Wednesday, 2014-08-13

« Tuesday, 2014-08-12 Index Thursday, 2014-08-14 »
*** nkinder has joined #openstack-barbican00:03
*** bdpayne has quit IRC00:37
*** SheenaG1 has joined #openstack-barbican00:49
*** SheenaG1 has left #openstack-barbican00:50
*** bdpayne has joined #openstack-barbican00:50
*** akoneru has quit IRC00:51
*** gchardon has joined #openstack-barbican01:01
*** gchardon has quit IRC01:28
*** rm_you has quit IRC01:41
*** rm_you has joined #openstack-barbican01:48
*** bdpayne has quit IRC01:54
*** crc32 has quit IRC01:55
*** woodster_ has quit IRC02:05
*** juantwo has joined #openstack-barbican03:02
*** juantwo has quit IRC03:02
*** juantwo has joined #openstack-barbican03:03
openstackgerritArun Kant proposed a change to openstack/barbican: Adding keystone notification listener support  https://review.openstack.org/11081703:12
*** ayoung has quit IRC03:50
*** rm_you| has joined #openstack-barbican05:02
*** rm_you has quit IRC05:04
*** rm_you|wtf has joined #openstack-barbican05:10
*** rm_you| has quit IRC05:13
openstackgerritOpenStack Proposal Bot proposed a change to openstack/barbican: Imported Translations from Transifex  https://review.openstack.org/11276406:10
openstackgerritJuan Antonio Osorio Robles proposed a change to openstack/barbican: Refactor secret_store for consistency  https://review.openstack.org/11353107:17
*** jaosorior has joined #openstack-barbican07:18
*** gerchardon has quit IRC08:54
*** gyee has quit IRC10:00
*** gyee has joined #openstack-barbican10:07
*** gyee has quit IRC10:23
*** gyee has joined #openstack-barbican10:30
*** juantwo has quit IRC11:28
*** jamielennox is now known as jamielennox|away12:02
*** juantwo has joined #openstack-barbican12:12
*** juantwo has quit IRC12:13
*** juantwo has joined #openstack-barbican12:14
*** nkinder has quit IRC13:11
*** woodster_ has joined #openstack-barbican13:12
openstackgerritArun Kant proposed a change to openstack/barbican: Adding keystone notification listener support  https://review.openstack.org/11081713:21
jaosoriorHey, if any core dev has time to review the cliff framework patch: https://review.openstack.org/#/c/107587/ it would be cool to keep that going :) or maybe even merged :D13:26
*** ayoung has joined #openstack-barbican13:32
*** paul_glass has joined #openstack-barbican13:42
*** paul_glass1 has joined #openstack-barbican13:45
*** paul_glass has quit IRC13:49
*** paul_glass1 is now known as paul_glass13:53
*** nkinder has joined #openstack-barbican13:55
*** akoneru has joined #openstack-barbican13:59
jvrbanacalee, redrobot, reaperhulk, woodster_, rellerreller ^^14:11
redrobotjvrbanac jaosorior it's on my to-do list.  Unfortunately I'm not familiar with Cliff, so I have to spin up on that before I can review14:12
*** paul_glass1 has joined #openstack-barbican14:15
*** openstackgerrit has quit IRC14:16
*** openstackgerrit has joined #openstack-barbican14:18
*** paul_glass has quit IRC14:18
*** juantwo has quit IRC14:28
*** juantwo has joined #openstack-barbican14:29
*** juantwo has quit IRC14:30
*** juantwo has joined #openstack-barbican14:30
hyakuheiMorning all.14:35
*** SheenaG1 has joined #openstack-barbican15:12
*** paul_glass1 is now known as paul_glass15:15
paul_glassredrobot15:15
jvrbanachyakuhei, morning15:16
redrobothyakuhei good morning!15:16
redrobotpaul_glass sup?15:16
*** jaosorior has quit IRC15:32
*** atiwari has joined #openstack-barbican15:43
woodster_atiwari: Would you be able to address comments in this CR this week perhaps?: https://review.openstack.org/#/c/8740515:46
*** juantwo has quit IRC15:55
openstackgerritVenkat Sundaram proposed a change to openstack/python-barbicanclient: remove tenant-id from uri  https://review.openstack.org/11214916:01
*** gyee_ has joined #openstack-barbican16:02
*** juantwo has joined #openstack-barbican16:06
*** juantwo_ has joined #openstack-barbican16:07
*** juantwo has quit IRC16:10
*** SheenaG1 has quit IRC16:12
rm_workredrobot: you get a chance to look at blueprinting the client changes?16:18
aleewoodster_, ping16:24
woodster_alee: hello16:25
aleewoodster_, hey! so any further thoughts on the certificates interface?16:25
aleewoodster_, should we try and define an interface or leave it as plugin specific meta-data16:26
woodster_alee: had a discussion a SME on certificates over here, and it seems that the differences between the cert plugins and the types of certs generated would make it hard make a generic meta key/value block that covers all situations, and then let plugins decide if they support it or not.  It seems more practical to define up front I want issuer xyz (Dogtag,16:27
woodster_Symatec, etc.) and them provide key/value info appropriate for that vendor.16:27
aleewoodster_, SME?16:28
woodster_atiwari: are you there? Curious about CR https://review.openstack.org/#/c/8740516:28
woodster_alee: subject matter expert...so, our team that handles customer certs over here16:29
aleewoodster_, I tend to agree. it would take awhile to design something generic enough - whereas you could much more easily get to even complicated certs quickly with just a pass through ..16:29
*** rm_work has quit IRC16:29
woodster_heading off to a meeting for 30 min, back after that...16:30
woodster_alee: agreed16:30
aleewoodster_, ok - lets chat after that16:30
*** rm_work has joined #openstack-barbican16:30
aleeand maybe figure out how we can expose which cert plugins are available.16:30
*** paul_glass has quit IRC16:34
*** rm_work has quit IRC16:34
*** paul_glass has joined #openstack-barbican16:34
*** rm_work has joined #openstack-barbican16:35
*** SheenaG1 has joined #openstack-barbican16:35
*** rm_work has quit IRC16:35
*** rm_work has joined #openstack-barbican16:35
*** paul_glass1 has joined #openstack-barbican16:36
*** paul_glass has quit IRC16:39
*** atiwari has quit IRC16:40
*** rm_work has quit IRC16:41
*** rm_work has joined #openstack-barbican16:42
*** rm_work has quit IRC16:42
*** rm_work has joined #openstack-barbican16:42
*** paul_glass1 has quit IRC16:46
*** rm_work has quit IRC16:47
*** rm_work has joined #openstack-barbican16:47
*** juantwo_ has quit IRC16:51
*** rm_work has quit IRC16:52
*** rm_work has joined #openstack-barbican16:53
*** rm_work has quit IRC16:53
*** rm_work has joined #openstack-barbican16:53
*** gyee_ has quit IRC16:53
*** juantwo has joined #openstack-barbican16:55
*** juantwo_ has joined #openstack-barbican16:56
*** juantwo__ has joined #openstack-barbican16:56
*** juantwo__ has quit IRC16:57
*** juantwo__ has joined #openstack-barbican16:57
*** rm_work has quit IRC16:58
*** rm_work has joined #openstack-barbican16:58
*** rm_work has quit IRC16:58
*** rm_work has joined #openstack-barbican16:58
*** juantwo_ has quit IRC17:00
*** juantwo has quit IRC17:00
*** akoneru is now known as akoneru_lunch17:02
*** rm_work has quit IRC17:03
*** rm_work has joined #openstack-barbican17:04
*** rm_work has quit IRC17:04
*** rm_work has joined #openstack-barbican17:04
*** rm_work has quit IRC17:08
*** rm_work has joined #openstack-barbican17:10
woodster_alee: back for a little before heading out for errands. Yes, we would need a facility for seeing what cert plugins are supported17:14
woodster_alee: maybe a GET orders variation?17:15
*** rm_work has quit IRC17:15
aleewoodster_, perhaps.  we're in the process of adding a feature to dogtag that would allow an admin to define multiple lightweight sub-ca's within a single ca.17:18
aleewoodster_,  that would be very useful for example, if you wanted a particular project to have its own subca17:19
aleeto restrict the scope of certificates within the project.17:19
*** rm_work has joined #openstack-barbican17:19
*** rm_work has quit IRC17:19
*** rm_work has joined #openstack-barbican17:19
woodster_alee: is there a mechanism exposed to list out all of those available cas/subcas? One we might adopt in barbican perhaps?17:20
aleewoodster_,  I'd like to design with that feature in mind -- basically, the idea that a single plugin might serve multiple ca's17:20
aleewoodster_, there will be :)17:20
*** juantwo__ has quit IRC17:20
aleewoodster_,  its in the planning phases now - but the idea is definitely to expose this to barbican17:21
*** juantwo has joined #openstack-barbican17:21
aleeand maybe even expose through some barbican admin inteface, the ability to create/ request creation of subca's ..17:21
*** juantwo has quit IRC17:22
aleewoodster_, I'll probably propose that as a possible design session at the K summit17:22
aleewoodster_,  in any case, I think we may need a table to map ca/subca identifier to plugin17:23
aleewoodster_, and we might want to expose that data to the end-user.17:24
aleeso that the client knows which ca's are available17:24
aleewoodster_,  there is also the possibility that we could define some kind of usage() function that plugins could choose to implement17:25
aleeto get the current required/optional fields ..17:25
*** rm_work has quit IRC17:26
woodster_alee: that's interesting....maybe a GET from orders?certificates or GET orders?certificates=dogtag would return that info?17:27
*** rm_work has joined #openstack-barbican17:27
aleewoodster_, yup17:27
aleewoodster_, more like GET orders?ca_id=X perhaps?17:28
aleewhere ca_id will map to a plugin17:29
woodster_alee: sounds interesting17:29
aleewoodster_, yeah - so do you think the certificate id's should be a resource in and of themselves?17:31
*** rm_work has quit IRC17:32
*** rm_work has joined #openstack-barbican17:32
*** juantwo has joined #openstack-barbican17:32
woodster_alee: I was hoping we could use names like 'symantec-xxxx' or 'dogtag-xxxx' that the plugins could produce maybe?17:33
aleewoodster_, incidentally we just found a need for plugin_name in get_plugin() in certificate_manager() -- and sooner than I expected :)17:33
woodster_alee: ha, yes. This is a little more sophisticated though...like a supports() call but with that ca_id passed in?17:34
aleewoodster_, I think that may be a little brittle - especially if eventually folks may get the ability to add subca's17:35
aleefolks for example might want to name their subca -- dev-ca for instance17:35
aleeand allowing them to do that seems much more user friendly.17:36
*** rm_work has quit IRC17:37
*** atiwari has joined #openstack-barbican17:38
*** rm_work has joined #openstack-barbican17:38
*** rm_work has quit IRC17:38
*** rm_work has joined #openstack-barbican17:38
aleewoodster_, if we allow more generic names, then we need to keep those names in barbican-core and barbican-core should do the id->plugin lookup.17:39
aleeand we should just have get_plugin(plugin_name=dogtag)17:39
woodster_alee: interesting. Ok, it might be good to have a blueprint to capture that, or at least an etherpad.  Heading out now for a bit, let's talk about this a bit later17:39
aleesure17:39
aleewoodster_, I'll start an etherpad17:40
*** rm_work has quit IRC17:43
*** rm_work has joined #openstack-barbican17:44
*** rm_work has quit IRC17:44
*** rm_work has joined #openstack-barbican17:44
*** rm_work has quit IRC17:49
*** rm_work has joined #openstack-barbican17:49
*** rm_work has quit IRC17:50
*** rm_work has joined #openstack-barbican17:50
*** rm_work has quit IRC17:54
*** akoneru_lunch is now known as akoneru17:59
*** rm_work has joined #openstack-barbican18:05
*** rm_work has quit IRC18:05
*** rm_work has joined #openstack-barbican18:05
*** rm_work has quit IRC18:11
*** rm_work has joined #openstack-barbican18:12
*** rm_work has quit IRC18:16
*** rm_work has joined #openstack-barbican18:16
*** ayoung has quit IRC18:26
*** rm_work has quit IRC18:29
*** rm_work has joined #openstack-barbican18:29
*** SheenaG1 has quit IRC18:31
*** rm_work has quit IRC18:33
*** rm_work has joined #openstack-barbican18:33
*** SheenaG11 has joined #openstack-barbican18:33
*** paul_glass has joined #openstack-barbican18:36
*** openstackgerrit has quit IRC18:41
*** jenkins-keep has quit IRC18:41
*** openstackgerrit has joined #openstack-barbican18:50
openstackgerritRavi Sankar Penta proposed a change to openstack/barbican: Install sqlite-devel package on fedora  https://review.openstack.org/11340418:53
redrobotrm_work kinda... still WIP though18:53
rm_workredrobot: like, WIP and posted? or WIP on your hard drive :P19:01
redrobotrm_work the latter.19:01
*** jenkins-keep has joined #openstack-barbican19:05
woodster_atiwari: are you there?19:13
rm_workredrobot: k19:14
openstackgerritConstanze Kratel proposed a change to openstack/barbican: IMaking updates to ck-devguide.xml after tech review Updating dev guide and WADL after tech review and edit Incorporated review comments from John Wood into WADL and devguide  https://review.openstack.org/10548819:16
openstackgerritA change was merged to openstack/python-barbicanclient: Introduce cliff for cli framework  https://review.openstack.org/10758719:25
rm_workredrobot: if you finish that BP even partially (like for Secrets or Orders individually) you should go ahead and post it as a WIP, I can get started on one of those, even if there would be minor changes19:26
rm_workredrobot: though, I think I know what you are thinking anyway (using the Container BP to extrapolate), so I might be able to get started regardless19:26
rm_workthough it's funny to think about writing the code first and following up with a BP spec :P\19:27
hockeynutthat way you know the code matches the spec ;-)19:28
rm_worktruth19:37
*** ayoung has joined #openstack-barbican20:15
*** paul_glass has quit IRC20:20
openstackgerritJohn Vrbanac proposed a change to openstack/barbican: Updating symantec plugin docstrings  https://review.openstack.org/11401720:35
*** crc32 has joined #openstack-barbican20:37
jvrbanacalee, reaperhulk, redrobot, codekobe_, dstufft, could one of y'all workflow this CR https://review.openstack.org/#/c/110144/20:38
redrobotjvrbanac -1  symantecssl should be added to requirements.txt20:44
redrobotor maybe not?  ...20:45
* redrobot still unclear about optional requierements20:45
redrobotwhich is an oxymoron...  if it's optional it's not required, right?20:45
* redrobot shrugs20:45
openstackgerritJohn Wood proposed a change to openstack/barbican: Install of a Barbican RPM fails due to missing keystonemiddleware  https://review.openstack.org/11402220:50
woodster_redrobot: I don't think anything outside of global reqs should show up in requirements.txt20:51
woodster_redrobot: so we are using Ade's Dogtag approach as a guideline, whereby only if dogtag is activated via stevedore do we try to do anything with it20:51
woodster_redrobot: said another way, Symantec and Dogtag are optional deployment modes for Barbican, but not out of the box for general OpenStack installs20:53
reaperhulkyep, so they don't belong in global reqs (and indeed we should remove that pykcs11 todo in there)20:53
redrobotright, but symantec is listed as the default plugin out of the box20:57
redrobothttps://github.com/openstack/barbican/blob/master/barbican/plugin/interface/certificate_manager.py#L3620:57
redrobotI think?20:57
*** rm_work has quit IRC20:57
*** rm_work has joined #openstack-barbican20:57
*** rm_work has quit IRC20:58
*** rm_work has joined #openstack-barbican20:58
*** juantwo has quit IRC20:58
woodster_good point...that should be removed, in favor of an as-yet-written default cert plugin but a plugin not found 500 is appropriate for now20:58
woodster_atiwari: are you there?20:59
openstackgerritArun Kant proposed a change to openstack/barbican: Adding keystone notification listener support  https://review.openstack.org/11081721:01
*** rm_work has quit IRC21:01
*** rm_work has joined #openstack-barbican21:04
woodster_ugh, devstack gate failed on a documentation CR...The Barbican (non-admin) API failed to respond within 20 seconds21:04
*** rm_work has quit IRC21:04
*** rm_work has joined #openstack-barbican21:04
woodster_....https://review.openstack.org/#/c/105488/21:04
jvrbanacredrobot, I figured it wasn't required right now21:04
jvrbanacredrobot, we don't depend on dogtag stuff either21:05
redrobot jvrbanac see comment about symantec being enabled out of the box21:05
redrobotjvrbanac unlike dogtag21:05
jvrbanacredrobot, I was trying to avoid taking over Kevin's CR. Oh well, let me toss that in.21:06
*** rm_work has quit IRC21:07
*** rm_work has joined #openstack-barbican21:07
*** rm_work has quit IRC21:07
*** rm_work has joined #openstack-barbican21:07
redrobotjvrbanac I think maybe disabling a default plugin would be best?  Seems we want to keep symantecssl out of requirements.txt21:07
woodster_redrobot: yeah, better to have no default plugin rather than a non-global-reqs one21:09
jvrbanacredrobot, woodster_ agreed.21:10
jvrbanacI think it's better to keep the initially activated plugins to a minimum.21:10
jvrbanacIt feels like more of a deployer choice.21:11
jvrbanacredrobot, so what do you want me to do with the symantec CR?21:13
woodster_jvrbanac: he just stepped away21:14
*** rm_work has quit IRC21:14
woodster_jvrbanac: do you mean in regards to that default plugin being set to symantec?21:14
*** rm_work has joined #openstack-barbican21:18
*** rm_work has quit IRC21:19
*** rm_work has joined #openstack-barbican21:19
jvrbanacwoodster_, yeah. I just want to know if I need to do something to resolve this CR21:21
woodster_jvrbanac: ...vs just accepting as is, and then quick following with a CR to tweak things?21:22
jvrbanacwoodster_, yeah. I added a follow-up CR already to address venkat's comments21:23
woodster_jvrbanac: that works for me, but once I see Douglas I'll see if he is ok with removing the -121:24
jvrbanacwoodster_, awesome thanks!21:24
SheenaG11Heads up, folks.  chellygel just got me set up so I can start wrecking house (read: committing code).21:25
SheenaG11It's about to go DOWN.21:25
rm_workoh shiiiii--21:25
reaperhulkI'll get my -2s ready21:26
SheenaG11Just the kind of love and acceptance I expected from this crew.21:26
chellygeli was really hoping she'd -1 yalls stuff instead21:26
reaperhulkI'll treat you the same as I treat everybody else21:26
reaperhulkPS figured out the IOError issue. Still thinking about the best way to fix it though21:27
chellygelthe adam-connection-reset-drama error?21:27
SheenaG11I'm really just trying to get set up to make doc changes, so hopefully it won't require a lot of -2s21:27
openstackgerritArvind Tiwari proposed a change to openstack/barbican: Reorganize code to use store crypto plug-in  https://review.openstack.org/11141221:27
reaperhulkchellygel: this is the one we had when trying to stand up our dev env21:29
woodster_SheenaG11: with great power comes great responsibility!21:29
reaperhulkunrelated to rm_work's uwsgi issue actually21:29
woodster_SheenaG11: are you planning to make changes after Constanze's CR merges?: https://review.openstack.org/#/c/105488/21:29
chellygelah21:29
SheenaG11Yeah - it's not a today thing, chellygel still has much to learn me21:30
SheenaG11Sorry, that was for woodster_21:30
chellygelwell im still pulling for SheenaG11 to start voting on that CR too.. so maybe thats next lesson!21:31
SheenaG11Noooooooooooo21:31
woodster_reaperhulk: is the ioerror issue C-related or stevedore/plugin-mgr related?21:32
reaperhulkplugin manager21:32
reaperhulkIf we singleton the PluginManager instance the issue completely disappears21:33
*** rm_work has quit IRC21:33
*** rm_work has joined #openstack-barbican21:33
*** rm_work has quit IRC21:35
*** rm_work has joined #openstack-barbican21:36
redrobotjvrbanac +2 on a failed gate.  Naughty, naughty! https://review.openstack.org/#/c/114017/21:38
jvrbanacredrobot, lol! Nothing like a recheck no bug ll21:39
jvrbanac^lol21:39
woodster_devstack gate is borking!21:39
woodster_did the admin api change recently?21:40
jvrbanacwoodster_, not that I know of21:40
woodster_Does anyone have an issue with a singleton plugin manager?21:41
woodster_reaperhulk: interesting that we hadn't noticed this issue before using pkcs11 though.  Hopefully only stevedore is our issue here.21:42
reaperhulkIt only happened after the plugin refactor21:42
reaperhulkSince prior to that we didn't instantiate this way21:42
reaperhulkMy last test was prior to that code being merged :)21:42
woodster_we had the extension manager before...was it singleton? I didn't think so21:43
reaperhulkI believe it was21:44
woodster_sorry, 'The Barbican (non-admin) API failed to respond within 20 seconds'...so the regular API is croaking in devstack it seems21:45
woodster_reaperhulk: So can you put up a bug and then CR to fix that?21:45
reaperhulkYeah I will as soon as I make up my mind how I want to fix it :)21:45
woodster_reaperhulk: sounds good, thanks for digging into that one...21:46
*** crc32 has quit IRC21:47
rm_workI can maybe look at that >_>21:48
* rm_work looks21:48
rm_worksince all my devstack stuff is spun up to mimick the gate as closely as possible21:48
*** SheenaG11 has quit IRC21:49
*** crc32 has joined #openstack-barbican21:53
redrobotlooks like it's pbr taking a shit on the version?21:58
*** nkinder has quit IRC22:05
jvrbanacarg!!22:09
rm_workAttributeError: 'Module_six_moves_urllib_parse' object has no attribute 'SplitResult'22:10
rm_workis what I'm getting22:10
rm_workwhich is in stevedore22:10
rm_workold six lib it looks like22:10
rm_workhmm or too new six lib :P22:10
redrobotsix 1.7.3 released on 6/29/14 is latest22:11
rm_workhmm22:11
rm_worksix.moves.urllib.parse.SplitResult exists in 3.3, doesn't in 2.x22:14
rm_workwtf22:14
rm_worki thought the whole POINT of 'six' was to provide compatibility from 2.x and 3.x22:14
rm_workis stevedore a new req?22:15
redrobotno, we've been stevedoring for a while now22:15
rm_workhmm22:15
redrobotand yes, that is weird... six IS supposed to level the 2 vs 3 field.22:16
rm_workthe stack is from http://pastebin.com/raw.php?i=UYLyxFD322:17
*** akoneru has quit IRC22:17
rm_workerr, stacktrace is in that link22:17
rm_workbarbican queue init -> oslo messaging transport -> stevedore -> six22:18
redrobotoh wow this is oslo.messaging-> stevedore -> six22:18
redrobotJINX!22:18
rm_workheh22:18
rm_workI beat you by like SIX seconds :P:P22:18
rm_workwhat plugin was stevedore trying to load22:22
redrobotwho knows... I'm not familiar with oslo.messaging internals22:22
rm_workhmmm22:23
rm_workhttps://github.com/openstack/barbican/blob/master/barbican/openstack/common/network_utils.py22:23
rm_workis the only place barbican uses SplitResult22:24
rm_workhttps://github.com/openstack/oslo.messaging/blob/master/oslo/messaging/openstack/common/network_utils.py22:24
rm_workis where they use it22:24
rm_workand yeah theirs is from six.moves22:24
redroboteh...  fun... that comes from oslo's "managed copy paste"22:25
redrobotmaybe we're due for a resync22:25
redrobotbecause, you know, libs aren't good enough for this stuff..22:25
redrobotrm_work https://github.com/openstack/oslo-incubator/blob/master/openstack/common/network_utils.py22:26
redrobotyeah, looks like there's a new version using six22:27
*** ayoung is now known as ayoung_is_a_Dad22:27
redrobotcongratulations ayoung_is_a_Dad !!!22:27
redrobotrm_work do you know how to resync oslo-incubator?22:28
rm_worka) no22:29
rm_workb) is that the problem?22:29
rm_worki thought the problem WAS that they are using six :P22:29
rm_workthey need to NOT be using six? :P22:29
rm_workthe version we have now is the version that uses six, and it's broken22:30
redrobotrm_work network_utils.py in barbican is outdated22:30
rm_workah22:30
rm_workyou mean OURS22:30
redrobotyep22:30
rm_workso... now ours will break too? :P i'm maybe missing the point22:30
rm_workif we sync... then barbican will ALSO have the broken version (the one using six)22:31
rm_workwell, broken for 2.x, working for 3.x22:31
redrobothttp://i.imgur.com/xVyoSl.jpg22:32
rm_workI think really, we need to do:22:33
rm_workfrom urllib import parseif six.moves.urllib.22:33
rm_workerr22:33
*** crc32 has quit IRC22:33
rm_workimport six22:36
rm_workif six.PY3:22:36
rm_work    from six.moves.urllib import parse22:36
rm_workelse:22:36
rm_work    from urllib import parse22:36
rm_work>_>22:36
rm_workshitty22:36
redrobotwow super shitty22:36
redrobotwhere is this needed?22:36
rm_workwhich is exactly what six is supposed to freaking prevent22:36
rm_workoslo messaging, I think22:36
rm_workwell, that file you linked22:37
redrobotnetwork_utils?22:37
rm_workanywhere that uses SplitResult22:37
rm_workyeah22:37
rm_worklet me try and see what happens22:37
redrobotright, so22:37
redrobotI was saying22:37
redrobotanything in barbican/openstack/common/* can't be edited by us22:37
rm_workright22:37
redrobotI just don't understand why this would start failing now... it's not like oslo.messaging got a new relese22:38
rm_workhmm22:39
*** juantwo has joined #openstack-barbican22:40
*** juantwo has quit IRC22:41
rm_workah it's oslo.messaging/oslo/messaging/transport.py22:41
*** juantwo has joined #openstack-barbican22:41
*** crc32 has joined #openstack-barbican22:49
rm_workok so22:51
rm_workimport six22:51
rm_workif six.PY3:22:51
rm_work    from six.moves.urllib import parse22:51
rm_workelse:22:51
rm_work    import urlparse as parse22:51
rm_workin openstack/common/network_utils.py22:52
rm_workand oslo/messaging/transport.py22:52
rm_workand then it works >_>22:52
rm_workreally just the first one matters for us22:53
rm_workbut both are broken22:53
rm_workit's not our code-tree though, it's the oslo import of the network_utils that's bad22:53
rm_workand yeah, I have no idea how that works, it's just a submodule we sync in periodically?22:53
rm_workhttps://github.com/openstack/oslo-incubator/blob/master/openstack/common/network_utils.py#L2322:55
rm_workI'm with you though on having no idea why this would pop up NOW22:56
redrobotyeah, its supposed to be synced whenever the project feels comfortable syncing22:56
redrobotoslo calls it "managed copy paste" or somethign to that effect22:57
redrobotI bitched about it a lot at the Atlanta summit22:57
rm_workhmm22:57
redrobotbecause you know, whe have these things called libraries in python where you can release code instead of copy-pasting it everywhere...22:57
rm_worktrying to see if oslo.messaging recently did that22:57
rm_worknot sure how to tell22:57
rm_worklooking through https://review.openstack.org/#/q/project:openstack/oslo.messaging,n,z22:57
redrobotrm_work https://review.openstack.org/#/q/project:openstack/oslo.messaging+status:merged,n,z works better22:58
rm_workerr, right22:59
rm_workfuck, i have no idea23:03
redrobotlmao23:03
redroboti'm right there with you bro23:03
rm_workwas this really merged in FEBRUARY!?23:04
rm_workhow the hell23:04
rm_workso obviously we just weren't triggering it until recently23:04
redrobotwhere did you grab that stacktrace from again?23:04
rm_workmy devstack box23:05
redrobotheh23:06
redrobotsooooooo23:06
rm_worksoooooooooooo?23:06
redrobotthe stacktrace I'm seeing23:06
redrobotis pbr shitting itself23:06
redrobothttp://logs.openstack.org/17/114017/1/check/gate-barbican-devstack-dsvm/4ebdcee/logs/screen-barbican.txt.gz23:06
rm_workhmmm23:06
rm_workrebuilding my devbox23:09
rm_workfrom scratch23:11
rm_workagain.23:11
*** nkinder has joined #openstack-barbican23:14
redrobotrm_work word... I'm going to take off.  I'll try to jump back on later.23:26
rm_workkk23:26
openstackgerritDouglas Mendizábal proposed a change to openstack/barbican: Revert remove version from setup.cfg  https://review.openstack.org/11408023:29
redrobotrm_work actually, gonna stick around for just a sec23:29
rm_workkk23:29
redroboti suspect this changed borked the gate https://review.openstack.org/#/c/109580/1023:29
rm_workthe pbr version bit?23:29
redrobotaye23:30
rm_workhow did IT pass the gate then?23:30
redrobotha!  good question23:31
* redrobot shrugs23:31
*** ravips has joined #openstack-barbican23:33
redrobotwell it doesn't look like zuul wants to test my theory any time soon, so I'm defnitely taking off23:47
rm_workhmm k23:48
rm_workredrobot: well looks like i was chasing ghosts23:48
rm_workthough i am curious WTF23:48
rm_workbecause even if it's not our error, i don't understand how that urllib parse code would work in 2.x <_<23:49
rm_worklooks like the PBR error is what I get now in a clean devstack23:49
rm_workand yeah, can't keep a devstack instance for more than like, 5 minutes it seems <_<23:49
*** jamielennox|away is now known as jamielennox23:51
« Tuesday, 2014-08-12 Index Thursday, 2014-08-14 »

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