| *** spatel has quit IRC | 00:04 | |
| openstackgerrit | Merged openstack/octavia stable/train: Fix healthmanager not update amphora health when LB disable https://review.opendev.org/724113 | 00:34 |
|---|---|---|
| *** gthiemon1e has quit IRC | 01:05 | |
| *** gthiemonge has joined #openstack-lbaas | 01:06 | |
| *** gregwork has quit IRC | 01:33 | |
| *** spatel has joined #openstack-lbaas | 01:50 | |
| *** spatel has quit IRC | 01:54 | |
| *** gthiemonge has quit IRC | 02:09 | |
| *** gthiemonge has joined #openstack-lbaas | 02:09 | |
| *** ramishra has joined #openstack-lbaas | 02:20 | |
| *** spatel has joined #openstack-lbaas | 02:23 | |
| *** sapd1_x has joined #openstack-lbaas | 03:12 | |
| *** sapd1_x has quit IRC | 03:17 | |
| *** sapd1_x has joined #openstack-lbaas | 03:31 | |
| *** psachin has joined #openstack-lbaas | 03:33 | |
| *** yamamoto has joined #openstack-lbaas | 03:35 | |
| *** yamamoto has quit IRC | 03:38 | |
| *** sapd1_x has quit IRC | 04:40 | |
| *** gcheresh has joined #openstack-lbaas | 04:51 | |
| *** sapd1_x has joined #openstack-lbaas | 04:54 | |
| *** spatel has quit IRC | 05:01 | |
| openstackgerrit | Merged openstack/octavia-dashboard master: Add py38 package metadata https://review.opendev.org/724991 | 05:31 |
| *** vishalmanchanda has joined #openstack-lbaas | 05:43 | |
| *** rpittau|afk is now known as rpittau | 06:33 | |
| *** ccamposr has joined #openstack-lbaas | 07:06 | |
| *** ccamposr__ has joined #openstack-lbaas | 07:08 | |
| *** ccamposr has quit IRC | 07:11 | |
| *** sapd1_x has quit IRC | 07:18 | |
| *** gcheresh has quit IRC | 07:30 | |
| *** sapd1_x has joined #openstack-lbaas | 07:30 | |
| *** gcheresh has joined #openstack-lbaas | 07:45 | |
| *** stingrayza has quit IRC | 08:15 | |
| *** stingrayza has joined #openstack-lbaas | 08:18 | |
| *** sapd1_x has quit IRC | 08:20 | |
| *** stingrayza has quit IRC | 08:32 | |
| *** sapd1_x has joined #openstack-lbaas | 08:34 | |
| *** stingrayza has joined #openstack-lbaas | 08:49 | |
| openstackgerrit | Carlos Goncalves proposed openstack/octavia master: Migrate grenade job to native Zuul v3 https://review.opendev.org/725098 | 08:52 |
| *** salmankhan has joined #openstack-lbaas | 09:00 | |
| *** stingrayza has quit IRC | 09:03 | |
| *** salmankhan1 has joined #openstack-lbaas | 09:04 | |
| *** salmankhan has quit IRC | 09:05 | |
| *** salmankhan1 is now known as salmankhan | 09:05 | |
| *** stingrayza has joined #openstack-lbaas | 09:07 | |
| *** stingrayza has quit IRC | 09:29 | |
| *** yamamoto has joined #openstack-lbaas | 09:50 | |
| *** yamamoto has quit IRC | 09:58 | |
| *** stingrayza has joined #openstack-lbaas | 10:01 | |
| *** gcheresh has quit IRC | 10:04 | |
| *** armax has joined #openstack-lbaas | 10:09 | |
| *** rpittau is now known as rpittau|bbl | 10:14 | |
| *** gcheresh has joined #openstack-lbaas | 10:53 | |
| *** kevinz has quit IRC | 11:04 | |
| *** yamamoto has joined #openstack-lbaas | 11:37 | |
| *** yamamoto has quit IRC | 11:37 | |
| *** servagem has joined #openstack-lbaas | 11:39 | |
| openstackgerrit | Carlos Goncalves proposed openstack/octavia master: Migrate grenade job to native Zuul v3 https://review.opendev.org/725098 | 12:01 |
| *** dmellado has quit IRC | 12:10 | |
| *** sapd1_x has quit IRC | 12:16 | |
| *** dmellado has joined #openstack-lbaas | 12:16 | |
| *** rpittau|bbl is now known as rpittau | 12:22 | |
| rm_work | hmmmm seems like we have network_tasks "PlugNetworks" and "UnPlugNetworks" that are only ever used in tests, lol | 12:29 |
| rm_work | can just be removed I think <_< | 12:29 |
| rm_work | keep getting blindly updated but aren't actually necessary anymore lol | 12:29 |
| cgoncalves | + GetPlumbedNetworks, GetMemberPorts | 12:35 |
| cgoncalves | + RetrievePortIDsOnAmphoraExceptLBNetwork, PlugPorts | 12:35 |
| rm_work | lol | 12:36 |
| rm_work | also all unused? | 12:36 |
| rm_work | would be great to run through and delete ALL our unused cruft | 12:36 |
| cgoncalves | unless something constructs the class names, only tests use them | 12:36 |
| rm_work | yeah | 12:36 |
| rm_work | hmm... | 12:36 |
| rm_work | in allowed_address_pairs.py it seems like `_delete_vip_security_group` and `_delete_security_group` are named backwards :D | 12:37 |
| rm_work | `_delete_vip_security_group` is a generic function that takes in a security_group name and deletes it... | 12:37 |
| rm_work | `_delete_security_group` does all kinds of logic about finding the sg for a vip, then calling the other one with that name XD | 12:38 |
| rm_work | gonna flip that in one of my patches... | 12:39 |
| *** spatel has joined #openstack-lbaas | 12:53 | |
| *** jamesden_ is now known as jamesdenton | 13:07 | |
| *** yamamoto has joined #openstack-lbaas | 13:11 | |
| *** psachin has quit IRC | 13:14 | |
| *** yamamoto has quit IRC | 13:16 | |
| *** KeithMnemonic has joined #openstack-lbaas | 13:31 | |
| KeithMnemonic | is this the right place to ask about this tempest test https://opendev.org/openstack/octavia-tempest-plugin/src/branch/master/octavia_tempest_plugin/tests/scenario/v2/test_traffic_ops.py | 13:32 |
| colin- | aap references give me ptsd :| | 13:33 |
| KeithMnemonic | I am getting a failure and was hoping to see if there were any docs describing what this test does in detail | 13:36 |
| *** salmankhan1 has joined #openstack-lbaas | 13:39 | |
| *** salmankhan has quit IRC | 13:40 | |
| *** salmankhan1 is now known as salmankhan | 13:40 | |
| johnsom | rm_work: failover deletes some cruft. We could start by merging it. Grin | 13:46 |
| rm_work | this is stuff i'm seeing on top of your failover patch | 13:47 |
| rm_work | like, i've got it checked out already | 13:47 |
| rm_work | so, in the case of the things i'm mentioning, merging that will not help :D | 13:48 |
| *** TrevorV has joined #openstack-lbaas | 13:48 | |
| johnsom | KeithMnemonic: It does what it says, runs tests that put traffic through the load balancer. | 13:48 |
| rm_work | johnsom: about to push a new take on the shared security group thing | 13:48 |
| johnsom | Ok | 13:49 |
| rm_work | still will rely on neutron adding the new sharing mode, but that looks likely to happen | 13:49 |
| johnsom | There is more I can remove when I do the job board version of failover too | 13:49 |
| KeithMnemonic | johnsom: so my failure is timeout jsut doing the resource_setup, but other tests run fine | 13:49 |
| johnsom | Hmmm, timeout in setup sounds like a sick cloud or the tempest config is not set right for the limitations of the cloud. | 13:51 |
| KeithMnemonic | specifically " cls.webserver1_response" | 13:51 |
| KeithMnemonic | but i can test webservers running on node manually and they work so trying to see how this is different | 13:52 |
| KeithMnemonic | somewhere here I guess https://opendev.org/openstack/octavia-tempest-plugin/src/branch/master/octavia_tempest_plugin/tests/test_base.py | 13:59 |
| rm_work | johnsom: also our "add_security_group_to_port" function is really "replace_security_groups_on_port" lol | 14:35 |
| johnsom | Yeah, I am slowly rewriting the whole network driver interface. I want it to be simple methods, not kitchen sink methods | 14:37 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Share the LB security group with the LB owner https://review.opendev.org/723735 | 14:38 |
| rm_work | johnsom: ^^ | 14:38 |
| rm_work | so, that makes a new SG "lb_access-$LBID" | 14:38 |
| rm_work | name pending :D | 14:38 |
| rm_work | empty, added to all ports we use (except management) | 14:39 |
| rm_work | so, vrrp ports, member ports... maybe vip port? actually not sure but it shouldn't matter | 14:39 |
| rm_work | i put it on top of failover | 14:39 |
| *** rpittau is now known as rpittau|afk | 16:08 | |
| johnsom | Legacy Free! Thanks Carlos for tracking the grenade changes and getting that gate updated. | 16:21 |
| *** gcheresh has quit IRC | 16:22 | |
| cgoncalves | o/ | 16:25 |
| cgoncalves | now I need to go fix grenade in stable/stein :/ it is blocking the gate | 16:26 |
| johnsom | I see that | 16:26 |
| openstackgerrit | Carlos Goncalves proposed openstack/octavia stable/stein: Do not run Tempest in octavia-grenade job https://review.opendev.org/725337 | 16:54 |
| cgoncalves | ^ shooting in the dark | 16:54 |
| *** vishalmanchanda has quit IRC | 17:15 | |
| *** salmankhan has quit IRC | 17:21 | |
| dawzon | johnsom I'm assuming that tls_versions will have to be represented as CSV in MySQL, correct? | 17:34 |
| *** maciejjozefczyk has joined #openstack-lbaas | 17:45 | |
| johnsom | dawzon Yeah or a json list. I would recommend not using python "pickle" as it changes formats over the different python versions and causes upgrade/backward compatibility issues. | 17:46 |
| *** maciejjozefczyk has quit IRC | 17:53 | |
| *** maciejjozefczyk has joined #openstack-lbaas | 17:57 | |
| *** maciejjozefczyk has quit IRC | 18:45 | |
| dawzon | johnsom Where would I put the code that turns the value from the DB back into a JSON array? | 18:53 |
| johnsom | That is a good question. | 18:54 |
| dawzon | I'm looking in the listener controller right now | 18:54 |
| dawzon | Maybe manipulating the result object?https://github.com/openstack/octavia/blob/07a93de755c0ec461f51b94e3465708700ff1549/octavia/api/v2/controllers/listener.py#L66 | 18:55 |
| johnsom | Well, I think it would be in one of the object layers. | 18:56 |
| cgoncalves | stable/stein octavia-grenade job fix is ready for reviews: https://review.opendev.org/#/c/725337/ | 18:59 |
| dawzon | Ah, found it I think https://github.com/openstack/octavia/blob/07a93de755c0ec461f51b94e3465708700ff1549/octavia/api/v2/types/listener.py#L68 | 18:59 |
| johnsom | Yeah, follow the allowed cidrs model | 19:00 |
| openstackgerrit | Merged openstack/octavia stable/train: Fix py3 amphora-agent cert-rotation type bug https://review.opendev.org/723389 | 19:05 |
| nmickus | johnsom dawzon Are we going ahead with switching from colon separated to comma seperated for the tls versions? | 19:21 |
| dawzon | nmickus The goal is to make it a JSON array on the client side | 19:22 |
| nmickus | dawzon would that just be passed into the cli using a comma separated list wrapped in brackets? | 19:26 |
| dawzon | Hmm | 19:26 |
| dawzon | However it works for allowed_cidrs | 19:27 |
| dawzon | It looks like allowed_cidrs uses repetitions of the flag to do multiple | 19:28 |
| *** gcheresh has joined #openstack-lbaas | 19:46 | |
| *** markguz_ has joined #openstack-lbaas | 19:59 | |
| markguz_ | Hi lbaas folks. I'm having problems with my magnum created k8s cluster deploying a load balancer. when i install nginx-ingress it starts | 20:01 |
| markguz_ | the install process but always fails with an error about using trust-scoped tokens to create another token....error: http://paste.openstack.org/show/793081/ | 20:02 |
| markguz_ | I've done some research. i enabled redelegation in keystone and allow_rescope_scoped | 20:04 |
| markguz_ | octavia version is U release from git | 20:05 |
| johnsom | markguz_ That is interesting. That code is fairly new and I haven't seen that before | 20:10 |
| markguz_ | johnsom: Yeah, i'm scratching my head a bit. when using magnum to deploy the k8s cluster, the load balancers deploy just fine. But when creating the ingress controller if blows up with that error | 20:11 |
| johnsom | markguz_ It is this patch and code that is causing the trouble: https://review.opendev.org/#/c/721550/4/octavia/common/clients.py | 20:15 |
| johnsom | The token being used by whatever called the Octavia API must be a special type. | 20:15 |
| johnsom | markguz_ Can you open a bug for us on this? Then, to get you working while we figure it out, change this setting in your octavia.conf: allow_invisible_resource_usage to true | 20:19 |
| markguz_ | johnsom: sure. which section does that go in ? DEFAULT ? | 20:19 |
| johnsom | [networking] | 20:20 |
| markguz_ | thx | 20:20 |
| markguz_ | johnsom: well that option definitely worked anyway | 20:23 |
| johnsom | Yeah, not a good long term solution, but will allow you to continue until we fix it | 20:24 |
| markguz_ | what does it do? disable auth? | 20:25 |
| johnsom | When the Octavia API verifies that the subnet submitted exists, with that "False" it attempts to use the user context to check the subnet. With it "True" it will use the Octavia service account context to check the subnet. | 20:26 |
| markguz_ | ah ok. | 20:28 |
| *** servagem has quit IRC | 20:28 | |
| markguz_ | johnsom: https://storyboard.openstack.org/#!/story/2007619 | 20:28 |
| johnsom | Thank you! | 20:29 |
| markguz_ | np | 20:29 |
| *** threestrands has joined #openstack-lbaas | 20:34 | |
| johnsom | lxkong FYI: https://storyboard.openstack.org/#!/story/2007619 | 20:43 |
| *** gcheresh has quit IRC | 21:15 | |
| openstackgerrit | Noah Mickus proposed openstack/python-octaviaclient master: Add the ability to specify TLS protocols for a listener https://review.opendev.org/722426 | 22:17 |
| *** rcernin has joined #openstack-lbaas | 22:22 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Refactor the failover flows https://review.opendev.org/705317 | 22:37 |
| *** spatel has quit IRC | 22:49 | |
| *** TrevorV has quit IRC | 23:00 | |
| *** ccamposr has joined #openstack-lbaas | 23:17 | |
| *** ccamposr__ has quit IRC | 23:20 | |
| rm_work | johnsom: just a rebase? or did you do more | 23:24 |
| johnsom | Found a log message I had dropped, so put it back | 23:25 |
| rm_work | seems a test is failing now | 23:25 |
| rm_work | typo? :D | 23:25 |
| rm_work | https://748441c9eeeb89eb36db-b17eeb704755458252956ad3afd603fe.ssl.cf5.rackcdn.com/705317/37/check/openstack-tox-py36/54f933f/testr_results.html | 23:25 |
| rm_work | hmm maybe not | 23:25 |
| rm_work | amphora.load_balancer_id -> amphora.load_balancer.id ? | 23:26 |
| johnsom | Probably | 23:26 |
| rm_work | hmm no, that should exist too | 23:26 |
| rm_work | not sure | 23:26 |
| rm_work | doesn't the model have both | 23:26 |
| johnsom | I am will have a look | 23:26 |
| rm_work | yeah it's literally used right above that | 23:26 |
| rm_work | yeah will stop wildly speculating and let you handle it :D | 23:26 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!