| jgriffith | mtaylor: Becuase when it ran in Jenkins it complained about the "table created without InnoDB" | 00:00 |
|---|---|---|
| *** markmcclain has joined #openstack-dev | 00:00 | |
| issackelly | mtaylor this is the code that is failing https://github.com/markmc/openstack-gitdm/blob/master/do-it.sh#L89 | 00:00 |
| mtaylor | jgriffith: well, when you ran it in jenkins it was running with the wrong user/passwords, yeah? | 00:00 |
| jgriffith | mtaylor: Nope... it ran in Jenkins fine but failed the InnoDB checks. I was trying to get it to run on my local system to debug it and find "which" table was supposedly wrong. | 00:01 |
| mtaylor | ahhh | 00:01 |
| mtaylor | issackelly: username different on your local machine than gerrit? | 00:01 |
| jgriffith | mtaylor: I've now figured out how to get it to run on my system, but it doesn't fail :( | 00:01 |
| issackelly | mtaylor oh probably | 00:02 |
| issackelly | i mean yes | 00:02 |
| mtaylor | jgriffith: is your local database set up with default-engine=myisam in your my.cnf? | 00:02 |
| mtaylor | jgriffith: we set that up on the openstack slaves specifically to catch things that were configured poorly for end users :) | 00:02 |
| mtaylor | issackelly: you might need to set up username | 00:02 |
| *** arosen has quit IRC | 00:02 | |
| mtaylor | issackelly: in your .ssh/config file | 00:03 |
| jgriffith | mtaylor: Nope, I'll try adding that to test_migrations.conf | 00:03 |
| issackelly | mtaylor cool, I don't need any special permissions to do that then, that's what I was worried about. thanks | 00:03 |
| mtaylor | issackelly: because it doesn't look like gitdm allows you to set a username | 00:04 |
| mtaylor | issackelly: nope! all you should need is normal access | 00:04 |
| jgriffith | mtaylor: it non likey that... alright, let me back up and look at things again | 00:04 |
| issackelly | mtaylor cool, I don't need any special permissions to do that then, that's what I was worried about. thanks | 00:04 |
| issackelly | wrong Up + enter window. sorry | 00:04 |
| mtaylor | jgriffith: I wouldn't add that to test_migrations.conf ... I'd add it to your system /etc/mysql/my.cnf and restart mysql | 00:04 |
| jgriffith | mtaylor: yeah, good point | 00:05 |
| issackelly | mtaylor this is working now. thanks much | 00:06 |
| *** markmcclain has quit IRC | 00:12 | |
| *** rods has quit IRC | 00:12 | |
| *** fnaval has joined #openstack-dev | 00:15 | |
| jgriffith | mtaylor: BoooYaaa | 00:17 |
| *** zigo has quit IRC | 00:18 | |
| jgriffith | mtaylor: So it needed 'default-storage-engine = MyISAM' under the mysqld heading in my.cnf | 00:18 |
| jgriffith | mtaylor: Also it turns out these calls aren't using those user/passwords, they're default root/'' | 00:18 |
| jgriffith | mtaylor: Thanks for your help, I can do all sort of destructive things now | 00:18 |
| *** Aaton is now known as Aaton_off | 00:19 | |
| mtaylor | jgriffith: AWESOME | 00:21 |
| *** issackelly has quit IRC | 00:22 | |
| *** johnpostlethwait has quit IRC | 00:24 | |
| *** winston-d has joined #openstack-dev | 00:30 | |
| *** blamar has quit IRC | 00:33 | |
| winston-d | jgriffith, around? | 00:33 |
| *** e1mer has joined #openstack-dev | 00:33 | |
| vishy | mtaylor: https://review.openstack.org/#/c/9093/ | 00:34 |
| vishy | we really need to get those security fixes in | 00:34 |
| *** littleidea has quit IRC | 00:34 | |
| jgriffith | winston-d: Hey winston-d | 00:34 |
| *** harlowja has quit IRC | 00:36 | |
| winston-d | jgriffith, i've invalidated bug 1003664 since cinder doesn't do notification for now. | 00:36 |
| uvirtbot | Launchpad bug 1003664 in nova "volume_update db api needs to return the volume_ref" [Undecided,Fix released] https://launchpad.net/bugs/1003664 | 00:36 |
| jgriffith | winston-d: I was hoping cp16net would come back to that one | 00:37 |
| jgriffith | winston-d: He started it, but Jenkins failed and it rotted | 00:37 |
| jgriffith | winston-d: But yes, thank you for invalidating it | 00:38 |
| winston-d | and for bug 992729, i've submitted patch based on Nirmal's previous patch for nova. | 00:38 |
| uvirtbot | Launchpad bug 992729 in cinder "Add 'host' name to initialize/terminate volume connection" [Undecided,In progress] https://launchpad.net/bugs/992729 | 00:38 |
| mtaylor | vishy: sorry - I got MUCH closer ... https://review.openstack.org/#/c/9093/ | 00:38 |
| mtaylor | oh, you already linked to that :) | 00:38 |
| winston-d | i am wondering if cinder needs notification? | 00:38 |
| jgriffith | winston-d: Excellent!! | 00:38 |
| *** steveb___ has quit IRC | 00:38 | |
| *** lts has quit IRC | 00:39 | |
| jgriffith | winston-d: as far as cinder "needing" notifications, it was more just a blanket "if it's in nova-vol put it in cinder" | 00:39 |
| *** ncode has quit IRC | 00:40 | |
| *** steveb_ has joined #openstack-dev | 00:40 | |
| jgriffith | winston-d: it would still be good to return the latest info on the update via the API don't you think? | 00:40 |
| *** cp16net is now known as cp16net|away | 00:41 | |
| winston-d | yes, i agree. I can submit a patch for that (without 'bug fix' title). | 00:41 |
| jgriffith | winston-d: That would be great, and more appropriate as you pointed out | 00:42 |
| *** cp16net|away is now known as cp16net | 00:42 | |
| jgriffith | speak of the devil cp16net :) | 00:42 |
| cp16net | what up jgriffith | 00:43 |
| cp16net | oh... talking about me huh? | 00:43 |
| jgriffith | cp16net: winston-d and I were just talking about the notificiation bug that you worked on | 00:43 |
| winston-d | :) we were talking about your fix to bug 1003664, i invalidated it for cinder. but your fix can still do good for cinder although cinder doesn't do notification now. | 00:43 |
| jgriffith | cp16net: Nahh... | 00:43 |
| cp16net | :) | 00:44 |
| jgriffith | cp16net: never say anything bad of course | 00:44 |
| jgriffith | cp16net: Besides, everything is logged :) | 00:44 |
| cp16net | yeah i was reading up... | 00:44 |
| jgriffith | cp16net: Sound good from your perspective? | 00:45 |
| *** e1mer has quit IRC | 00:45 | |
| *** e1mer has joined #openstack-dev | 00:45 | |
| *** e1mer has joined #openstack-dev | 00:45 | |
| cp16net | oh yeah i submitted a review for that issue but people wanted things made more common | 00:46 |
| cp16net | and i've been busy with other project since then | 00:46 |
| jgriffith | cp16net: Yeah, and then gerrit abandoned it due to no activity | 00:46 |
| *** salgado-afk has quit IRC | 00:46 | |
| jgriffith | cp16net: No worries at all | 00:46 |
| cp16net | yup | 00:46 |
| cp16net | https://review.openstack.org/#/c/7517/ | 00:46 |
| *** markmcclain has joined #openstack-dev | 00:47 | |
| cp16net | i remember talking with nirmal about it and he said he would take a look at it at some point but i think he got busy as well | 00:48 |
| jgriffith | cp16net: depending on what you've got going on now I think winston-d might be interested in revamping it a bit just to update the references on create/delete | 00:49 |
| jgriffith | cp16net: minus the notification piece for now | 00:49 |
| jgriffith | winston-d: Correct me if I'm signing you up for something you didn't volunteer for :) | 00:50 |
| cp16net | ok sounds good to me | 00:50 |
| cp16net | those references from the api are pretty simple | 00:50 |
| winston-d | np, i'd love to do that but it's cp16net's credit. :) | 00:50 |
| *** \etc\bin has joined #openstack-dev | 00:50 | |
| *** e1mer has quit IRC | 00:51 | |
| jgriffith | I'll leave it to the two of you... If you want to work on cinder I've got plenty of things on my list :) | 00:51 |
| winston-d | and i'd like to do more besides bug fix. | 00:52 |
| jgriffith | winston-d: Well then.... | 00:53 |
| jgriffith | winston-d: A few things I was hoping to get started on... | 00:53 |
| jgriffith | winston-d: I'm removing the quotas info from cinder db. My thought is that the quota checks can be moved out to the nova volume api calls | 00:54 |
| cp16net | how close is cinder to being used > | 00:54 |
| cp16net | ? | 00:54 |
| jgriffith | winston-d: In other words nova/volume/cinder.py | 00:54 |
| *** japage has quit IRC | 00:54 | |
| jgriffith | winston-d: When you make the create call you can do the quota checking there since you'll have all the access to the nova quota info | 00:54 |
| winston-d | right | 00:55 |
| *** salv-orlando has left #openstack-dev | 00:55 | |
| jgriffith | winston-d: otherwise there's the task of maintaining/configuring quota info across nova and cinder which is no good | 00:55 |
| jgriffith | You can look at the quota stuff that's in nova/volume/api.py and basicly it needs to be implemented in nova/volume/cinder.py | 00:55 |
| *** \etc\bin has quit IRC | 00:55 | |
| jgriffith | winston-d: The other thing that we need to figure out is availability zones between the two | 00:56 |
| jgriffith | winston-d: TBH I don't have any hints/ideas on that yet, I should, but I haven't really thought about it | 00:56 |
| *** e1mer has joined #openstack-dev | 00:56 | |
| winston-d | hmm, since it's two different services, does av_zone need to be sync-ed? | 00:57 |
| *** epim has joined #openstack-dev | 00:57 | |
| jgriffith | winston-d: yes, the question is how | 00:57 |
| winston-d | ok, i will look into this and discuss with you once i have some idea/thoughts | 00:58 |
| jgriffith | winston-d: That's great, there are some things after that as well. As we get some of this knocked out we'll get to things that you might enjoy working on more so hang in there with me :) | 00:59 |
| *** fnaval has quit IRC | 00:59 | |
| *** fnaval has joined #openstack-dev | 00:59 | |
| winston-d | sure. :D | 00:59 |
| *** spiffxp has quit IRC | 01:03 | |
| jgriffith | cp16net: depends on who you mean used by, I use it every day :) | 01:04 |
| jgriffith | cp16net: So it made a functioning release in F2 | 01:05 |
| jgriffith | cp16net: Can be used in devstack today, soon will be default in devstack | 01:05 |
| cp16net | jgriffith: oh thats cool i didnt realize it was fully functional | 01:05 |
| cp16net | jgriffith: last i had heard it wasnt there yet | 01:05 |
| jgriffith | cp16net: All of the basics at least | 01:05 |
| cp16net | jgriffith: awesome! thats great to hear! | 01:06 |
| jgriffith | cp16net: it's useable in a test env | 01:06 |
| jgriffith | cp16net: There's still some cleanup and detail work but it's coming along | 01:06 |
| *** blamar has joined #openstack-dev | 01:07 | |
| *** Mandell has quit IRC | 01:07 | |
| *** jtran has quit IRC | 01:08 | |
| winston-d | oh, i think of sth. i'd like to submit patch for cinder to use openstack-common. | 01:08 |
| jgriffith | winston-d: :) | 01:09 |
| jgriffith | winston-d: Funny you should mention that | 01:09 |
| jgriffith | winston-d: There's a BUNCH of work to be done there as well | 01:09 |
| jgriffith | winston-d: If you want to tackle it that's great | 01:09 |
| jgriffith | winston-d: joshm has started some of that for logging, but everything else is fair game | 01:09 |
| jgriffith | winston-d: joshm = jmckenty | 01:10 |
| *** dtroyer_zzz is now known as dtroyer | 01:12 | |
| winston-d | ok, noted. is he still working on that? i didn't see any open review about that | 01:12 |
| jgriffith | winston-d: We'll have to ping him and see where it's at | 01:13 |
| winston-d | ok | 01:14 |
| jgriffith | winston-d: He's got it in his own github I believe | 01:14 |
| jgriffith | winston-d: He was close last I heard so he'll likely finish up the logging one | 01:14 |
| jgriffith | winston-d: Like I said you can take a look at all the other remaining pieces, make it match up with what's currently in nova | 01:14 |
| winston-d | sure, i'll do that. | 01:15 |
| jgriffith | winston-d: cp16net: Ok, I need to get out of here but will be back later for a short bit. Otherwise I'll catch up with you tomorrow | 01:16 |
| winston-d | ok. see u | 01:16 |
| jgriffith | good night | 01:16 |
| winston-d | gdnite | 01:16 |
| *** dtroyer is now known as dtroyer_zzz | 01:24 | |
| *** adjohn has quit IRC | 01:24 | |
| *** dtroyer_zzz is now known as dtroyer | 01:25 | |
| *** dtroyer is now known as dtroyer_zzz | 01:31 | |
| *** novas0x2a|laptop has quit IRC | 01:34 | |
| *** markmcclain has quit IRC | 01:38 | |
| *** markmcclain has joined #openstack-dev | 01:40 | |
| *** markmcclain has quit IRC | 01:43 | |
| *** PotHix has quit IRC | 01:44 | |
| *** jtran has joined #openstack-dev | 01:45 | |
| *** danwent has quit IRC | 01:50 | |
| *** jdurgin has quit IRC | 01:53 | |
| *** lzyeval has joined #openstack-dev | 01:54 | |
| *** jakedahn is now known as jakedahn_zz | 01:58 | |
| *** blamar has quit IRC | 02:00 | |
| *** mnewby has quit IRC | 02:02 | |
| *** jtran has quit IRC | 02:05 | |
| *** anniec has quit IRC | 02:11 | |
| *** mdomsch has joined #openstack-dev | 02:11 | |
| *** thingee_zz has quit IRC | 02:16 | |
| *** edygarcia has joined #openstack-dev | 02:20 | |
| *** reed has quit IRC | 02:23 | |
| *** markmcclain has joined #openstack-dev | 02:24 | |
| *** markmcclain has quit IRC | 02:28 | |
| *** markmcclain has joined #openstack-dev | 02:28 | |
| *** nati_ueno has quit IRC | 02:32 | |
| *** shang has joined #openstack-dev | 02:36 | |
| *** littleidea has joined #openstack-dev | 03:01 | |
| *** blamar has joined #openstack-dev | 03:02 | |
| *** edygarcia has quit IRC | 03:05 | |
| *** andrewbogott has quit IRC | 03:07 | |
| *** Mandell has joined #openstack-dev | 03:07 | |
| *** adalbas has quit IRC | 03:11 | |
| *** nati_ueno has joined #openstack-dev | 03:19 | |
| *** jakedahn_zz is now known as jakedahn | 03:20 | |
| *** blamar has quit IRC | 03:21 | |
| *** winston-d has quit IRC | 03:28 | |
| comstud | uhm | 03:38 |
| *** nati_uen_ has joined #openstack-dev | 03:42 | |
| *** nati_uen_ has quit IRC | 03:44 | |
| *** nati_ueno has quit IRC | 03:44 | |
| *** nati_uen_ has joined #openstack-dev | 03:44 | |
| *** jakedahn is now known as jakedahn_zz | 03:44 | |
| *** jakedahn_zz is now known as jakedahn | 03:46 | |
| jgriffith | exit/away | 03:47 |
| *** nati_uen_ has quit IRC | 03:49 | |
| *** nati_ueno has joined #openstack-dev | 03:49 | |
| *** nati_uen_ has joined #openstack-dev | 03:53 | |
| *** nati_ueno has quit IRC | 03:54 | |
| *** littleidea has quit IRC | 03:57 | |
| *** littleidea has joined #openstack-dev | 03:59 | |
| *** anniec has joined #openstack-dev | 03:59 | |
| *** eglynn has joined #openstack-dev | 04:00 | |
| *** nati_ueno has joined #openstack-dev | 04:03 | |
| *** eglynn__ has quit IRC | 04:04 | |
| *** nati_uen_ has quit IRC | 04:04 | |
| *** novas0x2a|laptop has joined #openstack-dev | 04:05 | |
| *** anderstj has joined #openstack-dev | 04:23 | |
| *** anniec has quit IRC | 04:26 | |
| *** anniec has joined #openstack-dev | 04:27 | |
| *** danwent has joined #openstack-dev | 04:33 | |
| *** danwent has quit IRC | 04:43 | |
| *** nati_ueno has quit IRC | 04:47 | |
| *** sam_wan has joined #openstack-dev | 04:54 | |
| *** dolphm has joined #openstack-dev | 04:54 | |
| *** danwent has joined #openstack-dev | 05:03 | |
| *** andrewsmedina has quit IRC | 05:14 | |
| *** jtran has joined #openstack-dev | 05:25 | |
| *** danwent has quit IRC | 05:26 | |
| *** dolphm has quit IRC | 05:27 | |
| *** steveb_ has quit IRC | 05:28 | |
| *** anniec has quit IRC | 05:31 | |
| *** littleidea has quit IRC | 05:35 | |
| *** jtran has quit IRC | 05:35 | |
| *** zigo has joined #openstack-dev | 05:36 | |
| *** renier has quit IRC | 05:38 | |
| *** renier has joined #openstack-dev | 05:38 | |
| *** zigo has quit IRC | 05:41 | |
| *** nati_ueno has joined #openstack-dev | 05:57 | |
| *** zaitcev has quit IRC | 05:58 | |
| *** nati_ueno has quit IRC | 06:01 | |
| *** andrewsben is now known as andrewsben_is_a_ | 06:11 | |
| *** sam_wan has quit IRC | 06:11 | |
| *** anderstj has quit IRC | 06:13 | |
| *** sam_wan has joined #openstack-dev | 06:15 | |
| zykes- | will there be a v3 api in folsom 2 ? | 06:20 |
| *** adjohn has joined #openstack-dev | 06:21 | |
| zykes- | sam_wan: ping | 06:21 |
| sam_wan | @zykes: what's up? | 06:26 |
| *** mrunge has joined #openstack-dev | 06:37 | |
| *** dolphm has joined #openstack-dev | 06:38 | |
| *** dolphm has quit IRC | 06:42 | |
| zykes- | sam_wan: oracle using openstack ? *ponders* | 06:43 |
| *** mindpixel has joined #openstack-dev | 06:44 | |
| sam_wan | @zykes, i'm not sure | 06:57 |
| *** dachary has joined #openstack-dev | 06:59 | |
| *** vanchester has joined #openstack-dev | 07:00 | |
| *** Mandell has quit IRC | 07:08 | |
| *** reidrac has joined #openstack-dev | 07:12 | |
| *** Ryan_Lane has quit IRC | 07:16 | |
| *** eglynn has quit IRC | 07:18 | |
| *** markmcclain has quit IRC | 07:19 | |
| *** amotoki has joined #openstack-dev | 07:31 | |
| *** vanchester has quit IRC | 07:49 | |
| *** dachary has quit IRC | 07:58 | |
| *** lzyeval has quit IRC | 07:59 | |
| *** adjohn has quit IRC | 08:03 | |
| *** utlemming has quit IRC | 08:10 | |
| *** utlemming has joined #openstack-dev | 08:11 | |
| *** maploin has joined #openstack-dev | 08:11 | |
| *** maploin has quit IRC | 08:11 | |
| *** maploin has joined #openstack-dev | 08:11 | |
| *** maploin has quit IRC | 08:18 | |
| *** dachary has joined #openstack-dev | 08:22 | |
| *** jakedahn is now known as jakedahn_zz | 08:22 | |
| *** apevec has joined #openstack-dev | 08:26 | |
| *** openstackgerrit has joined #openstack-dev | 08:41 | |
| zykes- | sam_wan: aren't you from there ? ;) | 08:49 |
| *** danpb has joined #openstack-dev | 08:50 | |
| *** eglynn has joined #openstack-dev | 08:53 | |
| *** eglynn has quit IRC | 08:57 | |
| *** eglynn has joined #openstack-dev | 08:57 | |
| *** darraghb has joined #openstack-dev | 09:02 | |
| *** amotoki has quit IRC | 09:05 | |
| sam_wan | @zykes: how do you know? :) | 09:07 |
| *** sam_wan has left #openstack-dev | 09:14 | |
| *** rbasak has quit IRC | 09:17 | |
| *** maploin has joined #openstack-dev | 09:19 | |
| *** maploin has quit IRC | 09:20 | |
| *** maploin has joined #openstack-dev | 09:20 | |
| *** pixelbeat has joined #openstack-dev | 09:22 | |
| *** rbasak has joined #openstack-dev | 09:29 | |
| *** openstackgerrit has quit IRC | 09:30 | |
| *** rods has joined #openstack-dev | 09:30 | |
| *** openstackgerrit has joined #openstack-dev | 09:30 | |
| eglynn | mikal: there? | 09:31 |
| *** maploin has quit IRC | 09:36 | |
| *** salv-orlando has joined #openstack-dev | 09:53 | |
| *** e1mer has quit IRC | 09:57 | |
| *** e1mer has joined #openstack-dev | 09:59 | |
| mikal | eglynn: yeah, what's up? | 10:01 |
| *** e1mer has quit IRC | 10:02 | |
| eglynn | mikal: no worries, just clarifying your comment on https://review.openstack.org/9317 ... I just put it in a comment in gerrit instead | 10:02 |
| *** Leander_ has joined #openstack-dev | 10:03 | |
| * mikal clicks | 10:04 | |
| Leander_ | hello all! | 10:04 |
| Leander_ | does anyone around here have some knowledge on how the libvirt test cases work? | 10:05 |
| *** e1mer has joined #openstack-dev | 10:05 | |
| *** e1mer has joined #openstack-dev | 10:05 | |
| *** chrisfer has joined #openstack-dev | 10:11 | |
| *** eafonichev has quit IRC | 10:12 | |
| *** maploin has joined #openstack-dev | 10:16 | |
| *** maploin has quit IRC | 10:16 | |
| *** maploin has joined #openstack-dev | 10:16 | |
| *** RoboticCheese has quit IRC | 10:22 | |
| *** RoboticCheese has joined #openstack-dev | 10:23 | |
| zykes- | does one use flake in openstack ?= | 10:23 |
| *** maploin has quit IRC | 10:24 | |
| *** asalkeld has left #openstack-dev | 10:27 | |
| *** RoboticCheese has joined #openstack-dev | 10:28 | |
| *** e1mer has quit IRC | 10:39 | |
| *** e1mer has joined #openstack-dev | 10:42 | |
| *** e1mer has joined #openstack-dev | 10:42 | |
| *** eafonichev has joined #openstack-dev | 10:42 | |
| *** johngarbutt has joined #openstack-dev | 10:43 | |
| *** vanchester has joined #openstack-dev | 11:03 | |
| *** sandywalsh has quit IRC | 11:03 | |
| vanchester | Hello everyone. Question about keystone. Why there is a lot of layers with NovaTestRunner etc in comparison with swift tests? | 11:04 |
| *** zul has quit IRC | 11:13 | |
| *** rbasak has quit IRC | 11:21 | |
| *** rbasak has joined #openstack-dev | 11:35 | |
| *** sandywalsh has joined #openstack-dev | 11:37 | |
| *** Leander_ has quit IRC | 11:45 | |
| *** markmcclain has joined #openstack-dev | 11:50 | |
| *** sandywalsh_ has joined #openstack-dev | 11:53 | |
| *** salgado has joined #openstack-dev | 11:53 | |
| *** salgado has joined #openstack-dev | 11:53 | |
| *** sandywalsh has quit IRC | 11:53 | |
| *** e1mer has quit IRC | 11:58 | |
| *** e1mer has joined #openstack-dev | 12:03 | |
| *** mrunge has quit IRC | 12:03 | |
| *** apevec has quit IRC | 12:11 | |
| *** dolphm has joined #openstack-dev | 12:13 | |
| *** adalbas has joined #openstack-dev | 12:20 | |
| *** ncode has joined #openstack-dev | 12:21 | |
| *** ncode has quit IRC | 12:21 | |
| *** ncode has joined #openstack-dev | 12:21 | |
| *** alaski has joined #openstack-dev | 12:22 | |
| *** dolphm has quit IRC | 12:26 | |
| *** rods has quit IRC | 12:27 | |
| *** rods has joined #openstack-dev | 12:27 | |
| *** sandywalsh has joined #openstack-dev | 12:28 | |
| *** sandywalsh_ has quit IRC | 12:28 | |
| *** e1mer has quit IRC | 12:31 | |
| *** rbasak has quit IRC | 12:31 | |
| *** dolphm has joined #openstack-dev | 12:33 | |
| vanchester | Hello everyone. Question about keystone. Why there is a lot of layers with NovaTestRunner etc in comparison with swift tests? | 12:38 |
| *** e1mer has joined #openstack-dev | 12:40 | |
| *** rbasak has joined #openstack-dev | 12:44 | |
| *** e1mer has quit IRC | 12:45 | |
| *** vanchester has quit IRC | 12:47 | |
| *** dolphm has quit IRC | 12:52 | |
| *** Leander_ has joined #openstack-dev | 13:02 | |
| *** zul has joined #openstack-dev | 13:04 | |
| *** blamar has joined #openstack-dev | 13:14 | |
| *** vanchester has joined #openstack-dev | 13:16 | |
| *** Madkiss has joined #openstack-dev | 13:18 | |
| Madkiss | hi folks | 13:18 |
| Madkiss | could somebody possibly take a look at https://blueprints.launchpad.net/glance/+spec/ocf-support and https://blueprints.launchpad.net/keystone/+spec/ocf-support and possibly approve them? | 13:18 |
| *** sandywalsh_ has joined #openstack-dev | 13:25 | |
| *** sandywalsh has quit IRC | 13:26 | |
| *** mdomsch has quit IRC | 13:26 | |
| *** blamar has quit IRC | 13:29 | |
| vanchester | Hello everyone. Question about keystone. Why there is a couple of layers with NovaTestRunner etc in comparison with swift tests? | 13:33 |
| Leander_ | danbp: ping | 13:33 |
| Leander_ | dabpb: ping | 13:34 |
| Leander_ | danpb: ping* | 13:34 |
| *** edygarcia has joined #openstack-dev | 13:35 | |
| danpb | Leander_: pong ? | 13:35 |
| Leander_ | danpb: hello! thx for the code in the email you've sent! However, i've run into my infamous issue with the error "import libvirt" again. (about the ping, i'm not sure . but i've seen people use it around here like this before). | 13:37 |
| danpb | Leander_: i'm not sure what you mean by that - can you pastebin.com the output you see from the test | 13:39 |
| Leander_ | danpb: http://paste.openstack.org/show/19297/ | 13:39 |
| danpb | try changing | 13:41 |
| danpb | import libvirt | 13:41 |
| danpb | to | 13:41 |
| danpb | global libvirt | 13:41 |
| danpb | if libvirt is None: | 13:41 |
| danpb | libvirt = __import__('libvirt') | 13:41 |
| Leander_ | i already did this once, however on of the reviewers said this | 13:43 |
| Leander_ | Johannes Erdfel: It's unnecessary to make this that complicated. You can remove the libvirt global and change this code to: import libvirt - The import won't happen until get_diagnostics is called. | 13:43 |
| danpb | hmm, i don't think that's correct - in the test suite we need to be able to setup a fake libvirt import | 13:44 |
| danpb | which is what the code I post above allows for | 13:44 |
| *** maoy has joined #openstack-dev | 13:45 | |
| *** roge has joined #openstack-dev | 13:46 | |
| Leander_ | danpb: error and diff : http://paste.openstack.org/show/19298/ | 13:47 |
| *** dprince has joined #openstack-dev | 13:48 | |
| danpb | +import xml.etree | 13:48 |
| danpb | + | 13:48 |
| danpb | +libvirt = None | 13:48 |
| danpb | don't put in that global initializer | 13:48 |
| danpb | that's overriding what the test suite sets | 13:49 |
| *** cp16net is now known as cp16net|away | 13:49 | |
| *** cp16net|away is now known as cp16net | 13:49 | |
| Leander_ | danpb: NameError: global name 'libvirt' is not defined | 13:49 |
| Leander_ | after i removed the libvirt = None | 13:49 |
| danpb | oh i see why | 13:50 |
| danpb | to be honest you're creating pain for yourself here by having the code in the separate diagnostics.py file | 13:50 |
| danpb | the fake environment the test suite sets up only applies to the main driver.py | 13:51 |
| Leander_ | yeah, well i'm working as a researcher here at the university and they insist i put some kinda of recognition in the license | 13:52 |
| Leander_ | of the file. Is there any way i can add the file to the test set up? | 13:53 |
| Leander_ | danpb: ^ | 13:55 |
| danpb | what if you just add the following | 13:55 |
| danpb | @test.skip_if(missing_libvirt(), "Test requires libvirt") | 13:55 |
| danpb | immediately before | 13:55 |
| danpb | def test_diagnostic_full(self): | 13:55 |
| sdague | Leander_: you can add your copyright to driver.py in a patch with these changes over there | 13:55 |
| sdague | instead of the second file | 13:56 |
| Leander_ | sdage: do i add this after "Copyright (c) 2011 Piston Cloud Computing, Inc" | 14:00 |
| Leander_ | Copyright (c) 2012. | 14:00 |
| Leander_ | Universidade do Minho | 14:00 |
| Leander_ | Leander Beernaert | 14:00 |
| Leander_ | Miguel Matos | 14:00 |
| Leander_ | Jose Pereira | 14:00 |
| Leander_ | Ricardo Vilaca | 14:00 |
| Leander_ | Rui Oliveira | 14:00 |
| Leander_ | ? | 14:00 |
| Leander_ | sdague: ^ | 14:00 |
| danpb | Leander_: does each individual person really claim copyright over this code ? | 14:01 |
| danpb | Leander_: if the university claims copyright on your work, then it is more concise to just do | 14:02 |
| danpb | "Copyright (c) 2012 Universidade do Minho" | 14:02 |
| danpb | as was done with other copyright lines in the file | 14:02 |
| *** EmilienM has joined #openstack-dev | 14:02 | |
| *** anniec has joined #openstack-dev | 14:03 | |
| Leander_ | danpb: i'm gonna have to check with my superiors first, don't want there "wrath" upon me | 14:04 |
| Leander_ | however i'll work on putting everything in a single file | 14:04 |
| Leander_ | in the driver file* | 14:04 |
| *** mnewby has joined #openstack-dev | 14:05 | |
| Leander_ | their* | 14:05 |
| *** salv-orlando has quit IRC | 14:06 | |
| Leander_ | danpb: on another note; my function get_io_devices, can i added it to the driver class or do i declare the function inside the get_diagnostics method? | 14:07 |
| sdague | Leander_: you would add: Copyright (c) 2012 COPYRIGHT HOLDER NAME for every copy right holder. If it's really seperately held by 6 people so be it. If it's really university copyright, just put that | 14:07 |
| sdague | for instance my patches come in: Copyright (c) 2012 IBM Corp. | 14:08 |
| *** sandywalsh_ has quit IRC | 14:08 | |
| *** apevec has joined #openstack-dev | 14:09 | |
| *** rkukura has left #openstack-dev | 14:09 | |
| Leander_ | sdague: i'll check with my superiors first, but i believe the universtity copyright should suffice. Do you also add copy rights to the test files? | 14:09 |
| danpb | Leander_: either way works fine | 14:10 |
| *** blamar has joined #openstack-dev | 14:10 | |
| *** blamar has joined #openstack-dev | 14:11 | |
| *** nikhil has quit IRC | 14:11 | |
| *** markmcclain has quit IRC | 14:12 | |
| Leander_ | ok | 14:12 |
| Leander_ | thx | 14:12 |
| Leander_ | danpb: another question, there is a command, memory stats which can return other values | 14:13 |
| Leander_ | values which my libvirt version has no access to | 14:13 |
| Leander_ | how do i account that in the test, since currently it'll most likely fail should any any value be returned from memStat(); | 14:14 |
| danpb | that doesn't matter - the code copes with anmything | 14:14 |
| danpb | and test case demonstrates that | 14:14 |
| *** Mandell has joined #openstack-dev | 14:19 | |
| sdague | Leander_: typically if you are adding > 20 lines of code to a file, it's fine to add your copyright clause. | 14:20 |
| *** Dr_Who has joined #openstack-dev | 14:22 | |
| *** rkukura has joined #openstack-dev | 14:26 | |
| Leander_ | sdague: ok thx | 14:26 |
| *** nunosantos has joined #openstack-dev | 14:28 | |
| *** sacharya has joined #openstack-dev | 14:28 | |
| *** jgriffith has quit IRC | 14:29 | |
| *** mdomsch has joined #openstack-dev | 14:30 | |
| *** cp16net is now known as cp16net|away | 14:30 | |
| *** datsun180b has joined #openstack-dev | 14:30 | |
| *** markmcclain has joined #openstack-dev | 14:31 | |
| *** dtroyer_zzz is now known as dtroyer | 14:31 | |
| *** jgriffith has joined #openstack-dev | 14:31 | |
| *** sandywalsh has joined #openstack-dev | 14:31 | |
| *** anderstj has joined #openstack-dev | 14:36 | |
| *** Mandell has quit IRC | 14:39 | |
| *** jgriffith has quit IRC | 14:41 | |
| *** jgriffith has joined #openstack-dev | 14:42 | |
| *** fnaval has quit IRC | 14:43 | |
| Leander_ | danpb: is this the correct way to raise a libvirt error: raise libvirt_driver.libvirtError(0,0,"fake failure")? I always run into this error | 14:45 |
| Leander_ | AttributeError: 'str' object has no attribute 'libvirtError' | 14:45 |
| *** sandywalsh has quit IRC | 14:45 | |
| *** nikhil has joined #openstack-dev | 14:46 | |
| *** cp16net|away is now known as cp16net | 14:51 | |
| *** littleidea has joined #openstack-dev | 14:52 | |
| *** dubsquared has joined #openstack-dev | 14:54 | |
| *** sandywalsh has joined #openstack-dev | 14:55 | |
| *** mnewby has quit IRC | 14:58 | |
| Leander_ | can some explain to me what i'm doing wrong: http://paste.openstack.org/show/19299/ | 15:03 |
| Leander_ | i raise the exception with "raise libvirt.libvirtError('vcpu missing')" | 15:03 |
| *** fnaval has joined #openstack-dev | 15:05 | |
| *** heckj has joined #openstack-dev | 15:07 | |
| *** rnirmal has joined #openstack-dev | 15:07 | |
| *** sstent has quit IRC | 15:07 | |
| *** fnaval has quit IRC | 15:08 | |
| *** sstent has joined #openstack-dev | 15:08 | |
| ayoung | heckj, I'm trying to deal with the SQL tables for Signed tokens. I had punted on the migrate issue, but I see now that I have to deal with it. Do you have a moment? | 15:08 |
| *** fnaval has joined #openstack-dev | 15:09 | |
| heckj | ayoung: o/ | 15:09 |
| heckj | ayoung: still waking up with coffee though :-) | 15:10 |
| ayoung | OK, so the current table has an Id primary key | 15:10 |
| ayoung | heckj, me too, really. and we ran out of the Peets, so the coffee is meg | 15:10 |
| ayoung | ,meh | 15:10 |
| *** sleepsonzzz is now known as sleepsonthefloor | 15:10 | |
| ayoung | heckj, anyway, sql alchemy requires a primary key field, and the signed token is too big to be it | 15:11 |
| ayoung | let me post a link | 15:11 |
| *** danwent has joined #openstack-dev | 15:11 | |
| heckj | ayoung: I'm pulling up the code now - just a sec | 15:12 |
| ayoung | heckj, https://github.com/admiyo/keystone/blob/pki-token-generation/keystone/token/backends/sql.py | 15:12 |
| ayoung | https://github.com/admiyo/keystone/blob/pki-token-generation/keystone/token/backends/sql.py#L31 | 15:12 |
| sdague | what was the commit tag for "requires doc changes" ? | 15:13 |
| heckj | ayoung: yeah - sql.Column(sql.String(64), ...) | 15:13 |
| ayoung | heckj, the 'key' field is the new primary key, to keep SQL alchemy happy, and it is a hash of the id field | 15:13 |
| ayoung | but... | 15:13 |
| *** openstackgerrit has quit IRC | 15:14 | |
| ayoung | for migration sake, it would be easier if id stayed id, and then I could name the other field something like "marshalled:" or "token_data" or "I_dont_know_what_to_call_this" | 15:14 |
| *** anderstj has quit IRC | 15:14 | |
| ayoung | the thing is, all operations are done with the blob | 15:14 |
| *** blamar has quit IRC | 15:15 | |
| ayoung | er...technically clob, I guess.. | 15:15 |
| ayoung | and all of the APIs use token_id, which becomes id...and I can hide the hashing away from the rest of the world. | 15:15 |
| ayoung | heckj, so do I err on the side of simpler Python code, or simpler database migration | 15:16 |
| *** openstackgerrit has joined #openstack-dev | 15:16 | |
| ayoung | and which makes sense long term.... | 15:16 |
| ayoung | I made a quick attempt at renaming key->id and id->token_data, but getting it right in all of the tests is a bit of a struggle | 15:17 |
| heckj | ayoung: I'm not clear on the tradeoff and why you need to make it. Can you simply expand the definition of "id" to be 2048 or whatever is sufficient for the PKI token? Where does hash come in and why is it important? | 15:17 |
| ayoung | heckj, nope | 15:17 |
| ayoung | SQL alechmey won;'t let me | 15:17 |
| ayoung | alchemy | 15:17 |
| heckj | ahhhh | 15:17 |
| ayoung | aptly named | 15:18 |
| ayoung | heckj, token data is > than the maximum key size | 15:18 |
| heckj | I don't know much about the SQLAlchemy side of it... | 15:18 |
| ayoung | I think that is even a MySQL constraint | 15:18 |
| heckj | ayoung: ok, now I see the issue | 15:18 |
| heckj | hmmm | 15:18 |
| ayoung | we get something like 700 bytes | 15:18 |
| ayoung | which due to overhead comes out to about 500 chars for some reason that I've not investigated | 15:19 |
| heckj | ayoung: did we ever hear back from jroualt on the max-length issue he was having? | 15:19 |
| ayoung | nope | 15:19 |
| heckj | kk | 15:19 |
| ayoung | and frankly, I don't find it a compelling argument | 15:19 |
| ayoung | token length can easily increase in the future | 15:19 |
| heckj | ayoung: without anything really backing it up, neither do it | 15:19 |
| *** lts has joined #openstack-dev | 15:19 | |
| heckj | (do i) | 15:20 |
| ayoung | even without PKI | 15:20 |
| heckj | Okay - back to SQLAlchemy ick | 15:20 |
| ayoung | So I don't really understand the migration mechanism | 15:20 |
| ayoung | I assume that I would have to do something like this: | 15:20 |
| ayoung | add field named key | 15:21 |
| ayoung | drop index on id, add it on key | 15:21 |
| ayoung | drop column id | 15:21 |
| ayoung | add a new columnet named id with the larger size | 15:21 |
| sdague | any core devs want to Approve - https://review.openstack.org/#/c/9410/ and https://review.openstack.org/#/c/9411/ (I've got 2 more patches in the queue, but don't want to invalidate the reviewed check on those, so am waiting to push to review until after they land) | 15:21 |
| ayoung | now...that assumes that we just drop all of the currnet tokens. | 15:21 |
| *** reidrac has quit IRC | 15:21 | |
| heckj | I think your sense of adding a field to store the blob/clob/whatever (the big damn field of PKI plugh) is on the right track. I think I might just leave "id" as a UUID token thingy instead of trying to mess with a hash there - leave that as the primary key, but add an index on the key. Do you use the hash anywhere? | 15:22 |
| ayoung | heckj, yes, I use hash..here's how | 15:22 |
| ayoung | https://github.com/admiyo/keystone/blob/pki-token-generation/keystone/token/backends/sql.py#L57 | 15:22 |
| Leander_ | can some explain to me what i'm doing wrong: http://paste.openstack.org/show/19299/, i raise the exception with "raise libvirt.libvirtError('vcpu missing')" | 15:22 |
| heckj | In short, I'm leaning towards a bit more complexity in the python side than the SQL side | 15:22 |
| ayoung | I'm almost tempted to call neither column id | 15:23 |
| ayoung | maybe hash and data? | 15:23 |
| heckj | ayoung: don't - leave ID alone and autogenerate it, and make a new "key" and "hash" and index both or either of them. It'll make things much more clear | 15:23 |
| heckj | ayoung: hash and data is good too | 15:23 |
| ayoung | heckj, only one problem with that | 15:24 |
| ayoung | all of the APIs pass in 'id' | 15:24 |
| *** zns has joined #openstack-dev | 15:24 | |
| *** zns has left #openstack-dev | 15:24 | |
| ayoung | that value is the token from the wire | 15:24 |
| ayoung | so it maps to 'data' | 15:24 |
| heckj | ayoung: yeah, it means extending that internal API | 15:24 |
| ayoung | so if I drop the id column, at least there is no incorrect mapping | 15:25 |
| heckj | Or overloading it - I think I'd prefer to extend it, but I like to try and avoid overloading methods unless it's really clear what the options are for the parameters | 15:25 |
| heckj | you went to a fair bit of trouble in that code ot make the has something akin to ID there didn't you | 15:25 |
| ayoung | yes and no | 15:25 |
| ayoung | it was pretty easy to implement | 15:26 |
| heckj | So you could grab the token off the wire, get it's hash and look that up as an ID | 15:26 |
| ayoung | true | 15:26 |
| sdague | Leander_: actually, I think you are hitting something that danpb has a change going in to fix | 15:26 |
| sdague | you might just need to wait until that lands (hopefully today) | 15:26 |
| ayoung | the token code needs some cleanup anyway...it should be in keystone/token and not in keystone/service.py | 15:27 |
| ayoung | like the identity code is set up | 15:27 |
| heckj | ayoung: yep - and it's not using the base models.py either | 15:27 |
| Leander_ | sdague: i find it odd that it's only my code that triggers that error, the remaining tests run just fine :S | 15:27 |
| heckj | ayoung: gimme a sec to quickly review the token/core.py | 15:28 |
| sdague | Leander_: is this patch in the review queue, I can look | 15:28 |
| ayoung | heckj, one related thing: the authenticate call currently returns the service catalog. I think that it is correct for POST /tokens to do that, but why do the remote services need it? | 15:28 |
| ayoung | I, when I pass a token to nova, and nova validates, why does it need the service catalog? | 15:29 |
| Leander_ | sdague: not the latest changes i made, i will submit a draft in a minute | 15:29 |
| heckj | ayoung: so that they know the endpoints to communicate with to make relevant calls - right now it's minimal usage there, but growing | 15:29 |
| ayoung | The reason I ask is I have, in my latest hack, encoded the service catalog in the token, which grealt increases the token size | 15:29 |
| ayoung | greatly | 15:29 |
| ayoung | I am OK with doing that, just want to make sure it makes sense | 15:29 |
| heckj | ayoung: i.e. nova needs to know the API endpoint for cinder to make volume calls... | 15:29 |
| ayoung | heckj, does that vary per user? | 15:30 |
| heckj | ayoung: yeah - it really trudges up the whole kit on a response from POST /tokens | 15:30 |
| heckj | ayoung: yep - because some of the URIs are encoded with tenant_id and such (PITA) | 15:30 |
| ayoung | prior to that, the token was this long https://github.com/admiyo/keystone/blob/3aa04a627e1d9ca16c58719cfd81c90912baa77f/tests/signing/auth_token.pem | 15:31 |
| *** blamar has joined #openstack-dev | 15:31 | |
| *** blamar has quit IRC | 15:32 | |
| ayoung | heckj, afterwards it looks like this http://fpaste.org/V29m/ | 15:32 |
| *** blamar has joined #openstack-dev | 15:32 | |
| heckj | ayoung: yeah - about double in size | 15:33 |
| ayoung | heckj, that is assuming the service catalog stays small, too | 15:33 |
| heckj | ayoung: in most instances, it will be. Not entirely sure how large some of the service provider catalogs are - they haven't exactly been sharing that detail. | 15:34 |
| ayoung | heckj, I'll leave it for now, but we might want to revisit that decision | 15:35 |
| Leander_ | sdague: apparently the exception is not libvirt.libvirtError object but a string | 15:36 |
| Leander_ | is that the issue danpb is solving? | 15:36 |
| heckj | ayoung: sounds good | 15:37 |
| *** openstackgerrit has quit IRC | 15:37 | |
| *** openstackgerrit has joined #openstack-dev | 15:37 | |
| heckj | ayoung: It seems like your sense of using the hash for ID works really well with the token/core.py internal API, so I'd lean towards there, with the token/backends/signing.py (or hwatever you call it) taking up the heavy lifting where it's needed, and may some if/then work on the front to get that hash and work with it. I still think you'll need to somehow extend the core.py API though | 15:38 |
| *** belliott_ is now known as belliott | 15:43 | |
| ayoung | heckj, I don't think I do. It works as is. THe question is how are we going to handle upgrades...here is what I'd *like* to do | 15:45 |
| ayoung | Check token length. If it is the current length (64 bytes) or shorter, do an online lookup | 15:46 |
| eglynn | sdague: looks that alternate form of assertRaises is not supported in py26: https://jenkins.openstack.org/job/gate-nova-python26/1605/console | 15:46 |
| ayoung | if it is longer, use the PKI mechanism | 15:46 |
| eglynn | sdague: (with reference to https://review.openstack.org/9371) | 15:46 |
| ayoung | heckj, then, as part of the migrate, copy the id column into the hash column | 15:46 |
| ayoung | on the server side | 15:47 |
| *** danpb has quit IRC | 15:47 | |
| ayoung | if there is no data, create a new token using the authenticate code | 15:47 |
| heckj | ayoung: sounds pretty reasonable to me | 15:47 |
| ayoung | yeah...just more work | 15:48 |
| ayoung | could I possibly do that as a follow on patch? | 15:48 |
| *** belliott_ has joined #openstack-dev | 15:48 | |
| ayoung | for now, I can just truncate the table | 15:48 |
| *** belliott has quit IRC | 15:48 | |
| ayoung | I mean, it is during a keystone update.... | 15:48 |
| ayoung | so for development, we accept that it means starting with a blank token table...I think that is OK for the short run | 15:49 |
| sdague | eglynn: ok, good call. | 15:49 |
| heckj | ayoung: I think separate patches for that are just fine | 15:49 |
| *** zul has quit IRC | 15:49 | |
| sdague | eglynn: oops, I didn't catch that this was essex :) | 15:50 |
| ayoung | OK. I think then, that I will leave 'key' as is, and change 'id' to 'token_data' | 15:50 |
| ayoung | then resubmit this one | 15:50 |
| sdague | eglynn: once you submit the jenkins fix, I'll +1 it, sorry for leading you down the wrong path there | 15:51 |
| *** danwent_ has joined #openstack-dev | 15:51 | |
| eglynn | sdague: no problem! | 15:52 |
| *** jtran has joined #openstack-dev | 15:53 | |
| *** danwent has quit IRC | 15:55 | |
| *** danwent_ is now known as danwent | 15:55 | |
| *** dubsquared has quit IRC | 15:56 | |
| *** dubsquared has joined #openstack-dev | 15:56 | |
| *** andrewsben_is_a_ is now known as andrewsben | 15:58 | |
| Leander_ | did the authors file dissapear | 15:59 |
| Leander_ | from the repository? | 15:59 |
| Leander_ | i get this error message trying to fetch the latest code CONFLICT (modify/delete): Authors deleted in HEAD and modified in Adds diagnostics command for the libvirt driver.. Version Adds diagnostics command for the libvirt driver. of Authors left in tree. | 15:59 |
| *** belliott_ is now known as belliott | 15:59 | |
| *** cp16net is now known as cp16net|away | 16:03 | |
| *** andrewbogott has joined #openstack-dev | 16:03 | |
| *** anniec has quit IRC | 16:04 | |
| ayoung | heckj, actually...I think I need to leave id alone. The token as defined is used everywhere, and so things expect and Id field. I think my current approach is OK, as the APIs take in an id value, and just caluculate the hash value as a SQL optimization. | 16:05 |
| ayoung | heckj, I'll look into the migration...who knows that code base? | 16:05 |
| heckj | ayoung: termie is it for that one | 16:06 |
| maoy | vishy: ping | 16:08 |
| *** torgomatic has quit IRC | 16:08 | |
| Leander_ | Does anyone know what happend to the Authors file? | 16:10 |
| *** primeministerp has quit IRC | 16:10 | |
| *** primeministerp has joined #openstack-dev | 16:10 | |
| *** cp16net|away is now known as cp16net | 16:11 | |
| mtaylor | Leander_: it's auto generated | 16:12 |
| mtaylor | Leander_: so just go git rm Authors in your local repo to resolve the confluct | 16:13 |
| mtaylor | conflict | 16:13 |
| Leander_ | mtaylor: ah, ok. thx | 16:14 |
| *** issackelly has joined #openstack-dev | 16:17 | |
| *** issackelly has joined #openstack-dev | 16:19 | |
| *** zaitcev has joined #openstack-dev | 16:19 | |
| *** salgado is now known as salgado-lunch | 16:21 | |
| Leander_ | well i'm going, have a nice weekend | 16:25 |
| *** zul has joined #openstack-dev | 16:25 | |
| Leander_ | good bye | 16:25 |
| *** Leander_ has quit IRC | 16:25 | |
| *** mindpixel has quit IRC | 16:25 | |
| *** lloydde has joined #openstack-dev | 16:26 | |
| *** dubsquared has quit IRC | 16:30 | |
| andrewbogott | If any openstack-common reviewers aren't on vacation, I could use half a dozen or so reviews. Otherwise I may be forced to spend a weekend without coding :( | 16:32 |
| *** sandywalsh has quit IRC | 16:33 | |
| andrewbogott | well, that was fast | 16:33 |
| *** markmcclain has quit IRC | 16:36 | |
| *** dachary has quit IRC | 16:36 | |
| *** fnaval has quit IRC | 16:39 | |
| *** eglynn has quit IRC | 16:40 | |
| *** anniec has joined #openstack-dev | 16:41 | |
| *** anniec has quit IRC | 16:42 | |
| *** anniec has joined #openstack-dev | 16:42 | |
| *** anderstj has joined #openstack-dev | 16:47 | |
| *** salgado-lunch is now known as salgado | 16:48 | |
| termie | ayoung: migrations are in keystone/common/sql/migrate_repo/versions | 16:50 |
| termie | ayoung: here is an example of a migration from nova (uses the same system) | 16:50 |
| termie | ayoung: https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/migrate_repo/versions/089_add_volume_id_mappings.py | 16:50 |
| *** fnaval has joined #openstack-dev | 16:54 | |
| *** nati_ueno has joined #openstack-dev | 16:55 | |
| *** corXi has quit IRC | 16:56 | |
| *** dachary has joined #openstack-dev | 16:59 | |
| heckj | termie: thanks | 16:59 |
| *** sandywalsh has joined #openstack-dev | 17:01 | |
| *** eglynn has joined #openstack-dev | 17:01 | |
| *** reed has joined #openstack-dev | 17:01 | |
| *** darraghb has quit IRC | 17:02 | |
| jgriffith | ayoung: if you have any questions on anything in there let me know :) | 17:02 |
| *** harlowja has joined #openstack-dev | 17:03 | |
| *** _david_ has joined #openstack-dev | 17:03 | |
| harlowja | qq, so there is a new email list for openstack developers right | 17:03 |
| *** thingee has joined #openstack-dev | 17:04 | |
| *** derekh has joined #openstack-dev | 17:04 | |
| *** andrewsmedina has joined #openstack-dev | 17:08 | |
| *** _david_ has left #openstack-dev | 17:10 | |
| *** cp16net is now known as cp16net|away | 17:15 | |
| *** Mandell has joined #openstack-dev | 17:21 | |
| mtaylor | vishy: around? | 17:21 |
| ayoung | jgriffith mine should be two lines. alter table token change id `key` varchar (1024); alter table token add id varchar (2048); | 17:22 |
| mtaylor | vishy: I think I have all of the bits needed for jenkins in https://review.openstack.org/#/c/9093/ ... but it fails unittests even when running in run_tests.sh -V | 17:22 |
| ayoung | although I am converting key into id_hash just to make sure we avoid conflicts with the sql keyword | 17:23 |
| mtaylor | vishy: which makes me think that perhaps an upstream version increased and we're not pinning it down properly? | 17:23 |
| mtaylor | vishy: could you (or someone else who knows the codebase a bit better) grab 9093 and see if you can get unittests to run for it via run_tests.sh ? | 17:23 |
| jgriffith | ayoung: hehe... well guess that's pretty straight forward :) | 17:23 |
| ayoung | jgriffith, do I need to worry about different DB types? | 17:24 |
| ayoung | I think that is pretty straightforward | 17:24 |
| jgriffith | ayoung: Do you have foreign keys? | 17:24 |
| ayoung | nope | 17:24 |
| jgriffith | ayoung: and are you changing/modifying any primary keys? | 17:24 |
| ayoung | yes | 17:24 |
| ayoung | currently id is primary key | 17:25 |
| ayoung | changing it to id_hash | 17:25 |
| jgriffith | ayoung: modifying primary keys you'll likely have to have special cases for psql | 17:25 |
| ayoung | and adding a new column named id that is not primary key | 17:25 |
| *** torgomatic has joined #openstack-dev | 17:25 | |
| *** mnewby has joined #openstack-dev | 17:25 | |
| ayoung | jgriffith, it should not be auto incremente, though | 17:25 |
| jgriffith | ayoung: Ahhh... so it is very much like the uuid change | 17:25 |
| ayoung | do I need to drop the index? | 17:26 |
| jgriffith | ayoung: Take a look at how I did the volume id change, it's going to be almost exactly the same | 17:26 |
| ayoung | jgriffith, I thought a rename maintained the data? | 17:27 |
| jgriffith | ayoung: My understanding is that doesn't work for psql | 17:27 |
| ayoung | jgriffith, ok...I can test. Is there psql support in devstack yet? | 17:28 |
| jgriffith | ayoung: I broke it into two migrations 89 and 90 | 17:28 |
| jgriffith | ayoung: That I do not know | 17:28 |
| jgriffith | ayoung: But sqlite gets upset too as I recall | 17:28 |
| ayoung | I almost tempted to say that for psql we just truncate...it is only ephemeral tokens anyway | 17:28 |
| ayoung | hmmm...I wonder what will happen with memcached.... | 17:29 |
| jgriffith | ayoung: slightly different reasons on sqlite, but it doesn't like mucking with data types on primaries | 17:29 |
| jgriffith | ayoung: So just look at the 090 versions, the .py and the .sql and it should do exactly the same thing you're working on. | 17:31 |
| jgriffith | ayoung: Just cut out the extra cruft and look at the volume['id'] alone | 17:31 |
| bcwaldon | mtaylor: ring ring | 17:31 |
| jgriffith | ayoung: The rest is all just really the same thing for the other tables | 17:31 |
| *** cp16net|away is now known as cp16net | 17:32 | |
| jgriffith | ayoung: it's a pita but it works :) | 17:33 |
| *** eglynn has quit IRC | 17:33 | |
| ayoung | jgriffith, OK...so for Keystone, it doesn't look like we have any of those yet. I am guessing that I need to add: keystone/common/sql/migrate_repo/versions/002_pki_tokens.py | 17:33 |
| ayoung | do you mean to say that I also need | 17:33 |
| ayoung | keystone/common/sql/migrate_repo/versions/002_pki_tokens.sql | 17:33 |
| jgriffith | ayoung: yes, particularly for the sqlite case | 17:34 |
| *** sandywalsh has quit IRC | 17:36 | |
| *** PotHix has joined #openstack-dev | 17:37 | |
| jgriffith | ayoung: Oh... I see this is the first keystone migration other than the initial :) | 17:37 |
| ayoung | jgriffith, I don't see one there for psql. Is that dealt with in the .py file, or did we just not support psql back then? | 17:37 |
| *** lcheng has joined #openstack-dev | 17:37 | |
| jgriffith | ayoung: the .sql file will handle it | 17:37 |
| ayoung | ? | 17:37 |
| jgriffith | So you'll notice on some of the migrations in nova, you have a .py version and a .sql version | 17:38 |
| ayoung | so the sqlite*,sql files also handle psql? | 17:38 |
| jgriffith | ayoung: correct | 17:38 |
| ayoung | that is intuitive | 17:38 |
| jgriffith | ayoung: Well... there is that :) | 17:38 |
| ayoung | I'd almost rather drop the .py file and only use the SQL if it could deal with all of the DBMS | 17:39 |
| jgriffith | ayoung: You'll have to take that up with heckj :), it's not so bad really | 17:40 |
| ameade | names | 17:41 |
| ameade | bah | 17:41 |
| *** vanchester has quit IRC | 17:41 | |
| *** Ryan_Lane has joined #openstack-dev | 17:45 | |
| ameade | comstud, johannes, vishy: https://review.openstack.org/#/c/9450/ | 17:47 |
| comstud | lookin | 17:47 |
| *** markmcclain has joined #openstack-dev | 17:48 | |
| *** jakedahn_zz is now known as jakedahn | 17:48 | |
| *** matwood has quit IRC | 17:48 | |
| mtaylor | bcwaldon: pong! | 17:49 |
| bcwaldon | mtaylor: I forget! | 17:49 |
| bcwaldon | mtaylor: damn | 17:49 |
| mtaylor | bcwaldon: were you going to tell me that you were going to review my three nova patches related to setup and then fix the nova patch for stable/diablo so that it can pass unittests again? | 17:50 |
| bcwaldon | mtaylor: nope | 17:50 |
| bcwaldon | mtaylor: that'll have to wait until later today | 17:50 |
| mtaylor | bcwaldon: meh | 17:53 |
| mtaylor | bcwaldon: well, honestly, I'm stumped on the stable/diablo patch - I've got the jenkins build stuff added and theoretically working, but the tree doesn't pass unittests | 17:53 |
| mtaylor | which I'm guessing is a dependent library version bump, because nothing I did should have broken run_tests.sh -V at all | 17:53 |
| bcwaldon | mtaylor: I can help later, I've gotta run now | 17:54 |
| *** sandywalsh has joined #openstack-dev | 17:55 | |
| *** adjohn has joined #openstack-dev | 17:58 | |
| mtaylor | bcwaldon: awesome | 18:00 |
| *** vanchester has joined #openstack-dev | 18:01 | |
| *** jakedahn is now known as jakedahn_zz | 18:03 | |
| *** anniec has quit IRC | 18:04 | |
| *** eglynn has joined #openstack-dev | 18:04 | |
| *** anniec_ has joined #openstack-dev | 18:04 | |
| *** dachary has quit IRC | 18:04 | |
| *** novas0x2a|laptop has quit IRC | 18:05 | |
| *** dachary has joined #openstack-dev | 18:06 | |
| *** blamar has quit IRC | 18:09 | |
| sdague | any nova core folks around to land - https://review.openstack.org/#/c/9410/ and https://review.openstack.org/#/c/9411/ ? | 18:12 |
| *** zykes- has quit IRC | 18:17 | |
| *** novas0x2a|laptop has joined #openstack-dev | 18:24 | |
| *** andrewbogott has quit IRC | 18:24 | |
| *** blamar has joined #openstack-dev | 18:26 | |
| *** dubsquared has joined #openstack-dev | 18:38 | |
| *** johnpostlethwait has joined #openstack-dev | 18:41 | |
| *** johngarbutt has quit IRC | 18:42 | |
| *** mrunge has joined #openstack-dev | 18:45 | |
| *** fnaval has quit IRC | 18:47 | |
| *** blamar has quit IRC | 18:48 | |
| *** vanchester has joined #openstack-dev | 18:49 | |
| *** japage has joined #openstack-dev | 18:49 | |
| *** dubsquared has quit IRC | 18:50 | |
| *** eglynn has quit IRC | 18:58 | |
| mtaylor | anybody from swift around who can merge this: https://review.openstack.org/#/c/9253/ ? | 18:58 |
| *** andrewbogott has joined #openstack-dev | 19:01 | |
| *** andrewbogott has joined #openstack-dev | 19:01 | |
| *** fnaval has joined #openstack-dev | 19:01 | |
| *** derekh has quit IRC | 19:05 | |
| *** derekh has joined #openstack-dev | 19:05 | |
| *** dprince has quit IRC | 19:12 | |
| *** Aaton_off is now known as Aaton | 19:14 | |
| *** anniec_ has quit IRC | 19:14 | |
| *** derekh has quit IRC | 19:17 | |
| *** derekh has joined #openstack-dev | 19:17 | |
| vanchester | aha, there it is! swift-client! | 19:18 |
| *** eglynn has joined #openstack-dev | 19:19 | |
| *** arosen has joined #openstack-dev | 19:20 | |
| *** markmcclain has quit IRC | 19:21 | |
| *** openstackgerrit has quit IRC | 19:25 | |
| *** openstackgerrit has joined #openstack-dev | 19:25 | |
| *** edygarcia_ has joined #openstack-dev | 19:29 | |
| *** edygarcia has quit IRC | 19:32 | |
| *** edygarcia_ is now known as edygarcia | 19:32 | |
| *** markmcclain has joined #openstack-dev | 19:38 | |
| *** markmcclain has quit IRC | 19:38 | |
| *** markmcclain has joined #openstack-dev | 19:38 | |
| *** csolallo has joined #openstack-dev | 19:40 | |
| *** csolallo has left #openstack-dev | 19:41 | |
| *** zul has quit IRC | 19:44 | |
| *** fc__ has quit IRC | 19:51 | |
| *** blamar has joined #openstack-dev | 19:52 | |
| *** apevec has quit IRC | 19:52 | |
| *** harlowja has left #openstack-dev | 20:00 | |
| *** eglynn has quit IRC | 20:00 | |
| *** harlowja_ has joined #openstack-dev | 20:04 | |
| *** markmcclain has quit IRC | 20:04 | |
| *** anniec has joined #openstack-dev | 20:04 | |
| *** markmcclain has joined #openstack-dev | 20:05 | |
| *** zykes- has joined #openstack-dev | 20:06 | |
| *** rods has quit IRC | 20:10 | |
| *** heckj has quit IRC | 20:10 | |
| *** milner has quit IRC | 20:12 | |
| *** arosen has quit IRC | 20:14 | |
| *** reed has quit IRC | 20:14 | |
| *** vanchester has quit IRC | 20:16 | |
| *** blamar has quit IRC | 20:17 | |
| *** blamar has joined #openstack-dev | 20:17 | |
| *** reed has joined #openstack-dev | 20:21 | |
| jtran | dtroyer, ping | 20:23 |
| dtroyer | jtran: yo | 20:23 |
| jtran | hey w/ devstack's stack.sh i have to unset SERVICE_TOKEN SERVICE_PASSWORD right before the glance commands... | 20:24 |
| jtran | is this a known issue ? | 20:24 |
| jtran | i'm wondering why i'm the only one that has to do that | 20:25 |
| mtaylor | jtran: it's because you didn't pay your money to the mafia | 20:26 |
| jtran | i knew i forgot something | 20:26 |
| dtroyer | jtran: not sure, there were some changes recently to how the token for the glance commands is aquired to use the environment variables defined when keystone is set up. | 20:28 |
| dtroyer | jtran: how current is your devstack checkout? | 20:29 |
| jtran | just updated today | 20:29 |
| jtran | FTR i've always had thsi problem w/ devstack (months) | 20:29 |
| jtran | i'll try it again today see if it is fixed | 20:29 |
| dtroyer | do you define both SERVICE_TOKEN and SERVICE_PASSWORD in localrc? | 20:30 |
| jtran | yes | 20:30 |
| *** rods has joined #openstack-dev | 20:31 | |
| *** mdomsch has quit IRC | 20:33 | |
| dtroyer | SERVICE_PASSWORD shouldn't better for glance cli, it is only used to authenticate the services to keystone... | 20:33 |
| *** steveb___ has joined #openstack-dev | 20:33 | |
| dtroyer | s/better/matter/ | 20:33 |
| jtran | dtroyer: yes you're right now that i look at it again, it is just prior to the keystone cmds. it was the glance cmds that later compalin cuz $TOKEN doesn't exist due to keystone cmds not successful | 20:34 |
| jtran | i just updated stack.sh and it does look a lot diff. trying it now.. | 20:35 |
| dtroyer | jtran: right after keystone_data.sh runs the 4 OS_* env vars are exported, this was the bit required to fix keystone cli. If you don't have that you would see something like this | 20:35 |
| *** fnaval has quit IRC | 20:35 | |
| dtroyer | ah, ok | 20:35 |
| *** matwood has joined #openstack-dev | 20:38 | |
| *** issackelly has quit IRC | 20:39 | |
| *** issackelly has joined #openstack-dev | 20:40 | |
| *** sandywalsh has quit IRC | 20:42 | |
| *** alaski has quit IRC | 20:44 | |
| *** steveb___ has quit IRC | 20:45 | |
| jtran | dtroyer, i was still having the problem, i created bug and issued a patch https://bugs.launchpad.net/devstack/+bug/1018581 | 20:46 |
| uvirtbot | Launchpad bug 1018581 in devstack "stack.sh fails on keystone command needs unset SERVICE_TOKEN & SERVICE_PASSWORD" [Undecided,New] | 20:46 |
| jtran | https://review.openstack.org/#/c/9462/ | 20:47 |
| *** littleidea has quit IRC | 20:49 | |
| *** fnaval has joined #openstack-dev | 20:50 | |
| *** simo has joined #openstack-dev | 20:50 | |
| dtroyer | jtran: can you post a bit more detail to the bug, including your localrc with passwords redacted? | 20:53 |
| jtran | absolutely | 20:53 |
| jtran | i just posted my localrc non-redacted in the bug page | 20:54 |
| jtran | i'm working on getting more info from the err for you | 20:54 |
| *** simo has left #openstack-dev | 20:54 | |
| *** rbasak has quit IRC | 20:55 | |
| ayoung | derekh, have a moment to talk https://review.openstack.org/#/c/7437/5 ? | 21:00 |
| *** EmilienM has left #openstack-dev | 21:00 | |
| *** littleidea has joined #openstack-dev | 21:00 | |
| jtran | dtroyer, uploaded logfile as well | 21:01 |
| dtroyer | jtran: thx | 21:02 |
| *** dubsquared has joined #openstack-dev | 21:02 | |
| andrewbogott | dellmann, russellb: If you're both here, can we take a minute to hash out the relative merits of https://review.openstack.org/#/c/9075/ vs. https://review.openstack.org/#/c/9316/? (Or perhaps there's a different way to implement 9316 which is clearer?) | 21:03 |
| *** anniec has quit IRC | 21:07 | |
| *** derekh has quit IRC | 21:07 | |
| *** derekh has joined #openstack-dev | 21:07 | |
| *** japage has quit IRC | 21:13 | |
| derekh | ayoung: yup | 21:14 |
| *** dachary has quit IRC | 21:16 | |
| *** novas0x2a|laptop has quit IRC | 21:17 | |
| *** novas0x2a|laptop has joined #openstack-dev | 21:17 | |
| *** anniec has joined #openstack-dev | 21:19 | |
| ayoung | derekh, do you think that the URL segment /OS-KSCRUD/ is appropriate? | 21:22 |
| ayoung | seems to me that it is not really necessary | 21:23 |
| ayoung | should be | 21:23 |
| *** roge has quit IRC | 21:23 | |
| ayoung | PUT /users/{user_id}/OS-KSCRUD/password' | 21:23 |
| derekh | the admin api uses /OS-ADM/ so I simply changed it, but I'm willing to remove it if people thinks its more appropriate | 21:24 |
| *** lts has quit IRC | 21:25 | |
| derekh | ayoung: correction */OS-KSADM/ | 21:25 |
| ayoung | derekh, is there any technical reason to have anything there? | 21:25 |
| derekh | none that I know of | 21:26 |
| ayoung | like a certain router will own everything under CRUD... | 21:26 |
| ayoung | and taking it out just removes that segment of the string, right? | 21:26 |
| ayoung | no other side effects> | 21:26 |
| ayoung | ? | 21:26 |
| *** roge has joined #openstack-dev | 21:26 | |
| derekh | correct, I'll validate that but I'm pretty sure there would be no side effects | 21:26 |
| *** andrewbogott has quit IRC | 21:30 | |
| ayoung | OK...I'll run it once you get a new patch up there. | 21:35 |
| *** johnpostlethwait has quit IRC | 21:37 | |
| *** johnpostlethwait has joined #openstack-dev | 21:38 | |
| *** jakedahn_zz is now known as jakedahn | 21:39 | |
| *** littleidea has quit IRC | 21:43 | |
| derekh | ok | 21:43 |
| *** mrunge has quit IRC | 21:43 | |
| *** littleidea has joined #openstack-dev | 21:46 | |
| *** mdomsch has joined #openstack-dev | 21:47 | |
| *** Gordonz has quit IRC | 21:49 | |
| *** andrewbogott has joined #openstack-dev | 21:50 | |
| *** andrewbogott has joined #openstack-dev | 21:50 | |
| *** jtran has quit IRC | 21:53 | |
| *** sleepsonthefloor is now known as sleepsonzzz | 21:53 | |
| *** anniec has quit IRC | 21:56 | |
| *** Dr_Who has quit IRC | 22:02 | |
| *** anniec has joined #openstack-dev | 22:04 | |
| *** anniec_ has joined #openstack-dev | 22:05 | |
| derekh | ayoung: new version submitted | 22:06 |
| ayoung | derekh, thanks, I'll test soonish...think I'm getting dragged off to dinner here now... | 22:06 |
| *** bencherian has joined #openstack-dev | 22:07 | |
| derekh | ayoung: ok cool, no hurry | 22:07 |
| *** derekh has quit IRC | 22:07 | |
| *** mnewby has quit IRC | 22:07 | |
| *** anniec has quit IRC | 22:09 | |
| *** anniec_ is now known as anniec | 22:09 | |
| dragondm | is there a way to ping jenkins to re-run a check on a gerrit patch? It seems to have hiccuped on a pachset and came back with 'unstable' and o tests run :P I suspect a glitch. | 22:10 |
| *** camm has quit IRC | 22:12 | |
| *** thingee has quit IRC | 22:20 | |
| *** torgomatic has quit IRC | 22:28 | |
| *** torgomatic_ has joined #openstack-dev | 22:28 | |
| *** mdomsch has quit IRC | 22:32 | |
| *** fnaval has quit IRC | 22:34 | |
| *** johnpostlethwait has quit IRC | 22:35 | |
| *** johnpostlethwait has joined #openstack-dev | 22:37 | |
| *** fnaval has joined #openstack-dev | 22:38 | |
| *** sacharya has quit IRC | 22:42 | |
| *** dtroyer is now known as dtroyer_zzz | 22:46 | |
| *** andrewsmedina has quit IRC | 22:50 | |
| *** datsun180b has quit IRC | 22:50 | |
| jeblair | dragondm: yes: http://wiki.openstack.org/GerritJenkinsGithub#Gerrit.2C_Jenkins.2C_and_Testing | 22:53 |
| jeblair | dragondm: what change are you looking at? | 22:53 |
| dragondm | https://review.openstack.org/#/c/9455/ | 22:54 |
| *** nati_uen_ has joined #openstack-dev | 22:54 | |
| dragondm | but I am no looking at odd behavior in the testrunner. | 22:55 |
| dragondm | s/no/now/ | 22:55 |
| *** nati_ueno has quit IRC | 22:58 | |
| *** ncode has quit IRC | 23:03 | |
| *** rnirmal has quit IRC | 23:07 | |
| *** nati_uen_ has quit IRC | 23:09 | |
| *** nati_ueno has joined #openstack-dev | 23:09 | |
| *** lloydde has quit IRC | 23:10 | |
| *** eglynn has joined #openstack-dev | 23:10 | |
| *** anniec has quit IRC | 23:12 | |
| *** johnpostlethwait has quit IRC | 23:16 | |
| *** eglynn has quit IRC | 23:19 | |
| *** dtroyer_zzz is now known as dtroyer | 23:20 | |
| *** thingee has joined #openstack-dev | 23:21 | |
| *** ncode has joined #openstack-dev | 23:22 | |
| *** ncode has quit IRC | 23:26 | |
| *** nati_uen_ has joined #openstack-dev | 23:29 | |
| *** nati_ueno has quit IRC | 23:33 | |
| *** lloydde has joined #openstack-dev | 23:35 | |
| *** EmilienM has joined #openstack-dev | 23:43 | |
| *** EmilienM has left #openstack-dev | 23:44 | |
| *** edygarcia has quit IRC | 23:46 | |
| *** issackelly has quit IRC | 23:50 | |
| *** fnaval has quit IRC | 23:50 | |
| *** andrewbogott has quit IRC | 23:51 | |
| *** edygarcia has joined #openstack-dev | 23:51 | |
| *** johnpostlethwait has joined #openstack-dev | 23:55 | |
| *** edygarcia has quit IRC | 23:56 | |
| *** johnpostlethwait has quit IRC | 23:56 | |
| *** edygarcia has joined #openstack-dev | 23:56 | |
| *** fnaval has joined #openstack-dev | 23:57 | |
| *** jgriffit2 has joined #openstack-dev | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!