Thursday, 2016-11-03

« Wednesday, 2016-11-02 Index Friday, 2016-11-04 »
*** john-davidge has joined #openstack-sdks00:36
*** john-davidge has quit IRC00:40
*** markvoelker has quit IRC00:52
*** hoangcx has joined #openstack-sdks01:06
*** zhurong has joined #openstack-sdks01:18
*** yanyanhu has joined #openstack-sdks01:20
*** aohuanxuan has joined #openstack-sdks01:22
*** aohuanxuan has quit IRC01:41
*** aohuanxuan has joined #openstack-sdks01:41
*** markvoelker has joined #openstack-sdks01:53
*** markvoelker has quit IRC01:58
*** fzdarsky_ has joined #openstack-sdks02:15
*** fzdarsky has quit IRC02:18
*** yanyanhu has quit IRC02:48
*** yanyanhu has joined #openstack-sdks02:49
*** gouthamr has quit IRC03:04
*** xehbit has quit IRC03:05
*** xehbit has joined #openstack-sdks03:07
*** john-davidge has joined #openstack-sdks03:37
*** john-davidge has quit IRC03:41
*** markvoelker has joined #openstack-sdks03:54
*** markvoelker has quit IRC03:59
*** adriant has quit IRC05:39
openstackgerritYujun Zhang proposed openstack/js-openstack-lib: Ignore compiling output  https://review.openstack.org/39309705:41
*** openstackgerrit has quit IRC05:48
*** openstackgerrit has joined #openstack-sdks05:48
openstackgerritYujun Zhang proposed openstack/js-openstack-lib: Ignore compiling output  https://review.openstack.org/39309705:49
*** markvoelker has joined #openstack-sdks05:55
openstackgerritDongcan Ye proposed openstack/python-openstackclient: Add address pair options in port operation  https://review.openstack.org/39245705:56
*** markvoelker has quit IRC05:59
*** john-davidge has joined #openstack-sdks06:04
*** john-davidge has quit IRC06:08
*** annp has joined #openstack-sdks06:23
*** sheel has joined #openstack-sdks06:24
*** zhurong has quit IRC06:25
*** zhurong_ has joined #openstack-sdks06:25
openstackgerritmiaohb proposed openstack/python-openstacksdk: Add filters "policy_type" and "policy_name" for policy binding list  https://review.openstack.org/38787006:32
*** markvoelker has joined #openstack-sdks06:55
*** exploreshaifali has joined #openstack-sdks06:57
*** markvoelker has quit IRC07:00
*** zhurong_ has quit IRC07:03
*** zhurong has joined #openstack-sdks07:04
*** exploreshaifali has quit IRC07:08
*** openstackgerrit has quit IRC07:48
*** openstackgerrit has joined #openstack-sdks07:48
*** jpena|off is now known as jpena08:20
openstackgerritHuanxuan Ao proposed openstack/python-openstackclient: Refactor "snapshot" commands  https://review.openstack.org/36985408:21
*** jpena is now known as jpena|off08:32
*** jpena|off is now known as jpena08:35
*** aohuanxuan has quit IRC08:39
*** huanxuanao has joined #openstack-sdks08:40
*** huanxuanao has quit IRC08:40
*** huanxuan has joined #openstack-sdks08:41
*** zhiyong has joined #openstack-sdks08:54
*** zhiyong_ has joined #openstack-sdks08:55
*** zhiyong has quit IRC08:55
*** zhiyong_ has quit IRC08:55
*** jiahui has joined #openstack-sdks08:56
*** markvoelker has joined #openstack-sdks08:56
*** zhiyong has joined #openstack-sdks08:57
*** markvoelker has quit IRC09:00
*** amotoki has quit IRC09:04
*** john-davidge has joined #openstack-sdks09:10
*** zhurong has quit IRC09:14
*** huanxuan is now known as aohuanxuan09:18
*** aohuanxuan is now known as huanxuan09:18
*** lucas-afk is now known as lucasagomes09:22
openstackgerritDongcan Ye proposed openstack/python-openstackclient: [network] Add address pair options in port operation  https://review.openstack.org/39245709:30
*** yanyanhu has quit IRC09:38
*** yanyanhu has joined #openstack-sdks09:38
*** yanyanhu has quit IRC09:42
*** zhurong has joined #openstack-sdks09:49
*** gildub_ has quit IRC09:58
*** zhurong has quit IRC10:02
*** zhurong has joined #openstack-sdks10:04
*** zhurong has quit IRC10:04
openstackgerritDongcan Ye proposed openstack/python-openstackclient: [network] Add address pair options in port operation  https://review.openstack.org/39245710:14
*** hoangcx has quit IRC10:20
*** rtheis has joined #openstack-sdks10:22
*** cdent has joined #openstack-sdks10:42
*** cdent has quit IRC10:51
*** markvoelker has joined #openstack-sdks10:57
*** markvoelker has quit IRC11:02
*** huanxuan is now known as aohuanxuan11:02
*** aohuanxuan is now known as huanxuan11:02
*** huanxuan has quit IRC11:10
*** huanxuan has joined #openstack-sdks11:11
*** huanxuan has quit IRC11:12
*** annp has quit IRC11:22
*** rvasilets___ has joined #openstack-sdks11:33
*** cdent has joined #openstack-sdks11:52
*** jpena is now known as jpena|lunch11:57
*** edmondsw has joined #openstack-sdks12:05
*** jiahui has quit IRC12:10
*** jiahui has joined #openstack-sdks12:12
*** markvoelker has joined #openstack-sdks12:22
*** aohuanxuan has joined #openstack-sdks12:30
*** aohuanxuan is now known as huanxuan12:34
*** lucasagomes is now known as lucas-hungry12:35
*** john-davidge has quit IRC12:45
*** john-davidge has joined #openstack-sdks12:46
*** gouthamr has joined #openstack-sdks12:49
*** david-lyle has quit IRC12:55
*** jpena|lunch is now known as jpena12:59
*** Kiall_ is now known as Kiall13:19
*** _elmiko is now known as elmiko13:24
*** fguillot has joined #openstack-sdks13:31
*** lucas-hungry is now known as lucasagomes13:35
*** cleong has joined #openstack-sdks13:38
*** amotoki has joined #openstack-sdks13:50
*** sheel has quit IRC14:00
*** huanxuan has quit IRC14:14
*** singhj has joined #openstack-sdks14:26
*** annegentle has joined #openstack-sdks14:31
openstackgerritSindhu Devale proposed openstack/python-openstackclient: OSC Quota List  https://review.openstack.org/37981314:49
*** cdent has quit IRC14:51
*** amotoki has quit IRC14:53
openstackgerritSindhu Devale proposed openstack/python-openstackclient: SDK refactor: Prepare floating ip commands  https://review.openstack.org/39270914:54
*** d0ugal has quit IRC14:57
*** e0ne has joined #openstack-sdks14:59
openstackgerritSindhu Devale proposed openstack/python-openstackclient: OSC Extension Show  https://review.openstack.org/38202315:01
*** ALVAN has left #openstack-sdks15:03
*** jiahui has quit IRC15:07
*** jiahui has joined #openstack-sdks15:07
*** d0ugal has joined #openstack-sdks15:13
*** cdent has joined #openstack-sdks15:19
*** e0ne has quit IRC15:30
*** d0ugal has quit IRC15:39
*** ankur-gupta-f has left #openstack-sdks15:50
*** ankur-gupta-f has joined #openstack-sdks15:50
cdentetoews, elmiko you guys aligned with the current timezone weirdness?16:00
elmikoum, maybe?16:00
etoewsyes?16:00
*** thingee has joined #openstack-sdks16:23
openstackgerritShashank Kumar Shankar proposed openstack/python-openstackclient: [WIP] Add neutron dhcp-agent to OSC.  https://review.openstack.org/38761116:27
*** cdent has quit IRC16:44
*** edmondsw has quit IRC16:46
*** cdent has joined #openstack-sdks16:47
*** piet has joined #openstack-sdks16:51
elmikoit's a long shot, but, any of you heading to kubecon or apachecon?17:00
elmikoetoews, cdent ^^17:01
etoewsneither17:01
elmikoah well, had to ask =)17:01
cdentelmiko: I'd like to (on both) but not gonna happen17:01
elmikocdent: at least apachecon is on your side of the pond ;)17:02
cdent"hey people with the money, I know I was just in spain, but could I maybe go back to spain?"17:02
elmikohehe17:02
*** njohnston has left #openstack-sdks17:03
*** gouthamr has quit IRC17:05
*** singhj has quit IRC17:06
*** gouthamr has joined #openstack-sdks17:08
*** xehbit has quit IRC17:17
*** xehbit has joined #openstack-sdks17:17
*** flwang1 has joined #openstack-sdks17:19
flwang1hi api-wg folks, i have a question about the 'actions' of rest api17:19
flwang1what's the recommended way of an action api for a particular resource?17:19
flwang1thanks17:20
*** e0ne has joined #openstack-sdks17:20
openstackgerritAnkur proposed openstack/python-openstackclient: Add meter label rule to OSC  https://review.openstack.org/37805817:33
*** aarefiev has quit IRC17:34
elmikoflwang1: kinda depends on what you mean by "action"17:35
*** aarefiev_ has joined #openstack-sdks17:35
flwang1elmiko: like purge a queue17:35
flwang1in glance, it's using '/image_id/actions/activate'17:35
*** aarefiev_ has quit IRC17:36
flwang1in nova, it's using '/instance_id/actions/'17:36
elmikois there a way to set the queue length to 0 in the resource and update the model with a PUT or something?17:36
elmikoah17:36
flwang1in sahara, they are using '/job_id/cancel'17:36
elmikowell, in general we advise against making endpoints like the glance one or the sahara one17:36
flwang1i just want to figure out which is the best practice now, in openstack world17:36
elmikoiirc, the nova actions are a resource unto themselves, and that would be more inline with our best thinking on this topic17:37
elmikowe had been working on an actions guideline, but it is a complicated issue17:37
*** aarefiev_ has joined #openstack-sdks17:38
flwang1elmiko: so you mean, the way using '/resource_id/actions' and then post a body to let the server know what's the particular action  user is asking?17:38
elmikoright17:38
elmikotreat the actions as resources17:38
flwang1i see, hence why i asked17:38
flwang1since i don't want to create a new endpoint which is not api-wg like :)17:38
elmikobut, i would also recommend looking at your resources to see if you can avoid using custom actions by allowing model updates to dictate some actions17:39
elmiko=)17:39
flwang1in our case, the resource is queue(in zaqar world)17:39
elmikoso, in the queue example, is it possible to put an updated model back to the server with a requested queue length of 0, thus indicating you wish to empty the queu.17:39
flwang1and we would like to have an action to purge the queue but not delete the queue17:39
elmikoah, interesting17:40
flwang1purge means deleting all the resources under the queue17:40
flwang1but keep the queue since there are some metadata user may want to keep17:41
elmikoright, so you don't want to DELETE /queue/somequeueid17:41
flwang1exactly17:41
flwang1since we do have the delete queue behaviour17:41
flwang1they are different actions17:41
elmikoright17:41
flwang1in other words, we're using DELETE for deleting queue17:42
elmikois it possible to just PUT the queue model back to the server requesting a 0 length state?17:42
elmikoin this manner you are still just dealing with resource state transfers, and the server knows what to do17:42
flwang1elmiko: hmm.... that's not really the way i would like to go17:42
flwang1because17:42
*** aarefiev_ has quit IRC17:42
flwang1we're using PUT to create a queue17:43
elmikoPATCH perhaps?17:43
flwang1PATCH is used to update a queue :D17:43
flwang1but i can see your point17:43
elmikoright, this would be an update requesting a 0 length. or is that out of band?17:43
elmikoyeah, my rationale is to avoid creating specialized or unique action endpoints until all options have been exhausted17:44
elmikoi find it more idiomatic to deal with resource models and requesting the server to achieve a certain statefulness17:44
flwang1based on my understand, no matter using which method, we still need to define a http body like nova does, to tell the server what to do, right?17:44
elmikoit depends17:44
elmikoif i could send the queue model back with some length key set to 0, then no you don't need a new body17:45
elmikobut,17:45
* cdent catches up17:45
flwang1actually, in zaqar world, there is no length for queue17:45
elmikoif you create an actions endpoint, then i would treat each action as a resource unto itself and in that case you would create a new type of model to send back (ie. the action model)17:45
elmikocdent probably has some good ideas too17:46
cdentI agree with what elmiko has said about trying to change the state on the resource itself, and not using specific action endpoints17:46
cdenteven if that mean you need a new url to represent the state17:47
cdentfor a queue-like resource one can imagine that if the queue is: /queues/{id}17:47
cdentand you manage the queue by putting, patching, posting to it17:47
cdentthen you might have a /queues/{id}/status to manage it17:48
cdentit would be nicer, however, to manage it at /queues/{id}17:48
cdentperhaps with a different media type17:48
elmikoi gotta run, hopefully this is helpful flwang1. i think the only way you can go seriously wrong is if you make an endpoint like /queues/{id}/empty, or similar. good luck! =)17:49
flwang1ha17:50
flwang1elmiko: thanks17:50
cdentyeah, verby endpoints are bad news17:50
flwang1cdent: so i should totally avoid the way like glance and sahara did, right?17:50
elmikofwiw, sahara is factoring out the cancel endpoint in their v2 api17:51
cdentflwang1: yes the glance and saraha ways you've described are not great17:51
flwang1elmiko: cool17:52
flwang1cdent: ok, then i may go for using POST /queue/myqueue/actions to request a particular action17:53
flwang1since i don't want to find another http method for new actions in the future17:54
cdentflwang1: that seems to be the way that has become common in openstack, but for the record, I don't much care for it myself :)17:54
cdent(for all the reasons elmiko already said)17:55
flwang1cdent: ok, cool, thank you very much17:55
flwang1btw17:55
flwang1for now17:55
flwang1is it the recommended way to use PATCH for updating?17:55
flwang1or PUT is still the 'default standard'?17:55
*** lucasagomes is now known as lucas-afk17:59
*** jpena is now known as jpena|off18:03
cdentflwang1: probably PUT18:04
flwang1cdent: interesting...18:04
cdentas for at least some people PATCH requires a diff-like format18:04
cdentso unless you are using that a diff-like format, it can be easier to just PUT18:05
cdentbut, like so many things, that's open for debate18:05
flwang1hah18:05
flwang1we're migrating PUT to PATCH, hence asking18:05
cdentflwang1: I missed the part where you said who "we" is?18:06
flwang1cdent: sorry, zaqar team18:06
flwang1but like you said, we're not 100% sure if that's the best way18:07
cdentflwang1: if there are specs associated with that sort of stuff if you invite me, elmiko and etoews to come and look we can provide some input18:09
cdentsome of the time you'll get three different answers though :)18:09
flwang1hah, cool, thanks for the offer ;)18:09
cdentthis is an interesting blog post on patch: http://williamdurand.fr/2014/02/14/please-do-not-patch-like-an-idiot/18:10
cdenti'm not entirely in agreement with it, but it does help present some of the debates18:10
flwang1cdent: ah, i see think i read it before18:14
flwang1cdent: thank you very much18:18
cdentflwang1: you're welcome, but I'm pretty sure all I've done is said "there's no clear cut answer, good luck!" :)18:18
cdentbrb18:18
openstackgerritAnindita Das proposed openstack/python-openstackclient: [WIP] OSC Network Flavor  https://review.openstack.org/38253018:23
*** piet has quit IRC18:32
*** singhj has joined #openstack-sdks18:37
*** john-davidge has quit IRC18:41
*** annegentle has quit IRC18:43
*** flwang1 has quit IRC18:49
*** flwang1 has joined #openstack-sdks18:57
*** waj334 has joined #openstack-sdks19:01
waj334I'm working on adding Cinder v3 api support to the openstack client. Is there already progress being made on this effort?19:03
elmikocdent: re: PATCH mechanisms, i thought we had agreed that our guidance was PATCH should use either partial model update *or* json-diff, left up to the implementors discretion19:04
dtroyerI don't think we have looked at that yet.  does cinderclient have v3 support in its Python API lib?19:04
dtroyerwaj334: ^^19:04
elmikocdent: and that in general, PUT should be used for full model updates and PATCH for partials19:05
cdentelmiko: I can't remember and haven't had a chance to look yet19:05
elmikonot sure if that actually made it into the guidance, but i remember having a big discussion about it19:05
waj334dtroyer: Yes as far as I know. Currently v2 just makes calls from the v3 api.19:06
dtroyerwaj334: is the cinder v3 the same endpoint?  if cinderclient already translates v2 lib calls to v3 rest, what support is required?19:08
*** ankur-gupta-f has quit IRC19:08
*** e0ne has quit IRC19:12
waj334dtroyer: The endpoint should be different. v2 probably will not have awareness of microversions like the new v3 api has19:14
*** singhj has quit IRC19:27
*** flwang1 has quit IRC19:35
*** malloc83 has quit IRC19:45
*** singhj has joined #openstack-sdks19:54
*** rvasilets___ has quit IRC19:58
*** cdent has quit IRC19:59
*** rtheis has quit IRC20:12
*** flwang1 has joined #openstack-sdks20:45
openstackgerritKristi Nikolla proposed openstack/keystoneauth: Adds last_request_id to adapter and session classes  https://review.openstack.org/39348520:46
*** john-davidge has joined #openstack-sdks20:53
*** john-davidge has quit IRC20:57
*** john-davidge has joined #openstack-sdks20:57
*** gouthamr has quit IRC20:58
*** cleong has quit IRC20:59
*** john-davidge has quit IRC21:12
*** singhj has quit IRC21:16
*** flwang2 has joined #openstack-sdks21:18
*** flwang1 has quit IRC21:19
*** e0ne has joined #openstack-sdks21:35
*** manjeets has joined #openstack-sdks21:44
*** david-lyle has joined #openstack-sdks22:03
openstackgerritShashank Kumar Shankar proposed openstack/python-openstackclient: [WIP] Add neutron dhcp-agent to OSC.  https://review.openstack.org/38761122:18
*** singhj has joined #openstack-sdks22:20
*** gouthamr has joined #openstack-sdks22:21
*** flwang2 has quit IRC22:24
*** flwang1 has joined #openstack-sdks22:25
*** singhj has quit IRC22:25
*** gildub_ has joined #openstack-sdks22:27
*** adriant has joined #openstack-sdks22:36
*** annegentle has joined #openstack-sdks22:51
*** elmiko is now known as _elmiko23:00
*** annegentle has quit IRC23:12
*** john-davidge has joined #openstack-sdks23:24
*** john-davidge has quit IRC23:29
*** e0ne has quit IRC23:33
*** yuanying has joined #openstack-sdks23:41
« Wednesday, 2016-11-02 Index Friday, 2016-11-04 »

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