| *** jamesmcarthur has joined #openstack-keystone | 00:02 | |
| *** dave-mccowan has quit IRC | 00:07 | |
| *** jamesmcarthur has quit IRC | 00:17 | |
| *** jamesmcarthur has joined #openstack-keystone | 00:18 | |
| *** jamesmcarthur has quit IRC | 00:22 | |
| *** lbragstad has quit IRC | 00:32 | |
| *** ileixe has joined #openstack-keystone | 00:54 | |
| *** erus has quit IRC | 01:17 | |
| *** erus has joined #openstack-keystone | 01:18 | |
| *** gyee has quit IRC | 01:38 | |
| *** itlinux has quit IRC | 01:54 | |
| *** ileixe has quit IRC | 01:59 | |
| openstackgerrit | Merged openstack/keystone master: Update protocol policies for system reader https://review.openstack.org/625352 | 01:59 |
|---|---|---|
| *** ileixe has joined #openstack-keystone | 02:00 | |
| *** Dinesh_Bhor has joined #openstack-keystone | 02:09 | |
| *** itlinux has joined #openstack-keystone | 02:41 | |
| *** Dinesh_Bhor has quit IRC | 02:49 | |
| *** Dinesh_Bhor has joined #openstack-keystone | 02:53 | |
| *** erus has quit IRC | 02:53 | |
| *** erus has joined #openstack-keystone | 02:54 | |
| *** dave-mccowan has joined #openstack-keystone | 03:07 | |
| *** takamatsu has quit IRC | 03:25 | |
| *** itlinux has quit IRC | 03:29 | |
| openstackgerrit | Merged openstack/keystone master: Add tests for domain users interacting with roles https://review.openstack.org/622527 | 03:30 |
| *** dave-mccowan has quit IRC | 03:32 | |
| *** itlinux has joined #openstack-keystone | 03:33 | |
| *** awalende has joined #openstack-keystone | 03:43 | |
| *** erus has quit IRC | 03:43 | |
| *** erus has joined #openstack-keystone | 03:44 | |
| openstackgerrit | Merged openstack/keystone master: Add tests for project users interacting with roles https://review.openstack.org/622528 | 03:46 |
| openstackgerrit | Merged openstack/keystone master: Remove role policies from policy.v3cloudsample.json https://review.openstack.org/622529 | 03:46 |
| *** awalende has quit IRC | 03:47 | |
| *** markvoelker has joined #openstack-keystone | 03:48 | |
| *** vishakha has joined #openstack-keystone | 04:07 | |
| *** spsurya has joined #openstack-keystone | 04:14 | |
| *** markvoelker has quit IRC | 04:20 | |
| *** ileixe has quit IRC | 04:23 | |
| *** erus has quit IRC | 04:44 | |
| *** erus has joined #openstack-keystone | 04:45 | |
| openstackgerrit | Merged openstack/keystone master: Add protocol tests for system member role https://review.openstack.org/625353 | 04:50 |
| openstackgerrit | Vishakha Agarwal proposed openstack/keystone master: Drop py35 jobs https://review.openstack.org/639909 | 05:00 |
| *** ileixe has joined #openstack-keystone | 05:00 | |
| *** takamatsu has joined #openstack-keystone | 05:05 | |
| openstackgerrit | Vishakha Agarwal proposed openstack/python-keystoneclient master: Drop py35 jobs https://review.openstack.org/639910 | 05:09 |
| *** takamatsu has quit IRC | 05:10 | |
| openstackgerrit | Vishakha Agarwal proposed openstack/keystone master: Drop py35 jobs https://review.openstack.org/639909 | 05:13 |
| *** erus has quit IRC | 05:16 | |
| *** erus has joined #openstack-keystone | 05:16 | |
| *** markvoelker has joined #openstack-keystone | 05:17 | |
| *** shyamb has joined #openstack-keystone | 05:19 | |
| openstackgerrit | Vishakha Agarwal proposed openstack/keystonemiddleware master: Drop py35 jobs https://review.openstack.org/639913 | 05:28 |
| *** itlinux has quit IRC | 05:28 | |
| *** shyamb has quit IRC | 05:28 | |
| openstackgerrit | Vishakha Agarwal proposed openstack/python-keystoneclient master: Drop py35 jobs https://review.openstack.org/639910 | 05:29 |
| openstackgerrit | Vishakha Agarwal proposed openstack/keystoneauth master: Drop py35 jobs https://review.openstack.org/639915 | 05:35 |
| *** erus has quit IRC | 05:35 | |
| *** erus has joined #openstack-keystone | 05:36 | |
| openstackgerrit | Vishakha Agarwal proposed openstack/oslo.limit master: Drop py35 jobs https://review.openstack.org/639917 | 05:40 |
| openstackgerrit | Merged openstack/keystone master: Implement system admin role in protocol API https://review.openstack.org/625354 | 05:45 |
| openstackgerrit | Merged openstack/keystone master: Add tests for domain users interacting with protocols https://review.openstack.org/625355 | 05:45 |
| openstackgerrit | Merged openstack/keystone master: Add tests for project users interacting with protocols https://review.openstack.org/625356 | 05:45 |
| *** markvoelker has quit IRC | 05:51 | |
| *** ileixe has quit IRC | 05:52 | |
| *** ileixe has joined #openstack-keystone | 05:52 | |
| *** shyamb has joined #openstack-keystone | 05:53 | |
| openstackgerrit | Vishakha Agarwal proposed openstack/keystone master: Add py37 in setup.cfg https://review.openstack.org/639921 | 05:55 |
| openstackgerrit | Vishakha Agarwal proposed openstack/keystone master: Drop py35 jobs https://review.openstack.org/639909 | 05:58 |
| openstackgerrit | Vishakha Agarwal proposed openstack/ldappool master: Drop py35 jobs https://review.openstack.org/639924 | 06:13 |
| openstackgerrit | Vishakha Agarwal proposed openstack/pycadf master: Add py37 tox env and drop py35 https://review.openstack.org/639930 | 06:21 |
| *** markvoelker has joined #openstack-keystone | 06:48 | |
| *** shyamb has quit IRC | 06:51 | |
| *** shyamb has joined #openstack-keystone | 06:55 | |
| *** tkajinam_ has joined #openstack-keystone | 07:02 | |
| *** tkajinam has quit IRC | 07:05 | |
| *** markvoelker has quit IRC | 07:20 | |
| *** erus has quit IRC | 07:25 | |
| *** erus has joined #openstack-keystone | 07:26 | |
| *** shyamb has quit IRC | 07:39 | |
| *** shyamb has joined #openstack-keystone | 07:40 | |
| *** shyamb has quit IRC | 07:58 | |
| *** awalende has joined #openstack-keystone | 08:11 | |
| *** markvoelker has joined #openstack-keystone | 08:18 | |
| *** imacdonn_ has quit IRC | 08:18 | |
| *** imacdonn_ has joined #openstack-keystone | 08:18 | |
| *** pcaruana has joined #openstack-keystone | 08:32 | |
| *** mvkr has joined #openstack-keystone | 08:35 | |
| *** tkajinam_ has quit IRC | 08:38 | |
| *** erus has quit IRC | 08:39 | |
| *** erus has joined #openstack-keystone | 08:40 | |
| *** markvoelker has quit IRC | 08:50 | |
| *** shyamb has joined #openstack-keystone | 08:55 | |
| *** jaosorior has joined #openstack-keystone | 09:06 | |
| *** fiddletwix has quit IRC | 09:07 | |
| *** erus has quit IRC | 09:16 | |
| *** erus has joined #openstack-keystone | 09:17 | |
| *** shyamb has quit IRC | 09:21 | |
| *** shyamb has joined #openstack-keystone | 09:24 | |
| *** shyamb has quit IRC | 09:36 | |
| *** shyamb has joined #openstack-keystone | 09:36 | |
| *** shyamb has quit IRC | 09:41 | |
| *** markvoelker has joined #openstack-keystone | 09:47 | |
| *** rcernin has quit IRC | 09:48 | |
| *** prashkre has joined #openstack-keystone | 10:03 | |
| *** shyamb has joined #openstack-keystone | 10:14 | |
| *** markvoelker has quit IRC | 10:21 | |
| openstackgerrit | Vishakha Agarwal proposed openstack/keystone master: Add openstack_groups to assertion https://review.openstack.org/588211 | 10:22 |
| *** shyamb has quit IRC | 10:37 | |
| openstackgerrit | Vishakha Agarwal proposed openstack/keystone master: Add openstack_groups to assertion https://review.openstack.org/588211 | 10:49 |
| *** markvoelker has joined #openstack-keystone | 11:18 | |
| *** ileixe has quit IRC | 11:18 | |
| *** erus has quit IRC | 11:18 | |
| *** erus has joined #openstack-keystone | 11:19 | |
| *** shyamb has joined #openstack-keystone | 11:20 | |
| *** Dinesh_Bhor has quit IRC | 11:23 | |
| *** markvoelker has quit IRC | 11:50 | |
| *** prashkre has quit IRC | 11:56 | |
| *** shyamb has quit IRC | 12:14 | |
| *** raildo has joined #openstack-keystone | 12:15 | |
| *** erus has quit IRC | 12:38 | |
| *** erus has joined #openstack-keystone | 12:39 | |
| *** markvoelker has joined #openstack-keystone | 12:47 | |
| *** shyamb has joined #openstack-keystone | 12:53 | |
| *** shyamb has quit IRC | 12:58 | |
| erus | o/ | 13:00 |
| cmurphy | \o | 13:00 |
| *** shyamb has joined #openstack-keystone | 13:01 | |
| openstackgerrit | Merged openstack/keystone master: Implement system reader role for groups https://review.openstack.org/625732 | 13:08 |
| openstackgerrit | Merged openstack/keystone master: Implement system member test coverage for groups https://review.openstack.org/625733 | 13:08 |
| *** dave-mccowan has joined #openstack-keystone | 13:10 | |
| *** erus has quit IRC | 13:10 | |
| *** erus has joined #openstack-keystone | 13:11 | |
| openstackgerrit | Bernhard M. Wiedemann proposed openstack/python-keystoneclient master: Make tests pass in 2020 https://review.openstack.org/640024 | 13:12 |
| *** mchlumsky has joined #openstack-keystone | 13:13 | |
| openstackgerrit | Merged openstack/keystone master: Implement system admin role in groups API https://review.openstack.org/625734 | 13:16 |
| *** jamesmcarthur has joined #openstack-keystone | 13:16 | |
| openstackgerrit | Merged openstack/keystone master: Add shibboleth config to log output https://review.openstack.org/636966 | 13:17 |
| *** markvoelker has quit IRC | 13:20 | |
| *** jamesmcarthur has quit IRC | 13:23 | |
| *** jamesmcarthur has joined #openstack-keystone | 13:26 | |
| *** jamesmcarthur has quit IRC | 13:31 | |
| erus | hi cmurphy how are you today? ^.^ | 13:31 |
| *** shyamb has quit IRC | 13:32 | |
| cmurphy | i'm good erus how are you? | 13:32 |
| *** imacdonn_ has quit IRC | 13:36 | |
| *** vishakha has quit IRC | 13:43 | |
| *** jamesmcarthur has joined #openstack-keystone | 13:46 | |
| cmurphy | kmalloc: ugh i just found a place in the app cred code where we raise either a 400 or 404 for basically the same request :( http://git.openstack.org/cgit/openstack/keystone/tree/keystone/api/users.py#n519 | 13:53 |
| *** lbragstad has joined #openstack-keystone | 13:54 | |
| *** ChanServ sets mode: +o lbragstad | 13:54 | |
| *** jmlowe has quit IRC | 14:01 | |
| *** itlinux has joined #openstack-keystone | 14:01 | |
| *** jamesmcarthur has quit IRC | 14:04 | |
| *** jamesmcarthur has joined #openstack-keystone | 14:05 | |
| kmalloc | Bah | 14:06 |
| kmalloc | That isn't cool :( | 14:06 |
| openstackgerrit | Colleen Murphy proposed openstack/keystone master: WIP: Add role check to access rules https://review.openstack.org/640034 | 14:06 |
| *** itlinux has quit IRC | 14:08 | |
| openstackgerrit | Colleen Murphy proposed openstack/keystone master: Expose bad error handling API for app creds https://review.openstack.org/640035 | 14:09 |
| *** jamesmcarthur has quit IRC | 14:10 | |
| *** jistr is now known as jistr|call | 14:11 | |
| cmurphy | lbragstad: kmalloc today's the last day for non-client releases, we don't have 605043 yet (sorry i haven't really looked at it yet) should we propose a ksa release anyway? | 14:17 |
| *** jmlowe has joined #openstack-keystone | 14:17 | |
| *** jmlowe has quit IRC | 14:17 | |
| cmurphy | and do we want to try to get 633369 in or wait till train? | 14:17 |
| *** markvoelker has joined #openstack-keystone | 14:18 | |
| erus | i'm fine cmurphy :) the autumn is coming o/ | 14:19 |
| cmurphy | erus: yay :) how's it going with remote_id_attribute and get_remote_id_parameter? | 14:20 |
| erus | i have an idea so i'm going to test it today | 14:21 |
| cmurphy | awesome | 14:21 |
| erus | yay :D | 14:22 |
| lbragstad | cmurphy 605043 isn't going to impact existing users (outside of people using ksa directly), so I wouldn't be opposed to proposing a ksa release now and see if we can get that in after | 14:22 |
| cmurphy | okay | 14:24 |
| cmurphy | lbragstad: https://review.openstack.org/640041 | 14:27 |
| *** jmlowe has joined #openstack-keystone | 14:29 | |
| kmalloc | lbragstad: ++ | 14:32 |
| cmurphy | lbragstad: https://review.openstack.org/640045 too, we can do another if we end up wanting to later today | 14:33 |
| cmurphy | lbragstad: https://review.openstack.org/640047 | 14:35 |
| *** jamesmcarthur has joined #openstack-keystone | 14:38 | |
| *** vishakha has joined #openstack-keystone | 14:39 | |
| lbragstad | 0.0 | 14:43 |
| lbragstad | my in progress reviews dropped by 15 last night | 14:44 |
| *** erus has quit IRC | 14:44 | |
| *** jamesmcarthur has quit IRC | 14:44 | |
| *** erus has joined #openstack-keystone | 14:45 | |
| cmurphy | merging all the things | 14:45 |
| lbragstad | a pleasant surprise to see that number go down during your first cup of coffee for the day | 14:49 |
| *** markvoelker has quit IRC | 14:50 | |
| cmurphy | lbragstad: what's the plan for https://review.openstack.org/624218 ? | 15:02 |
| gagehugo | o/ | 15:03 |
| openstackgerrit | Lance Bragstad proposed openstack/keystone master: Add tests for project users interacting with services https://review.openstack.org/620623 | 15:03 |
| openstackgerrit | Lance Bragstad proposed openstack/keystone master: Remove service policies from policy.v3cloudsample.json https://review.openstack.org/619282 | 15:03 |
| * lbragstad digs up a link | 15:03 | |
| *** awalende has quit IRC | 15:04 | |
| lbragstad | cmurphy so - i think we need https://review.openstack.org/#/c/624794/ to land before that patch will pass tests | 15:04 |
| *** awalende has joined #openstack-keystone | 15:04 | |
| lbragstad | i've asked for some feedback from qa folks, but it's been quiet lately | 15:04 |
| lbragstad | i brought it up in our keystone meeting a few weeks ago, too | 15:05 |
| lbragstad | but I can add the depends on back in, not sure why i didn't add that back in | 15:05 |
| cmurphy | i remember now | 15:05 |
| cmurphy | i've been ignoring it due to the w-1 but i guess we'll have to harass the qa team about it | 15:06 |
| lbragstad | i'm not entirely sure how the tempest folks feel about ripping out the domain admin stuff they have in there | 15:06 |
| lbragstad | it conflicts with the ideas we have for how scope should work | 15:06 |
| cmurphy | the ideas we have or the ideas they have? | 15:07 |
| *** erus has quit IRC | 15:08 | |
| lbragstad | what's implemented in tempest for domain admin support conflicts with our ideas of what a domain user should be able to do | 15:08 |
| *** erus has joined #openstack-keystone | 15:08 | |
| cmurphy | i see | 15:09 |
| *** awalende has quit IRC | 15:09 | |
| *** prashkre has joined #openstack-keystone | 15:09 | |
| lbragstad | fwiw - i didn't really start understanding the tempest domain stuff until i started working on this | 15:10 |
| lbragstad | which is certainly a regret of mine | 15:10 |
| *** shyamb has joined #openstack-keystone | 15:10 | |
| lbragstad | i wish i would have been more available when those ideas about domain users were being formed or implemented in tempest | 15:10 |
| cmurphy | sounds like we'll have to step in and help fix/rewrite | 15:12 |
| lbragstad | yeah - i think so, too | 15:12 |
| *** jamesmcarthur has joined #openstack-keystone | 15:18 | |
| *** prashkre has quit IRC | 15:19 | |
| *** itlinux has joined #openstack-keystone | 15:21 | |
| lbragstad | cmurphy ah - i think https://review.openstack.org/#/c/623319/6 is in that same boat | 15:21 |
| lbragstad | ^ that's the patch i added the depends on to | 15:22 |
| lbragstad | but i should update the project series to be the same | 15:22 |
| *** jamesmcarthur has quit IRC | 15:22 | |
| cmurphy | oh should have noticed that wasn't merged yet | 15:23 |
| lbragstad | so many patches, tho | 15:23 |
| *** erus has quit IRC | 15:26 | |
| *** erus has joined #openstack-keystone | 15:27 | |
| *** awalende has joined #openstack-keystone | 15:32 | |
| lbragstad | looks like the qa team has office hours at 17:00 UTC - i'll make a point to swing by http://eavesdrop.openstack.org/#QA_Team_Office_hours | 15:33 |
| cmurphy | ++ | 15:33 |
| *** awalende has quit IRC | 15:37 | |
| *** jamesmcarthur has joined #openstack-keystone | 15:42 | |
| *** shyamb has quit IRC | 15:42 | |
| *** erus has quit IRC | 15:44 | |
| *** erus has joined #openstack-keystone | 15:45 | |
| *** jamesmcarthur has quit IRC | 15:46 | |
| *** jamesmcarthur has joined #openstack-keystone | 15:47 | |
| *** markvoelker has joined #openstack-keystone | 15:47 | |
| *** itlinux_ has joined #openstack-keystone | 15:54 | |
| *** itlinux has quit IRC | 15:56 | |
| *** jistr|call is now known as jistr | 16:00 | |
| *** itlinux_ has quit IRC | 16:07 | |
| *** dave-mccowan has quit IRC | 16:17 | |
| *** markvoelker has quit IRC | 16:21 | |
| openstackgerrit | Lance Bragstad proposed openstack/keystone master: Remove protocol policies from v3cloudsample.json https://review.openstack.org/625357 | 16:29 |
| *** itlinux has joined #openstack-keystone | 16:32 | |
| *** jmlowe has quit IRC | 16:43 | |
| openstackgerrit | Merged openstack/oslo.policy master: Provide more specific error when namespace is missing https://review.openstack.org/639822 | 16:46 |
| *** itlinux has quit IRC | 16:56 | |
| *** itlinux has joined #openstack-keystone | 16:56 | |
| *** markvoelker has joined #openstack-keystone | 17:18 | |
| *** raildo has quit IRC | 17:19 | |
| *** erus has quit IRC | 17:22 | |
| *** erus has joined #openstack-keystone | 17:23 | |
| *** jamesmcarthur has quit IRC | 17:27 | |
| *** jamesmcarthur has joined #openstack-keystone | 17:37 | |
| *** raildo has joined #openstack-keystone | 17:43 | |
| *** gyee has joined #openstack-keystone | 17:44 | |
| *** jamesmcarthur has quit IRC | 17:45 | |
| *** vishakha has quit IRC | 17:46 | |
| *** jamesmcarthur has joined #openstack-keystone | 17:48 | |
| *** raildo has quit IRC | 17:48 | |
| *** markvoelker has quit IRC | 17:51 | |
| *** jamesmcarthur has quit IRC | 17:52 | |
| *** jamesmcarthur has joined #openstack-keystone | 17:53 | |
| *** jmlowe has joined #openstack-keystone | 18:02 | |
| kmalloc | lbragstad, i'd like some eyes on https://review.openstack.org/#/c/605043 i would like to land that if we can. the functional test will come from SDK consuming it. | 18:13 |
| *** itlinux has quit IRC | 18:13 | |
| kmalloc | lbragstad: but we're at somewhat of an impasse, we can't consume it without landing it and releasing it. | 18:13 |
| lbragstad | hmm | 18:15 |
| kmalloc | it has unit tests, it does what it should do as far as i can tell. | 18:15 |
| *** itlinux has joined #openstack-keystone | 18:15 | |
| lbragstad | kmalloc have you tested it out yet? | 18:15 |
| kmalloc | yeah, locally | 18:15 |
| kmalloc | and the interface in session wont change even if we need to change some stuff under the hood. | 18:16 |
| kmalloc | it is also opt-in. | 18:16 |
| kmalloc | please feel free to test it locally as well. | 18:16 |
| lbragstad | looks like the release note is the only documentation | 18:16 |
| lbragstad | yeah - i'd like to | 18:16 |
| kmalloc | yeah, i'm inclined to add more documentation on it when we consume it in SDK. | 18:17 |
| kmalloc | mostly want to have a clear functional test landing with the documentation. | 18:17 |
| kmalloc | but we can add the docs earlier as well (followup patch) | 18:18 |
| kmalloc | it's super straightforward consuming it. | 18:18 |
| kmalloc | i also admit the tests were a little dated, but i have it unblocked so we can have real eyes on it as well. I plan to revisit local tests soon(ish) setting up an environment for it. | 18:19 |
| kmalloc | lbragstad: this is an important feature for SDK and zuul. | 18:20 |
| kmalloc | alsow orking on getting a fresh zuul response that isn't 2+ months old | 18:21 |
| *** jamesmcarthur has quit IRC | 18:37 | |
| *** erus has quit IRC | 18:37 | |
| *** erus has joined #openstack-keystone | 18:37 | |
| *** prashkre has joined #openstack-keystone | 18:41 | |
| *** markvoelker has joined #openstack-keystone | 18:47 | |
| *** prashkre_ has joined #openstack-keystone | 18:49 | |
| *** prashkre has quit IRC | 18:52 | |
| *** mvkr has quit IRC | 18:52 | |
| lbragstad | kmalloc i spun up the change, testing it locally | 18:58 |
| lbragstad | i have an instance of client with a session that's using the semaphore | 18:58 |
| lbragstad | how did you test it locally? | 18:58 |
| kmalloc | basically set the concurrency and rate limit | 18:59 |
| kmalloc | and then watch the traffic | 18:59 |
| kmalloc | tcpdump or similar | 18:59 |
| lbragstad | yep - i have a concurrency of 2 and delay of 5 | 18:59 |
| kmalloc | we could add a debug log for it i guess | 19:00 |
| kmalloc | i used tcpdump and just watched the traffic | 19:00 |
| *** itlinux has quit IRC | 19:07 | |
| kmalloc | and you need to send enough requests to hit the rate-limit/concurrency limit | 19:09 |
| openstackgerrit | Merged openstack/keystone master: Add tests for project users interacting with services https://review.openstack.org/620623 | 19:10 |
| *** rexor has joined #openstack-keystone | 19:13 | |
| *** pcaruana has quit IRC | 19:14 | |
| *** whoami-rajat has quit IRC | 19:20 | |
| *** markvoelker has quit IRC | 19:21 | |
| lbragstad | ok - so i have a concurrency of 2 and delay of 5? | 19:22 |
| kmalloc | yeah, that should mean you have at most 2 concurrent requests | 19:24 |
| kmalloc | and a rate_limit of 5? | 19:25 |
| kmalloc | not a delay. | 19:25 |
| kmalloc | so max 2 concurrent requests and 5 requests per second. | 19:25 |
| lbragstad | https://review.openstack.org/#/c/605043/6/keystoneauth1/_fair_semaphore.py@34 says rate_delay? | 19:29 |
| lbragstad | so - if i really want to throttle this down | 19:29 |
| lbragstad | i should do concurrency of 1 and 0? | 19:30 |
| kmalloc | or rate of 0.5 | 19:30 |
| lbragstad | which would do? | 19:30 |
| kmalloc | it's a float so you can do 1 request ever 2 seconds for example | 19:30 |
| lbragstad | ahh | 19:30 |
| lbragstad | got it | 19:30 |
| kmalloc | :) | 19:30 |
| lbragstad | ok - so i did FairSemaphore(1, 0.5) | 19:32 |
| lbragstad | and i made 10 requests before apparently getting rate limited | 19:32 |
| lbragstad | ah nvm | 19:33 |
| kmalloc | got it? | 19:33 |
| kmalloc | this stuff is hard(tm) to test. | 19:33 |
| lbragstad | maybe? | 19:34 |
| lbragstad | so if i set a delay of 0.1 is should make things even slower | 19:35 |
| *** erus has quit IRC | 19:35 | |
| lbragstad | once every 10 seconds? | 19:35 |
| kmalloc | it should | 19:36 |
| *** erus has joined #openstack-keystone | 19:36 | |
| kmalloc | and 0.01 should be one every 100 seconds. | 19:36 |
| lbragstad | hmm | 19:36 |
| kmalloc | this should be set though the adapter/session not directly on the fair semaphore. | 19:38 |
| kmalloc | the fair semaphore does things with a delay | 19:38 |
| kmalloc | the adapter takes a rate_limit option | 19:38 |
| lbragstad | https://pasted.tech/pastes/1ea741d6916e7e79946d9b372c5e16ac7c26fd20.raw | 19:38 |
| kmalloc | if you're setting values on the semaphore, you are probably getting the inverse | 19:38 |
| kmalloc | ah, ok so yeah, you're setting values on the semaphore | 19:38 |
| lbragstad | i'm setting the values on the semaphone and rebuilding the sessions | 19:39 |
| kmalloc | ahhh ok | 19:39 |
| kmalloc | so then you are talking about direct delay | 19:39 |
| lbragstad | that snippet should get rate limited, right? | 19:39 |
| kmalloc | https://git.openstack.org/cgit/openstack/keystoneauth/tree/keystoneauth1/adapter.py?h=refs/changes/43/605043/6#n96 | 19:39 |
| kmalloc | it should be rate limited. but at... sec | 19:39 |
| lbragstad | the vm i'm using has 8 cores | 19:40 |
| lbragstad | if that matters | 19:40 |
| kmalloc | no. | 19:40 |
| kmalloc | the rate_delay on the semaphore should od the inverse of the rate_limit on the adapert | 19:40 |
| kmalloc | so 0.1 should be waiting 0.1 seconds between requests | 19:41 |
| kmalloc | a rate_delay of 10 should be a delay of 10s. | 19:41 |
| lbragstad | bah | 19:41 |
| kmalloc | the fair semaphore is an internal interface | 19:41 |
| lbragstad | s = _fair_semaphore.FairSemaphore(1, 10) | 19:41 |
| kmalloc | yeah that should force a 10s delay. | 19:42 |
| lbragstad | i thought the session made you pass it in? | 19:42 |
| kmalloc | no. | 19:42 |
| *** raildo has joined #openstack-keystone | 19:42 | |
| kmalloc | https://git.openstack.org/cgit/openstack/keystoneauth/tree/keystoneauth1/adapter.py?h=refs/changes/43/605043/6#n163 | 19:43 |
| kmalloc | if you use the adapter, you get the fairsemaphore | 19:43 |
| kmalloc | depending on values sent in | 19:43 |
| lbragstad | ah | 19:43 |
| kmalloc | session allows for passing in a semaphore if you want | 19:43 |
| lbragstad | that's what i was doing | 19:43 |
| kmalloc | but that is an artifact of how the adapter(s) work | 19:43 |
| kmalloc | the adapter passes the semafore to the session as a kwarg. | 19:47 |
| openstackgerrit | Colleen Murphy proposed openstack/keystonemiddleware master: Add validation of app cred access rules https://review.openstack.org/633369 | 19:49 |
| lbragstad | mmm | 19:53 |
| lbragstad | https://pasted.tech/pastes/d1c6dd7eed93f04ae58dc17d4688da6f43671b55.raw | 19:53 |
| lbragstad | ^ reworked | 19:53 |
| lbragstad | that's not working for me, but i'm also building the session manually | 19:55 |
| lbragstad | well - actually | 19:55 |
| lbragstad | it's required by the Adapter constructor | 19:55 |
| lbragstad | the usability of that seems a little backwards? | 19:56 |
| *** jamesmcarthur has joined #openstack-keystone | 19:56 | |
| lbragstad | if you have to specify arguments to the adapter's constructor for rate limits *and* supply a session? | 19:56 |
| *** prashkre_ has quit IRC | 19:58 | |
| kmalloc | hm. | 20:06 |
| kmalloc | well, it is meant to be an opt-in on the adapter | 20:07 |
| kmalloc | adapter is added to the session | 20:07 |
| lbragstad | https://pasted.tech/pastes/cd2adc7035545f1de50606f1efdb3a0af5eddc57.raw | 20:07 |
| kmalloc | you in theory can ahve multiple adapters per session afair | 20:07 |
| lbragstad | gives me https://pasted.tech/pastes/1f7f8de74ca36fa364db6b7704f9510ccb7e5aaf.raw | 20:08 |
| kmalloc | so 10 requests per second | 20:08 |
| lbragstad | my rate limiting args are being ignored | 20:08 |
| kmalloc | 1 concurrency | 20:08 |
| kmalloc | is what i read that as the intention | 20:09 |
| lbragstad | they're not attributes on the adapt | 20:09 |
| lbragstad | adapter* | 20:09 |
| kmalloc | rate_delay is 0.1, which is what is on the semaphore | 20:09 |
| kmalloc | 10 /s = 0.1 delay | 20:09 |
| kmalloc | that looks correct to me | 20:10 |
| * lbragstad shakes head | 20:10 | |
| lbragstad | ok - yeah... that works | 20:10 |
| lbragstad | setting to 0.1 slows things way down | 20:10 |
| kmalloc | then the rate limit is doing what it should | 20:10 |
| kmalloc | concurrency is when you deal with multithreading | 20:11 |
| lbragstad | took me a while to understand the aesthetics | 20:11 |
| kmalloc | yeah sorry, you were down in the internal interface :) | 20:11 |
| kmalloc | and i know that is a little confusing | 20:11 |
| kmalloc | we tried to make it more straight forward faceing the user(s) | 20:12 |
| kmalloc | rate_limit is N Requests / Sec | 20:12 |
| lbragstad | that should be in the name :) | 20:12 |
| lbragstad | concurrency is pretty standard | 20:12 |
| kmalloc | well, rate limit means to me X requests per Y Time (1s would be my default assumption) | 20:13 |
| kmalloc | i am happy to be wrong and we can re-name it if you want. | 20:13 |
| *** erus has quit IRC | 20:13 | |
| *** erus has joined #openstack-keystone | 20:14 | |
| lbragstad | so rate_limit=0.5 translates to 1 request every 2 seconds | 20:16 |
| kmalloc | yes | 20:16 |
| lbragstad | and rate_limit=10 translates to 1 request every 0.1 seconds | 20:16 |
| kmalloc | correct. | 20:17 |
| kmalloc | or 10 per second. | 20:17 |
| kmalloc | and 0.5 is 1/2 reuqest per second, which really means one every 2 seconds because 1/2 a request... lol | 20:17 |
| lbragstad | right... | 20:17 |
| lbragstad | i'm just wondering if this is a me thing or if there is a different way we can relay the same information | 20:18 |
| kmalloc | not sure how to make it more intuative | 20:18 |
| openstackgerrit | Colleen Murphy proposed openstack/keystonemiddleware master: Fix debug tox environment https://review.openstack.org/640183 | 20:18 |
| kmalloc | best i could say is rename it to: requests_per_second | 20:18 |
| kmalloc | but i like this as a future looking where we might allow X requests over Y time (both configurable) | 20:19 |
| lbragstad | sure | 20:19 |
| kmalloc | easier than trying to do wonky math to figure out 75 requests over a 30 minute window | 20:19 |
| kmalloc | (that'd be 2.5 today, but saying 75 over 1800 might be more direct math wise) | 20:20 |
| kmalloc | and easier for a consumer to understand | 20:21 |
| kmalloc | and we do evenly space the requests over the period of time, so no one can currently do something like rate_limit = 1000 and put 1000 requests in the first .01 seconds | 20:23 |
| lbragstad | psh - let's just use canned defaults | 20:25 |
| lbragstad | "would you like rate limiting? ok, small, medium, or large? are you sure we can't super-size that for you? | 20:26 |
| kmalloc | :P | 20:26 |
| kmalloc | "Would you like us to concurrently make you fries with that?" | 20:26 |
| kmalloc | i'm pretty happy with the functionality of this | 20:27 |
| lbragstad | +2 | 20:27 |
| kmalloc | it has a slight issue though, it does force a slight serialization of all requests in a massivly concurrent environment that we previously didn;t have | 20:28 |
| kmalloc | that is because of the mutex | 20:28 |
| kmalloc | it should have next to zero impact. | 20:28 |
| kmalloc | but just a known quantity | 20:28 |
| lbragstad | ok - posted my comments | 20:30 |
| kmalloc | thnx | 20:31 |
| kmalloc | lbragstad: ok, so i'll commit to additional docs ASAP on this, but any other concerns? | 20:33 |
| kmalloc | tyring to get this in for the upcoming release. | 20:34 |
| kmalloc | if possible. | 20:34 |
| kmalloc | for S-cycle that is. | 20:34 |
| kmalloc | the functionality that is* | 20:34 |
| lbragstad | i don't think so | 20:34 |
| kmalloc | well we have a current passing zuul | 20:35 |
| kmalloc | if it's ok that docs land early next week / outside of the S-cycle cutoff, we can +2/+A this and get it spun into stein | 20:35 |
| kmalloc | it'll def. help out the infra folks and simplify a lot of their code. | 20:36 |
| lbragstad | do you know thomas' irc nick? | 20:36 |
| kmalloc | nope. i don't :( | 20:36 |
| lbragstad | does mordred? | 20:36 |
| kmalloc | maybe | 20:36 |
| kmalloc | cmurphy: might | 20:37 |
| kmalloc | he has an @suse.com email address | 20:37 |
| lbragstad | oh - good call | 20:37 |
| kmalloc | mordred, cmurphy: ^ either know Thomas Bechtold's IRC nic? | 20:37 |
| lbragstad | i'd just like him to ack the comments | 20:37 |
| kmalloc | we're trying to close the loop and land the fair semaphore/rate limit code in KSA | 20:37 |
| lbragstad | stepping away for a minute | 20:41 |
| *** erus has quit IRC | 20:41 | |
| *** erus has joined #openstack-keystone | 20:42 | |
| cmurphy | kmalloc: toabctl | 20:43 |
| cmurphy | kmalloc: he's CET though so probably won't respond tonight | 20:47 |
| mordred | kmalloc, lbragstad: fwiw - the ksa rate limit argument is the inverse of what the nodepool rate lmit is - becaus that one is expressed in, delay between requests (converting between the two is a simple 1/value luckily) - but setting rate limit of 0.01 - and that being slower than 0.001 ... hurt people's heads | 20:55 |
| mordred | but I see in the scrollback that you discussed rate delay as well :) | 20:56 |
| *** dave-mccowan has joined #openstack-keystone | 21:11 | |
| lbragstad | :) | 21:12 |
| *** dave-mccowan has quit IRC | 21:22 | |
| *** erus has quit IRC | 21:23 | |
| *** erus has joined #openstack-keystone | 21:24 | |
| *** markvoelker has joined #openstack-keystone | 21:48 | |
| *** erus has quit IRC | 21:48 | |
| *** erus has joined #openstack-keystone | 21:48 | |
| *** mchlumsky has quit IRC | 21:50 | |
| lbragstad | kmalloc mordred if you want to respin that patch for the kwarg comment and whatnot, i think we can get it in this release | 21:58 |
| lbragstad | even though cmurphy just released ksa | 21:58 |
| cmurphy | you'd have to ask the release team really nicely, the deadline's today | 21:59 |
| *** fiddletwix has joined #openstack-keystone | 21:59 | |
| mordred | ls | 22:02 |
| mordred | gah | 22:02 |
| mordred | kmalloc: I make kwargs comment update patch real quick | 22:05 |
| mordred | lbragstad: which thing are we responning for? | 22:06 |
| *** erus has quit IRC | 22:06 | |
| mordred | just updating the docstring? | 22:07 |
| lbragstad | thomas had a comment on the doc string, the default kwarg, and some looping stuff | 22:07 |
| *** erus has joined #openstack-keystone | 22:07 | |
| *** erus has quit IRC | 22:13 | |
| *** erus has joined #openstack-keystone | 22:14 | |
| openstackgerrit | Monty Taylor proposed openstack/keystoneauth master: Add support for client-side rate limiting https://review.openstack.org/605043 | 22:14 |
| mordred | lbragstad, cmurphy, kmalloc: ^^ | 22:14 |
| mordred | how do those changes look? | 22:14 |
| lbragstad | good - testing it locally | 22:15 |
| mordred | is it possible for time.sleep() to not actually sleep as long as you ask it to? | 22:16 |
| lbragstad | i'm not sure | 22:18 |
| lbragstad | er - i'm not sure why that would be? | 22:18 |
| mordred | because computers are devious and like lying? | 22:19 |
| mordred | (I was thinking maybe it was originally coded as a loop like that as a defense against sleep being mischevious? | 22:19 |
| *** markvoelker has quit IRC | 22:22 | |
| lbragstad | oh | 22:22 |
| lbragstad | interesting | 22:22 |
| lbragstad | what would you do in the case of a computer being devious? | 22:22 |
| lbragstad | log something? | 22:22 |
| * lbragstad has to run quick | 22:23 | |
| *** erus has quit IRC | 22:23 | |
| lbragstad | i'll keep an eye on that patch is someone else +Ws it | 22:24 |
| lbragstad | and i'll see if we can get a new ksa release out | 22:24 |
| *** erus has joined #openstack-keystone | 22:24 | |
| mordred | coolio! | 22:25 |
| kmalloc | +2/+A | 22:28 |
| kmalloc | mordred: i think the only case where sleep() doesn't really sleep long enough is if something like ntp ticks forward | 22:29 |
| *** erus has quit IRC | 22:30 | |
| mordred | yeah - and I don't think that loop would correct for that | 22:30 |
| clarkb | kmalloc: sleep is interrupted by any signal iirc | 22:30 |
| *** erus has joined #openstack-keystone | 22:31 | |
| kmalloc | clarkb: true. | 22:32 |
| *** raildo has quit IRC | 22:42 | |
| *** tkajinam has joined #openstack-keystone | 22:55 | |
| *** awalende has joined #openstack-keystone | 23:05 | |
| *** awalende has quit IRC | 23:10 | |
| *** vishwanathj has quit IRC | 23:14 | |
| *** markvoelker has joined #openstack-keystone | 23:18 | |
| * mnaser looks at past conversation | 23:20 | |
| * mnaser checks channel name | 23:20 | |
| kmalloc | mnaser: funny how there are crossovers in functionality | 23:24 |
| *** markvoelker has quit IRC | 23:24 | |
| *** markvoelker has joined #openstack-keystone | 23:24 | |
| *** erus has quit IRC | 23:31 | |
| *** erus has joined #openstack-keystone | 23:32 | |
| lbragstad | yeah - probably could have had that in sdk ;) | 23:34 |
| *** itlinux has joined #openstack-keystone | 23:39 | |
| *** itlinux_ has joined #openstack-keystone | 23:41 | |
| *** itlinux has quit IRC | 23:43 | |
| *** erus has quit IRC | 23:45 | |
| *** jamesmcarthur has quit IRC | 23:46 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!