Tuesday, 2015-04-07

elmikoredrobot: no prob, glad to help =)00:00
*** nkinder has joined #openstack-barbican00:09
*** SheenaG has quit IRC00:10
jamielennoxhi, what's the status of https://review.openstack.org/#/c/169481/2 ?00:32
jamielennoxdoes that mean there is another review replacing it or it's no longer required..?00:33
*** SheenaG has joined #openstack-barbican01:36
openstackgerritJohn Vrbanac proposed openstack/barbican: Adding simple log.info's to the SecretController  https://review.openstack.org/16877401:45
*** SheenaG has quit IRC01:53
woodster_jamielennox: we just talked about that one today, I don't think it's being replaced if that's what you mean?01:54
jamielennoxwoodster_: i was just looking at https://review.openstack.org/#/c/148450/01:55
jamielennoxwoodster_: and the last comment refers to it - but it's been -1ed after a +2, i was just wondering what was going on01:55
woodster_jamielennox: it looks like that last gate is borking and Thomas' CR might fix that. After that your CR should be all green?02:00
jamielennoxwoodster_: not sure - i was just reading the comment02:00
woodster_jamielennox: my read is that it could fix that broken gate. The issue is that the API is returning new fields now (for order sub-status) that broke the client.02:02
jamielennoxwoodster_: that's ok, i was just circling back to it, i can leave it for another few days i was just wondering if it was worth rebasing onto an existing fix or something02:05
woodster_jamielennox: we'll be trying to put some heat behind these CRs over the next couple of days, but if you are in a hurry you could certainly add those changes to your CR (and the one Steve mentions at the end of this CR: https://review.openstack.org/#/c/169481/2/barbicanclient/barbican_cli/orders.py,cm)02:16
woodster_jamielennox: fwiw, we talked about making the client less sensitive to new optional fields added to responses from the API so that these new fields don't break the client everytie02:19
openstackgerritDave McCowan proposed openstack/barbican: Refactor Stored Key Certificate Order Validator Code  https://review.openstack.org/17102302:19
jamielennoxwoodster_: would make sense, client shouldn't be tied to only working with certain server versions02:20
woodster_jamielennox: yeah we had a big discussion about that this afternoon. If feels 'dirty' adding a **kwargs at the end of the Order __init__() though...or plucking values out of the kwargs dict directly. :)02:21
jamielennoxhaving done the kwargs thing i'd hold off as long as possible02:22
woodster_jamielennox: the action items was to see what other projects do to handle this one, or the API workig group02:22
woodster_jamielennox: I was in the **kwargs camp if the alternative to is to release more versions of the client though the API URI version isn't changing.02:23
jamielennoxwoodster_: why is the client checking the server side info at all?02:24
jamielennoxyou can add this stuff to get() and list() etc02:24
jamielennoxor if you are talking about creating the object, i don't know about reconstructing from kwargs, we normally pass the response from the response as a dict and you should only provide accessors to things that the client knows about02:25
woodster_jamielennox: the response is coming back as a dict that is then passed to the Order model which has specific keyword args with no ending **kwargs to 'soak up' the remaining 'new' dict items that might be in the response.02:26
jamielennoxwoodster_: hmm, ok - not sure which way you want to go, just showing how we deal with it02:28
jamielennoxand the client should be able to handle these new options02:29
woodster_jamielennox: so you mean the dict is passed in to the __init__ and then the constructor does '.get('my_value')' sort of calls to it to set its 'self' attributes, correct?02:29
jamielennoxwoodster_: more or less, they tried to standardize this a while ago, i'll find it02:30
jamielennoxwoodster_: https://github.com/openstack/oslo-incubator/blob/master/openstack/common/apiclient/base.py#L44002:31
woodster_jamielennox: which project are you referring to? The API WG?02:31
jamielennoxwoodster_: that was an older one, the way we tried to do it in sdk you still pass the dictionary, __getitem__ gives you access to the whole resource response so that if you need new values before the client you can get them02:32
woodster_jamielennox: Nice!02:32
jamielennoxand attributes are provided for objects that the client knows about02:33
jamielennoxthat oslo-incubator module is deprecated, but there are plenty of clients that are using more or less that02:34
jamielennoxthis is a way better way but it's a big file https://github.com/stackforge/python-openstacksdk/blob/master/openstack/resource.py02:35
woodster_jamielennox: this looks good. I need to spin up on what it takes to get barbican into the openstack sdk. I think we'll be discussing that more in liberty.02:41
dave-mccowanwoodster_  https://review.openstack.org/171023  this is the code moving the validator function out of orders.py, per your code review comment. what do you think?02:42
briancurtinwoodster_: there's already some minimal support for it, just need to rename it from keystore to whatever you moved to (can't remember)02:42
jamielennoxwoodster_: yea, that's long term a good idea, but the clients will stick around for a while yet anyway02:42
*** chlong has quit IRC02:42
woodster_dave-mccowan: nice, I'll take a look02:42
jamielennoxbriancurtin: that resource file is so much bigger than i remember, i need to get back in there02:43
*** chlong has joined #openstack-barbican02:44
woodster_briancurtin: ha, I was thinking work wouldn't start until after the Liberty summit. So maybe we just fret over the current client we have until Kilo final release and then replace it?02:45
briancurtinwoodster_: i dont work on barbican so i have no input there, but once we in SDK wrap up some current things, we could have full support of the REST API of pretty much any service very quickly. we did a small amount of a bunch of services just to see how things look, and will push more for completeness probably starting in a few weeks02:46
woodster_jamielennox: frameworks can be that way I figure :)02:46
woodster_briancurtin: nice, so meeting up in Liberty will be good timing then it sounds like02:47
briancurtini believe so02:47
openstackgerritJohn Vrbanac proposed openstack/barbican: Updating hacking version for py3pep8 tox job  https://review.openstack.org/17102703:14
openstackgerritJohn Vrbanac proposed openstack/barbican: Fixing python 3 imcompatiblity in common.utils  https://review.openstack.org/17103103:20
*** jamielennox is now known as jamielennox|away03:47
*** gyee has quit IRC03:50
*** kebray has joined #openstack-barbican04:00
*** dave-mccowan has quit IRC04:08
*** kebray has quit IRC04:44
*** kebray has joined #openstack-barbican04:51
rm_workwoodster_: ah i hope you had good luck tonight -- got disconnected, so may have missed if you did ping me05:46
openstackgerritOpenStack Proposal Bot proposed openstack/barbican: Imported Translations from Transifex  https://review.openstack.org/17105506:11
*** woodster_ has quit IRC06:50
*** everjeje has joined #openstack-barbican06:57
*** tkelsey has joined #openstack-barbican07:28
*** tkelsey has quit IRC07:30
*** chlong has quit IRC07:34
*** kebray has quit IRC07:46
openstackgerritThomas Herve proposed openstack/python-barbicanclient: Fix order listing on the command line.  https://review.openstack.org/16948107:56
*** jaosorior has joined #openstack-barbican08:05
*** gitorres1 has joined #openstack-barbican09:55
*** gitorres has quit IRC09:55
*** chlong has joined #openstack-barbican10:07
*** darrenmoffat has quit IRC10:20
*** darrenmoffat has joined #openstack-barbican10:21
jaosorioranybody knows what version of openssl is being used in the gate?11:39
*** chlong has quit IRC11:49
*** chlong has joined #openstack-barbican11:51
*** woodster_ has joined #openstack-barbican12:10
*** dave-mccowan has joined #openstack-barbican12:21
*** alee has quit IRC12:26
*** rellerreller has joined #openstack-barbican12:41
openstackgerritMerged openstack/barbican: Updating hacking version for py3pep8 tox job  https://review.openstack.org/17102712:55
openstackgerritJohn Wood proposed openstack/barbican: Add retry server and functional tests to DevStack  https://review.openstack.org/17089612:58
openstackgerritDave McCowan proposed openstack/barbican: Refactor Stored Key Certificate Order Validator Code  https://review.openstack.org/17102313:02
*** openstackgerrit has quit IRC13:07
*** SheenaG has joined #openstack-barbican13:07
*** openstackgerrit has joined #openstack-barbican13:07
rellerrellerwoodster_ redrobot Are people discussing content type issues today? When I logged off yesterday there was talk of a Google hangout.13:13
woodster_rellerreller I think the idea was to meet in IRC after the weekly meeting, but I think you had to cut out shortly thereafter. A google hangouts today would work for me though. What time works for you?13:20
rellerrellerwoodster_ All times in CT, I am busy from 9:00-9:30 and after 12:00. Basically anytime before noon except 9:00-9:30 is good with me.13:24
woodster_so effectively now thru the next 2 and 1/2 hours?13:25
woodster_rellerreller,  sorry ^^^13:26
woodster_redrobot, are you there?13:26
woodster_rellerreller, I don't see alee either. Maybe give them another hour13:27
*** alee has joined #openstack-barbican13:40
dave-mccowanwoodster_, rellerreller,  yesterday redrobot said he was going to spend today sorting out and documenting all the flows and formats, then find a slot on Wednesday for a hangout.13:44
*** SheenaG has quit IRC13:45
*** chlong has quit IRC13:51
aleejaosorior, I would guess its 0.9.8  something , based on the failures I saw before.13:58
jaosorioralee: I didn't post an update on that question. But from what I saw it seems it's 1.0.1f13:59
aleejaosorior, ok14:00
jaosorioralee:  1.0.1f-1ubuntu2.11 to be specific14:00
jaosoriorI'm getting errors on some tests from the snakeoil CA14:00
jaosoriorand I'm trying to understand what's going on14:01
*** kfarr has joined #openstack-barbican14:01
*** zz_dimtruck is now known as dimtruck14:03
aleejaosorior, ok14:03
*** paul_glass has joined #openstack-barbican14:04
*** paul_glass has quit IRC14:14
*** paul_glass has joined #openstack-barbican14:15
*** igueths has joined #openstack-barbican14:23
iguethsHi all.14:23
woodster_igueths, morning14:25
iguethsMorning woodster_.14:28
*** SheenaG has joined #openstack-barbican14:32
*** kebray has joined #openstack-barbican14:49
*** atiwari has joined #openstack-barbican14:56
*** kfarr has quit IRC15:02
*** SheenaG has quit IRC15:04
*** xaeth_afk is now known as xaeth15:08
*** SheenaG has joined #openstack-barbican15:43
jvrbanacjaosorior, yo!15:43
jvrbanacjaosorior, sooo I went and looked at Nova and Keystone and it looks like LOG.infos are used very sparingly, but they are used in the same fashion for what I did in that CR. They only use LOG.infos to report discrete positive units of work within the controller call directly. I'm inclined to follow that pattern for now.15:52
*** gyee has joined #openstack-barbican15:54
jaosoriorjvrbanac: alright. What's the link to the CR?16:07
jvrbanacjaosorior, https://review.openstack.org/#/c/168774/16:15
jvrbanacjaosorior, I think this is a good sidebar discussion for us to have over Liberty, but currently, we have zero logging that is acceptable for production cases, so I would like to atleast get our minimum unit-of-work messages in and then iterate.16:17
*** SheenaG1 has joined #openstack-barbican16:21
*** SheenaG has quit IRC16:22
openstackgerritMerged openstack/barbican: Fixing python 3 imcompatiblity in common.utils  https://review.openstack.org/17103116:26
elmikojvrbanac: i've been watching that review (168774), would it be worthwhile to add similar logging to the containers controller?16:41
*** alee is now known as alee_lunch16:49
*** tkelsey has joined #openstack-barbican16:49
*** dave-mccowan has quit IRC16:58
*** rellerreller has quit IRC16:58
*** dave-mccowan has joined #openstack-barbican17:26
jaosoriorjvrbanac: I agree. And we can refactor it later if we find another solution that we prefer17:46
jaosoriorjvrbanac: I agree. And we can refactor it later if we find another solution that we prefer17:48
*** alee_lunch is now known as alee17:49
*** tkelsey has quit IRC18:00
*** jamielennox|away is now known as jamielennox18:01
jvrbanacjaosorior, thx18:19
*** atiwari has quit IRC18:19
jvrbanacelmiko, yeah I agree. I didn't want to start adding logging to other things until we had agreed on a solution for the time being.18:20
elmikojvrbanac: yea same, sounds like i could make a patch for it?18:22
jvrbanacelmiko, I believe so. Btw, here is the Openstack logging guideline if you didn't already have it: http://specs.openstack.org/openstack/openstack-specs/specs/log-guidelines.html18:24
elmikojvrbanac: awesome, thanks!18:24
openstackgerritAde Lee proposed openstack/barbican: Changes to get remaining cert functional tests working  https://review.openstack.org/16960018:53
aleewoodster_, redrobot , dave-mccowan ^^18:54
aleerellerreller, kfarr ^^18:54
aleewoodster_, redrobot , dave-mccowan - this gets all the cert order functional tests working and gets things working as the current code is envisioned to work.18:55
redrobotalee sounds good18:55
redrobotalee will hopefully get to it sometime soon18:55
aleewoodster_, redrobot , dave-mccowan - if we decide we want to change how things work, we can do that in a separate cr.18:56
aleebut this will at least give us a framework and tests for what needs to work in case we change anything18:56
aleeredrobot, yes - I think this will be instructive for you as you research how we manage content type.18:57
aleeits in all the code you need to examine.18:57
aleejvrbanac, jaosorior ^^ please take a look as well.18:58
aleedave-mccowan, going to try out your tests now to see if they pass ..18:58
dave-mccowanalee awesome.  looking at your CR now.18:59
aleereaperhulk, we're probably going to need some input from you too.19:00
aleedave-mccowan, I don't think your test cases are valid -- I'll comment on the CR.19:10
*** bpoulos has joined #openstack-barbican19:13
dave-mccowanalee quite possible that they need to be debugged now there is working code underneath.19:15
aleedave-mccowan, yup19:15
aleedave-mccowan, I'd wait till redrobot figures out what we want to do and we have the google hangout tomorrow.19:16
aleeredrobot, have you scheduled the hangout yet?19:16
dave-mccowanalee, i hacked them a couple different ways trying to work around the underlying code; i don't remember which state I left them in. :-)  but please, comment on CR.19:17
redrobotalee not yet19:17
aleedave-mccowan, I'll put in some pointers on the CR and maybe you can try them ..19:17
redrobottrying to figure out if we even need one19:17
redrobotI think the only thing I don't like currently is that there is no content type for a public key19:18
redrobotso we're using application/octet-stream for a public key19:19
aleeredrobot, I think what we are currently doing is pretty complicated.19:21
redrobotalee I've been mostly looking at it from an API user pov19:22
aleeredrobot, so its worth reviewing.  check out my cr to see what needs to be done to get things working19:22
aleeredrobot, sure - lets start from there.19:22
*** openstackgerrit has quit IRC19:22
*** openstackgerrit has joined #openstack-barbican19:23
openstackgerritMerged openstack/barbican: Imported Translations from Transifex  https://review.openstack.org/17105519:24
*** bpoulos has left #openstack-barbican19:29
*** tkelsey has joined #openstack-barbican19:46
*** tkelsey has quit IRC19:50
*** SheenaG has joined #openstack-barbican20:12
*** SheenaG1 has quit IRC20:16
*** SheenaG1 has joined #openstack-barbican20:17
*** SheenaG has quit IRC20:19
aleedave-mccowan, can you run the functional tests and see if the stored_key_test still fails?20:33
aleewoodster_, ^^20:33
aleeit still seems to be failing in the gate for some reason20:34
dave-mccowanalee ok20:39
dave-mccowanalee FAILED (SKIP=10, errors=3, failures=8)20:45
aleedave-mccowan, phooey20:45
aleedave-mccowan, which ones failed?20:45
dave-mccowanalee, functional.test_certificate_orders.20:46
aleedave-mccowan, thats a lot of failures -- more even than in the gate20:47
aleedave-mccowan, can you post the results?20:48
aleedave-mccowan, just confirming that you applied the patch ok and to a clean tree, because on the gate only one test fails20:50
dave-mccowanalee, git log looks ok, but i usually hose my repo whenever i try to cherry-pick.  so, i wouldn't rule that out.20:52
aleedave-mccowan, ok -- so just to confirm, lets do this ..20:53
aleegit checkout master; git pull20:53
dave-mccowanalee, my output http://ur1.ca/k4koe20:53
aleegit checkout -b alee_test; git fetch ssh://alee@review.openstack.org:29418/openstack/barbican refs/changes/00/169600/3 && git cherry-pick FETCH_HEAD20:55
aleethen reinstall your barbican server -- not sure how you do it -- I use barbican.sh install20:55
aleedave-mccowan, and run tests20:56
aleedave-mccowan, if that doesn't work, you might try blowing away the db20:56
dave-mccowanalee, my bad.  i forgot to restart my server.  standby for new results.20:58
dave-mccowanalee, trying to listen to a conference call at the same time.  I got OK (SKIP=10) this time.20:59
aleedave-mccowan, yay20:59
aleedave-mccowan, can you confirm that the stored_key case was run?20:59
dave-mccowanRan 5 tests, OK (SKIP=1)21:02
dave-mccowanalee nosetests functionaltests/api/v1/functional/test_certificate_orders.py:CertificatesTestCase.test_create_stored_key_order ran OK21:08
aleedave-mccowan, great thanks21:08
openstackgerritJohn Wood proposed openstack/barbican: Add retry server and functional tests to DevStack  https://review.openstack.org/17089621:20
*** chlong has joined #openstack-barbican21:26
*** alee has quit IRC21:31
*** openstackgerrit has quit IRC21:37
*** openstackgerrit has joined #openstack-barbican21:37
redrobotanyone ever notice this content-type on Barbican responses: ?21:42
redrobotContent-Type: application/octet-stream; charset=UTF-821:42
redrobotit doesn't seem correct that application/octet-stream would have a charset21:43
*** tkelsey has joined #openstack-barbican21:47
*** tkelsey has quit IRC21:51
openstackgerritJohn Wood proposed openstack/barbican: Add retry server and functional tests to DevStack  https://review.openstack.org/17089621:53
*** xaeth is now known as xaeth_afk22:08
*** gyee has quit IRC22:11
*** gyee has joined #openstack-barbican22:14
openstackgerritJohn Wood proposed openstack/barbican: Add retry server and functional tests to DevStack  https://review.openstack.org/17089622:21
*** paul_glass has quit IRC22:30
*** alee has joined #openstack-barbican22:32
*** jaosorior has quit IRC22:52
*** dimtruck is now known as zz_dimtruck22:56
*** dave-mccowan has quit IRC23:04
*** igueths has quit IRC23:06
openstackgerritMerged openstack/barbican: Adding simple log.info's to the SecretController  https://review.openstack.org/16877423:33
*** alee has quit IRC23:37
*** alee has joined #openstack-barbican23:50
*** kebray has quit IRC23:57

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