1.0, but we needed pip>=1.4.16:12
*** khyati has joined #openstack-trove16:13
vgnbkrAlso, are there upstart scripts available for the trove services?  If so, where would I find them?  If not, where should they be?16:14
shalini_Hi hub_cap and all16:14
vgnbkrSlickNik: re trove state.  No, I didn't use the reset API, I realized that my problem was that the guest-agent hadn't contacted the taskmanager due to networking issues.  Once that was fixed, it updated it's state properly and I could move forward.16:16
SlickNikvgnbkr: glad you got that figured out.16:18
*** jmontemayor has quit IRC16:18
SlickNikvgnbkr: https://github.com/openstack/trove-integration/blob/master/scripts/files/trove-guest.upstart.conf16:18
vgnbkrSlickNik: thanks, I have the upstart for the guest-agent, I was asking about upstart scripts for the trove services (api, conductor, taskmanager).16:19
hub_caphi shalini_16:23
vgnbkrhub_cap: Are you aware of any upstart scripts for starting the Trove services?16:24
hub_capvgnbkr: we have one for the guest, but thats it16:25
shalini_Please review https://review.openstack.org/#/c/66499/16:25
shalini_It is being delayed for so long.16:25
vgnbkrhub_cap: I assume that for the icehouse release we will need to have a package pushed to the apt repos, and as part of that there will need to be upstart scripts.  Is that correct?16:26
SlickNikvgnbkr: I don't think we have specific ones for the services. tripleo-image-elements does have an element (os-svc-install) that given an OpenStack project generates these on the fly, iirc.16:28
SlickNikhttps://github.com/openstack/tripleo-image-elements/blob/6cdec90acb28c62fa83b64611e8e8f8b5416c46e/elements/os-svc-install/bin/os-svc-daemon16:28
shalini_vgnbkr: helo16:28
shalini_You are working on trove uses soon-to-be-deprecated first-boot scripts16:28
shalini_?16:28
vgnbkrSlickNik: Great, thanks, that's what I needed to know.  So for now I will just hand code some.16:28
hub_capvgnbkr: we dont do that work tho, but yes, it should happen16:29
shalini_@ vgnbkr Bug #1212436...16:29
shalini_Ok16:29
hub_capvgnbkr: https://packages.debian.org/search?keywords=trove16:29
vgnbkrshalini_: I did that, but I've been sidetracked getting an OpenStack/Trove environment set up so that we can move forward with installation docs.16:29
hub_capu can see the trove ones are already in debian testing16:29
*** demorris has joined #openstack-trove16:32
shalini_So what are we planing to do in that.. can still do something with first-boot.d ?16:32
*** jmontemayor has joined #openstack-trove16:34
vgnbkrNo, you can't use first-boot.d.  What are you trying to do?16:35
shalini_As trove is already using it. I was just trying to add something for my own purpose16:37
shalini_Just a R & D16:37
*** Ajaeger1 has joined #openstack-trove16:37
*** rramirez has joined #openstack-trove16:37
shalini_So what are we planing to do with first-boot.d16:37
Ajaeger1Trove team, if you like to see Trove translated, please review https://review.openstack.org/#/c/81532/ - the sooner this is setup, the quicker you get translations rolling...16:38
shalini_And when will it get eliminated completely?16:38
vgnbkrWe didn't actually need to use first-boot.d, so I just did what it did in a different way.16:40
vgnbkrMy commit wrt that will eliminate it's use in Trove.16:40
hub_capAjaeger1: if this review thats loading is the transifex stuff i will find u and hug u16:41
hub_capAjaeger1: it is!! HORRAY!!! we will merge this for icehouse16:41
esphub_cap: greetings16:42
espgrapex: I had to do this recently on my mac to install some of our python packages.  export ARCHFLAGS=-Wno-error=unused-command-line-argument-hard-error-in-future16:45
hub_capesp: i cannot see your pm's16:45
espno prb16:45
espit's probably me16:45
hub_caplet me try to fix esp16:45
hub_capwell esp i did just redo my laptop....16:46
shalini_Vgnbkr : like the way written in readme of disk-image-builder ?16:46
*** iartarisi has quit IRC16:47
shalini_So when can we see your commit in openstack gerrit? I wish to see how we are doing it .... :)16:48
hub_capesp: victory16:48
espnice work hub_cap16:49
Ajaeger1hub_cap: let's hug in Atlanta, ok?16:53
hub_capAjaeger1: aye16:53
*** SushilKM has joined #openstack-trove16:53
vgnbkrshalini_: No, what Trove's first-boot script was doing could just as well be done in pre-boot, so I did it there (in the chroot).16:55
*** amcrn has joined #openstack-trove16:55
*** SushilKM has quit IRC16:56
*** SushilKM has joined #openstack-trove16:56
*** yogesh has joined #openstack-trove16:57
juiceall: the meeting is an hour later due to day light savings drift correct?16:58
shalini_ Vgnbkr: okay..so files in pre-boot.d also gets executed after booting of instance?16:59
hub_capso is it meeting time now? w dst changes?17:00
amcrnlast time we did it an hour from now, since technically it's 17:00 utc right now (if i remember correctly)17:00
hub_capright it changes in 2 wks then i think17:01
hub_capor something like that, right?17:01
vgnbkrNo, there just wasn't anything we needed to do post boot.  (and there is no pre-boot.d, I don't think - by "pre-boot", I meant in the dib elements)17:01
vgnbkrsorry, all, ^^^ was for shalini_17:02
esphub_cap: no idea.  amcrn is correct tho.  last time we meet an hour after.17:02
shalini_SlickNick: I have added new patch. You can please review  https://review.openstack.org/#/c/66499/17:03
*** eguz has joined #openstack-trove17:04
shalini_Vgnbkr : but we should also have something to do for post-boot kind of thing17:04
hub_capshalini_: SlickNik that bp is not listed in rc1, so we wont get to it for a bit i think17:04
*** eguz has quit IRC17:04
*** eguz has joined #openstack-trove17:04
grapexestp: Thanks for the advice, but no luck on my end. :(17:07
*** eghobo has quit IRC17:07
*** SnowDust has joined #openstack-trove17:08
*** harlowja_away is now known as harlowja17:09
vgnbkrshalini_: at this point, I don't think there is anything that we need to do.  The original bug report suggests how it is supposed to be done.17:11
*** tattabbum has quit IRC17:11
vgnbkrshalini_: Do you have something that can't be done in the chroot?17:11
vgnbkrshalini_: Basically, anything that is done to the guest-agent image must survive reboot, so generally should be configured through init/upstart/systemd scripts.17:13
SnowDustis this about firstboot.d we are talking ?17:18
shalini_SnowDust yes17:19
*** SushilKM has quit IRC17:19
shalini_Vgnbkr : m not sure abt chroot but I just wish thay17:20
SnowDustshalini_ : so as per vgnbkr we will do upstart script for running services that survive reboot right ?17:20
hub_capvgnbkr: is saying thats the way systems do it, generally, and to do it another way would not be the right way (i think)17:23
shalini_Vgnbkr: we will be having something similar to first-boot in which comands get executed after instance boot.17:23
vgnbkrshalini_: we won't, because everything needs to work in a heat environment.  The original first-boot bug report tells you how to do the equivalent of "first-boot.d" if you really need it.17:25
*** harlowja has quit IRC17:26
*** michael-yu has joined #openstack-trove17:26
shalini_Okay...17:26
*** SushilKM has joined #openstack-trove17:27
*** rramirez has quit IRC17:28
shalini_hub_cap : so when are we going to take heat based instance workflow bp ..?17:29
*** rramirez has joined #openstack-trove17:30
*** rramirez_ has joined #openstack-trove17:30
*** michael-yu has quit IRC17:31
*** michael-yu has joined #openstack-trove17:31
shalini_hub_cap : talking about the one, for which you was saying that it is not listed in rc117:32
*** rramirez_ has quit IRC17:33
*** casanch1 has quit IRC17:33
*** casanch1 has joined #openstack-trove17:34
*** shalini_ has quit IRC17:34
*** shalini_ has joined #openstack-trove17:35
*** harlowja has joined #openstack-trove17:36
*** datsun180b has quit IRC17:37
Ajaeger1hub_cap: regarding your question on how this works: There's both a post job and a periodic job run by Zuul/Jenkins.17:39
Ajaeger1openstack-infra/config has modules/jenkins/files/slave_scripts/upstream_translation_update.sh17:39
Ajaeger1and calls it with "upstream_translation_update.sh trove" after each and every change17:40
*** shalini_ has quit IRC17:40
SlickNikSnowDust: got a sec?17:41
SnowDustyes !17:41
Ajaeger1and then there's propose_translation_update.sh which is called once every day (after 6:00 UTC) and that one proposes any updated translations back to you17:41
*** datsun180b has joined #openstack-trove17:41
SnowDustSlickNIk: tell me17:41
*** SushilKM has quit IRC17:41
SnowDustSlickNik: i am there17:42
*** shalini_ has joined #openstack-trove17:42
SlickNikSnowDust: noticed you added an new agenda item to the meeting.17:42
shalini_Hubcap:  ...?17:42
Ajaeger1sorry, it wasn't hub_cap asking - will add to gerrit now17:42
SlickNikSnowDust: Am a bit unclear around what it entails. Is there a bp or wiki page regarding it?17:42
SnowDustSlickNik its around earlier discussion in the community .. let me confirm from yogesh17:43
SnowDustMay be we should discuss in the open forum ..17:44
SlickNikshalini_: I think we'll need to wait for the juno branch to open for checkins before we can merge non rc1 bps / bugs.17:44
SlickNikSnowDust: I think that's a good idea, Let's discuss it on the ML / IRC channel, and then when we have a recommendation (wiki page / bp / whatever) let's bring it up in the meeting.17:45
SlickNikSnowDust: Trying not to have a free-for-all, design this right now impromptu session during the meeting. :)17:47
*** khyati has quit IRC17:47
*** cweid has joined #openstack-trove17:47
*** shalini_ has quit IRC17:49
*** zuqiang has joined #openstack-trove17:49
*** shalini_ has joined #openstack-trove17:49
yogeshSlickNik: I remember we wanted to have a discussion on this earlier...is there some movement...container vs Join?17:49
*** michael-yu has quit IRC17:50
openstackgerritA change was merged to openstack/trove: Remove IDENTIFIED BY clause for MySQL REVOKE  https://review.openstack.org/8069417:53
cp16netaway17:54
cp16netwhoops17:54
hub_caplol cp16net17:54
*** michael-yu has joined #openstack-trove17:55
cp16nethi hub_cap17:55
grapexcp16net: You were about to fly out the Window again when you realized you had a meeting, right?17:55
SlickNikyogesh: IIRC, not really. We discussed the pros and cons of the two approaches, but decided to concentrate on just replication for v1. amcrn and the ebay folks were still looking into the cluster vs. join approach, and we were going to discuss it further once we are done with icehouse and v1. You might want to sync with amcrn to see what their thoughts are so far.17:55
cp16netwell i was attempting to return17:55
cp16netbut missed the /17:55
*** ViswaV has quit IRC17:55
shalini_SlickNik : k thanks...17:56
*** ViswaV has joined #openstack-trove17:56
shalini_Cp16net : Hi wanted to discuss about your comment on https://review.openstack.org/#/c/66499/17:57
yogeshSlickNik: sure17:58
shalini_Cp16net : you17:58
cp16netsure17:58
shalini_Commented that we should add them in group to make them run in gate.17:59
cp16netcan we chat after the meeting? shalini_17:59
*** ViswaV_ has joined #openstack-trove17:59
*** ViswaV has quit IRC17:59
shalini_How much timevwill it take..o17:59
shalini_?17:59
cp16netyou can add it to the GROUP and it would be jsut the same17:59
shalini_Which GROUP ?17:59
cp16netthere is a master name called GROUP18:00
shalini_The one which is defined in instance.py as GROUP?18:00
cp16netyea18:00
shalini_I tried but when I executed ./redstack int-tests18:00
shalini_My tests did not run.18:01
*** Ajaeger1 has left #openstack-trove18:01
SlickNikmeeting in #openstack-meeting-alt18:01
*** ashishsnigam has joined #openstack-trove18:01
shalini_As that grp was not imported in int-test.py I suppose18:01
*** shalini_ has quit IRC18:02
*** shalini_ has joined #openstack-trove18:02
*** amcrn has quit IRC18:02
*** Barker has joined #openstack-trove18:05
*** rramirez_ has joined #openstack-trove18:06
*** rramirez has quit IRC18:06
mayfield.18:07
shalini_Cp16net : summarizing it. I add18:08
shalini_Added GROUP Along with GROUP_HEAT , but when I executed int-tests my heat tests did not run. So I suppose instances.GROUP is not imported in int-tests.py thats y my tests did not run.18:11
*** ViswaV_ has quit IRC18:12
*** khyati has joined #openstack-trove18:14
*** khyati has quit IRC18:16
*** khyati has joined #openstack-trove18:18
*** khyati has quit IRC18:23
*** khyati has joined #openstack-trove18:24
*** ViswaV has joined #openstack-trove18:27
*** tanisdl has joined #openstack-trove18:33
*** yogesh has quit IRC18:33
*** robertmy_ is now known as robertmyers18:43
*** ViswaV has quit IRC18:43
*** shakayumi has joined #openstack-trove18:47
*** shakayumi has quit IRC18:47
*** SnowDust has quit IRC18:53
*** saurabhs has joined #openstack-trove18:57
*** shakayumi has joined #openstack-trove18:58
*** shakayumi has quit IRC18:58
*** SnowDust has joined #openstack-trove19:00
hub_capSnowDust: fwiw, when i said offline i meant we should discuss it here after the meeting19:01
SnowDustyes19:01
SnowDusti got that19:01
kevinconwaySnowDust: really he wan't you to mail him a letter19:02
hub_capcool so since no one is talking about the container vs joins we can finish19:02
*** shakayumi has joined #openstack-trove19:02
*** shakayumi has quit IRC19:02
SnowDusthub_cap i am here to talk about container vs joins19:02
hub_capand then i can sing an awesome joe cocker song19:02
hub_capahh so what is that about then SnowDust ? did u add it to the agenda?19:02
SnowDustyes hub_cap i did that19:02
hub_capok what does it mean?19:03
SnowDustfor me its instance(db) lifecycle management we should focus before we do replication19:03
*** shakayumi has joined #openstack-trove19:04
SnowDustwhat i understood is container = instance( db) thats dbaas19:04
*** freyes has joined #openstack-trove19:05
SnowDustand joins is group of db nodes to form an instance19:06
SnowDustanyone there still ?19:07
*** ashishsnigam has quit IRC19:08
*** SnowDust has quit IRC19:09
*** SnowDust has joined #openstack-trove19:09
SnowDusthello again19:09
hub_capSnowDust: post to the mailing list, maybe one of the tesora guys will reply19:09
SnowDusthub_cap : sure19:09
hub_caplooks like everyone dissapeared :)19:10
SnowDusthub_cap : mass disconnections19:10
*** yogesh has joined #openstack-trove19:10
hub_capok im too going to disconnect, i have to finish setting up my lappy19:10
SnowDusthub_cap : is bp approved ?19:11
SnowDusti dont have the clear idea after discussion19:11
hub_capyes SnowDust it is19:11
hub_capbut we want u to remove the stuff in cfg. that u added19:11
hub_capthere is no reason to put it in there19:11
SnowDusthub_cap ,19:11
hub_capas soon as anything which needs that option gropu imports it19:11
SnowDustcfg until register an opt/group19:11
hub_capitll load into CONFIG19:11
SnowDustwe cannot refer them using CONF19:11
hub_capinto *CONF* sorry19:12
hub_capimport path.to.new.mysql.config19:12
hub_capCONF.mysql.new_value19:12
hub_capwill work19:12
hub_capmaybe u can just, in the datastore manager, import path.to.#{datastore}.config19:13
hub_capso itll be "autoloaded" when the manager is created19:13
SnowDusthub_cap ok got it ..19:13
hub_caprelink the bp and ill approve it19:13
SnowDusthttps://blueprints.launchpad.net/trove/+spec/refactoring-datastore-options-in-cfg19:14
*** yogesh has quit IRC19:14
SnowDusthub_cap : my understanding of oslo.config.cfg.import_group  made me implement that way19:15
SnowDustit was written when we want to import setting "globally" from an external module19:15
hub_capwell SnowDust ill ask u a question19:17
SnowDustyes plz19:17
hub_capso the manager.py needs to get values from all the diff configs right?19:17
hub_caplike CONF.get(datastore).blah19:17
hub_capand those all need to be loaded if they are in the datastore.config, right?19:17
SnowDustyes19:17
SnowDustright19:18
hub_capso if u just import datastore.config, it should work19:18
hub_captry it by just importing it dynamically when u dynamically load the manager impl19:18
SnowDustbut its a chicken egg thing19:18
SnowDustwe dont know datastore until we import cfg19:18
hub_capsure import cfg, i never said not to19:19
SnowDustok so two imports19:19
*** shakayumi has quit IRC19:19
hub_caprestate your question w/ chicken/egg19:19
SnowDustcfg.default_datstore / cfg.datastore_manager19:19
SnowDustdetail what is our manager19:19
hub_capsure those are in DEFAULT right?19:20
SnowDustyeah19:20
hub_capthere is code that takes cfg.datastore.... and turns it into an impl19:20
SnowDustright19:20
hub_capand then the impl uses the CONF.get(datastore_manager).blah19:20
hub_capso right before u load cfg.datastore, just load the datastore.config file19:21
hub_capand itll load those19:21
SnowDustso .. another import from the impl.config19:21
SnowDustyeah19:21
hub_capright19:21
SnowDustgot that ..19:21
*** saju_m has quit IRC19:21
hub_capbut its exposing something ugly i think, but i dont think its your fault, its just we dont have a good separation between the common code and the impls19:22
*** saju_m has joined #openstack-trove19:22
SnowDusthub_cap : thats my point in the BP :)19:22
SnowDustlets do the separation as much as possible19:22
SnowDust:)19:22
hub_capsure :)19:23
hub_capok now i must go19:23
hub_capto fix my laptop19:23
SnowDustjust approve and go :D19:23
SnowDusthttps://blueprints.launchpad.net/openstack/?searchtext=refactoring-datastore-options-in-cfg19:23
*** hub_cap has quit IRC19:28
*** tanisdl has quit IRC19:34
*** tanisdl has joined #openstack-trove19:36
*** ViswaV has joined #openstack-trove19:42
*** ViswaV_ has joined #openstack-trove19:43
*** zuqiang has quit IRC19:43
*** ViswaV has quit IRC19:47
*** amcrn has joined #openstack-trove19:52
*** SnowDust has quit IRC19:54
*** michael-yu has quit IRC19:56
*** saju_m has quit IRC20:08
*** eguz has quit IRC20:08
*** eghobo has joined #openstack-trove20:08
*** saju_m has joined #openstack-trove20:08
*** eguz has joined #openstack-trove20:14
*** eghobo has quit IRC20:17
*** demorris has quit IRC20:38
*** harlowja is now known as harlowja_away20:52
*** yidclare has quit IRC21:01
*** harlowja_away is now known as harlowja21:14
*** demorris has joined #openstack-trove21:17
*** amytron has quit IRC21:17
*** hub_cap has joined #openstack-trove21:24
*** radez is now known as radez_g0n321:24
*** amytron has joined #openstack-trove21:25
*** tanisdl_ has joined #openstack-trove21:26
*** casanch1 has quit IRC21:27
*** saju_m has quit IRC21:28
*** tanisdl has quit IRC21:28
*** tanisdl_ is now known as tanisdl21:28
*** yidclare has joined #openstack-trove21:29
openstackgerritKhyati Sheth proposed a change to openstack/trove: Add support for 'trove root-enable' in Couchbase  https://review.openstack.org/8164021:37
*** amytron has quit IRC21:40
*** amytron has joined #openstack-trove21:41
*** pdmars has quit IRC21:43
*** freyes has quit IRC21:46
*** shalini_ has quit IRC22:03
*** shalini_ has joined #openstack-trove22:03
openstackgerritJenkins proposed a change to openstack/trove: Updated from global requirements  https://review.openstack.org/7964822:11
*** tanisdl has quit IRC22:16
*** robertmyers has quit IRC22:19
*** Barker has quit IRC22:22
*** rramirez_ has quit IRC22:37
*** achampion has quit IRC22:39
*** thedodd has quit IRC22:42
imsplitbitso... who's the tox pimp on the channel?22:49
imsplitbitI've got a bunch of unittests that pass when run through nosetests and pass when run with testr22:50
imsplitbitwhen I run tox -e py27 the explode with a bunch of sqlalchemy "not persisted" errors22:50
amcrnimsplitbit: tox version?22:50
imsplitbitpresumably because tox is parallelizing them and they're stomping on eachother?22:51
imsplitbithmm22:51
imsplitbittox 1.6.122:51
amcrnalright, cool, figured we'd check that first22:51
imsplitbitI believe is what it's says it's running22:51
imsplitbitI can paste errors22:52
amcrnthere's a known issue of ephemeral failures (see https://bugs.launchpad.net/trove/+bug/1294850); so out of curiosity, does the sqlalchemy error happen every single run?22:52
imsplitbitevery single run22:52
imsplitbitevery single test22:52
imsplitbithttp://pastie.org/895185522:52
imsplitbitI get a bunch of that stuff22:52
imsplitbitso the setup for each test clears the db and dumps in clean test data22:53
imsplitbitwell the setup for the test22:53
amcrnhrm, can you activate the .venv inside of .tox and pip freeze and see the version of sqlalchemy it's using?22:54
amcrnand cross compare to the version you see pip freeze'd in the public gerrit gate?22:54
amcrnsmells like a version drift22:55
amcrnif not, i'm out of immediate ideas22:55
imsplitbitI sure can22:56
imsplitbitthe very second you remind me how to do that :)22:56
amcrnhehe22:57
amcrnone sec22:57
amcrnshould be like: source .tox/py27/bin/activate22:57
amcrnthen pip freeze | grep sqlalchemy22:57
amcrnand you'd compare it to http://logs.openstack.org/40/81640/1/check/gate-trove-python27/f405eac/console.html, which has it as SQLAlchemy==0.7.1022:58
*** harlowja is now known as harlowja_away22:59
*** jcru has quit IRC22:59
imsplitbitsame vers23:00
imsplitbit0.7.1023:00
*** grapex has quit IRC23:02
amcrndo the newly introduced tests rely on the same db?23:03
amcrnif so, for kicks, try chaining them with "after" directives23:03
*** amytron has quit IRC23:03
imsplitbitamcrn: how does one do that in unittests?23:05
imsplitbitunittests don't have proboscis to set order23:06
amcrni haven't seen your new tests, but is it possible for each to push to a different table (that's augmented with an arbitrary ascending prefix or suffix)23:07
imsplitbitwell they use the instance_metadata model23:07
imsplitbitthat I've added23:07
imsplitbitand they're unittests that exercise the model and controller23:07
amcrnah, then pretty much no23:08
amcrn;D23:08
amcrnhrm, not sure then23:08
amcrnsorry i couldn't be of more help23:08
*** amytron has joined #openstack-trove23:11
*** rhodgin has quit IRC23:12
*** mattgriffin has quit IRC23:19
imsplitbitno worries23:20
imsplitbitI'll keep working on it23:20
imsplitbitit's just frustrating to know that the tests are good23:20
imsplitbitthey work23:20
imsplitbittox just hates my soul23:20
*** shalini_ has quit IRC23:21
openstackgerritNikhil Manchanda proposed a change to openstack/trove: Tox tests should not require errors in rigid order  https://review.openstack.org/8166823:21
imsplitbitlol23:22
imsplitbitmaybe he can propose a fix to make my tests work23:22
imsplitbit:D23:22
imsplitbitready, 1.2.3. go23:22
SlickNikimsplitbit: I haven't seen the "sqlalchemy persists" issues you speak of.23:24
SlickNik But I definitely found issues that surface when running tests through tox that don't surface when you run them through run_tests or testr directly.23:25
SlickNikMostly have to do with the validation returning errors in a different order than what the tests expect.23:25
SlickNikhttps://bugs.launchpad.net/trove/+bug/129485023:26
*** amytron has quit IRC23:28
SlickNikalso imsplitbit, try blowing away your .tox folder and starting clean23:29
imsplitbitdid that this morning23:29
SlickNikrm -rf ./.tox23:29
imsplitbitno love there23:29
juiceimsplitbit: i have had other issues with testr concurrency23:33
juiceimsplitbit: but it only seems to be a problem with mac os x23:33
*** jmontemayor has quit IRC23:34
juiceimsplitbit: try this "echo test_run_concurrency=echo 1 >> .testr.conf" and see if it resolves the issue23:34
*** harlowja_away is now known as harlowja23:38
openstackgerritNikhil Manchanda proposed a change to openstack/trove: Tox tests should not require errors in rigid order  https://review.openstack.org/8166823:44
*** kevinconway has quit IRC23:49
imsplitbitk thanks23:52
imsplitbitthat didn't change anything but I'll keep working on it23:52
imsplitbitgotta jet for now23:52
imsplitbitttyl23:52
juiceimsplitbit: take it easy23:53
*** demorris has quit IRC23:58
« Tuesday, 2014-03-18 Index Thursday, 2014-03-20 »

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!