| *** amotoki has joined #openstack-lbaas | 01:42 | |
| openstackgerrit | Joe Mills proposed openstack/neutron-lbaas: api test: centralize health monitor protocol https://review.openstack.org/419324 | 02:10 |
|---|---|---|
| *** links has joined #openstack-lbaas | 02:25 | |
| *** links has quit IRC | 02:38 | |
| *** links has joined #openstack-lbaas | 02:46 | |
| *** bana_k has joined #openstack-lbaas | 03:16 | |
| *** sticker has joined #openstack-lbaas | 03:24 | |
| *** bana_k has quit IRC | 03:29 | |
| openstackgerrit | Tuan Luong-Anh proposed openstack/neutron-lbaas: Replaces uuid.uuid4 with uuidutils.generate_uuid() https://review.openstack.org/406768 | 03:57 |
| *** amotoki_ has joined #openstack-lbaas | 05:09 | |
| *** amotoki has quit IRC | 05:12 | |
| *** kevo has joined #openstack-lbaas | 06:37 | |
| *** saju_m has joined #openstack-lbaas | 06:47 | |
| openstackgerrit | JingLiu proposed openstack/octavia: Fix file mode https://review.openstack.org/420519 | 06:50 |
| *** greghaynes has quit IRC | 06:52 | |
| *** greghaynes has joined #openstack-lbaas | 06:58 | |
| *** greghaynes has quit IRC | 07:04 | |
| *** yamamoto has quit IRC | 07:04 | |
| *** greghaynes has joined #openstack-lbaas | 07:15 | |
| *** nrado has joined #openstack-lbaas | 07:16 | |
| *** kevo has quit IRC | 07:16 | |
| *** tesseract has joined #openstack-lbaas | 07:21 | |
| *** nmagnezi has joined #openstack-lbaas | 07:25 | |
| *** kevo has joined #openstack-lbaas | 07:31 | |
| *** gcheresh_ has joined #openstack-lbaas | 07:33 | |
| *** pcaruana has joined #openstack-lbaas | 07:34 | |
| *** jsheeren has joined #openstack-lbaas | 07:52 | |
| *** kobis has joined #openstack-lbaas | 07:56 | |
| *** eezhova has joined #openstack-lbaas | 07:59 | |
| openstackgerrit | JingLiu proposed openstack/neutron-lbaas: Fix file mode https://review.openstack.org/420551 | 08:00 |
| *** amotoki has joined #openstack-lbaas | 08:07 | |
| *** amotoki_ has quit IRC | 08:11 | |
| *** eezhova has quit IRC | 08:47 | |
| *** eezhova has joined #openstack-lbaas | 09:12 | |
| *** kevo has quit IRC | 09:32 | |
| *** amotoki has quit IRC | 10:01 | |
| *** anilvenkata has joined #openstack-lbaas | 10:03 | |
| *** amotoki has joined #openstack-lbaas | 10:08 | |
| *** amotoki has quit IRC | 10:24 | |
| *** saju_m has quit IRC | 10:28 | |
| *** saju_m has joined #openstack-lbaas | 10:55 | |
| *** nmagnezi has quit IRC | 11:04 | |
| *** nmagnezi has joined #openstack-lbaas | 11:04 | |
| *** nmagnezi has quit IRC | 11:06 | |
| *** nmagnezi_ has joined #openstack-lbaas | 11:14 | |
| *** nmagnezi_ is now known as nmagnezi | 11:15 | |
| *** amotoki has joined #openstack-lbaas | 11:37 | |
| *** nrado has quit IRC | 11:59 | |
| *** beagles_afk is now known as beagles | 13:08 | |
| *** nrado has joined #openstack-lbaas | 13:33 | |
| nmagnezi | johnsom, ping, a question about static routes in amphora (when you are here) | 14:10 |
| nmagnezi | johnsom, I created a loadbalancer with a subnet that has static routes, expected to get those in the file created by https://github.com/openstack/octavia/blob/master/octavia/amphorae/backends/agent/api_server/templates/plug_port_ethX.conf.j2#L26-L27 | 14:12 |
| nmagnezi | johnsom, but didn't get any | 14:12 |
| nmagnezi | johnsom, i tried it both with static routes who are part of the subnet (lbaas-loadbalancer-create --name test <subnet>) | 14:12 |
| nmagnezi | johnsom, also, tried to add this as part of the router (which seems like a long shot but i still tried it) | 14:13 |
| nmagnezi | johnsom, the reasoning behind all this is to have an equivalent placement for static routes in the centos templates. | 14:13 |
| nmagnezi | johnsom, so as a first step, I tried to see how it looks with master code + ubuntu based amphora | 14:14 |
| *** links has quit IRC | 14:30 | |
| *** beagles is now known as beagles_brb | 14:57 | |
| *** jsheeren has quit IRC | 15:00 | |
| *** anilvenkata has quit IRC | 15:10 | |
| *** fnaval has joined #openstack-lbaas | 15:25 | |
| *** beagles_brb is now known as beagles | 15:41 | |
| *** ducttape_ has joined #openstack-lbaas | 15:52 | |
| *** ducttape_ has quit IRC | 15:55 | |
| *** amotoki has quit IRC | 15:57 | |
| *** gcheresh_ has quit IRC | 16:00 | |
| *** kobis has quit IRC | 16:07 | |
| *** beardedeagle has joined #openstack-lbaas | 16:08 | |
| *** armax has joined #openstack-lbaas | 16:26 | |
| *** ducttape_ has joined #openstack-lbaas | 16:34 | |
| *** eezhova has quit IRC | 16:37 | |
| *** cody-somerville has joined #openstack-lbaas | 16:39 | |
| *** cody-somerville has quit IRC | 16:39 | |
| *** cody-somerville has joined #openstack-lbaas | 16:39 | |
| *** cody-somerville has quit IRC | 16:39 | |
| johnsom | nmagnezi The static routes come in from the "host_routes" on the subnet used: http://developer.openstack.org/api-ref/networking/v2/?expanded=create-port-detail,create-subnet-detail#create-subnet | 16:44 |
| *** nrado has quit IRC | 16:45 | |
| *** cody-somerville has joined #openstack-lbaas | 16:47 | |
| *** cody-somerville has quit IRC | 16:47 | |
| *** cody-somerville has joined #openstack-lbaas | 16:47 | |
| *** amotoki has joined #openstack-lbaas | 16:57 | |
| *** ducttape_ has quit IRC | 16:58 | |
| *** nrado has joined #openstack-lbaas | 16:58 | |
| *** armax has quit IRC | 17:05 | |
| *** eezhova has joined #openstack-lbaas | 17:12 | |
| *** tesseract has quit IRC | 17:29 | |
| *** johnsom has quit IRC | 17:30 | |
| *** johnsom has joined #openstack-lbaas | 17:30 | |
| *** frippe75 has joined #openstack-lbaas | 17:40 | |
| *** armax has joined #openstack-lbaas | 17:42 | |
| frippe75 | maybe a pure neutron question? looking in the routers schema there is a STATUS field which indicates that a router is ACTIVE. In the lbaas_loadbalancer the "same field" is called OPERATING_STATUS and it is ONLINE. Is there no namingconvention? | 17:43 |
| johnsom | Well, OPERATING_STATUS != STATUS | 17:43 |
| johnsom | If I remember right, STATUS is more like our PROVISIONING_STATUS | 17:44 |
| frippe75 | ok thought that it could be something like that. | 17:44 |
| johnsom | https://github.com/openstack/octavia/blob/master/octavia/common/constants.py#L64 | 17:44 |
| johnsom | Yeah, which uses ACTIVE | 17:44 |
| frippe75 | Ok so ther is a provisioning_status for loadbalancer and it is ACTIVE. Missed that. but it's actually named provisioning_status and not status. But maybe to be clear as to the operating_status. | 17:45 |
| johnsom | OPERATING_STATUS is actively monitored and the current functional status of the object. Correct. We tried to make it more clear what it is representing | 17:46 |
| *** armax has quit IRC | 17:47 | |
| *** kevo has joined #openstack-lbaas | 17:47 | |
| frippe75 | ok thanks! took the "easy way" of copying some code and spent some time wondering why it didn't work... | 17:47 |
| frippe75 | :-) | 17:47 |
| johnsom | NP | 17:48 |
| nrado | Hi, can someone explain to me the workflow of creating a TLS listener with barbican? Why do I need the service_auth in the neutron-lbass.conf and why do I have to add the admin user according to these instructions: https://wiki.openstack.org/wiki/Network/LBaaS/docs/how-to-create-tls-loadbalancer | 17:50 |
| johnsom | Hi nrado | 17:51 |
| johnsom | Take a look at: http://docs.openstack.org/developer/octavia/guides/basic-cookbook.html#deploy-a-tls-terminated-https-load-balancer | 17:51 |
| johnsom | Barbican has changed how we can access containers, so TLS listeners have changed a bit. We are still working with barbican to improve this in a future release. | 17:52 |
| johnsom | nrado This may also help with your user question: http://docs.openstack.org/developer/octavia/guides/dev-quick-start.html#create-octavia-user | 17:53 |
| *** ducttape_ has joined #openstack-lbaas | 17:59 | |
| *** ducttape_ has quit IRC | 18:03 | |
| *** bana_k has joined #openstack-lbaas | 18:05 | |
| nrado | Thx johnsom, but it's still not clear to me. What is the neutron or rather the octavio user used for? Let's say I've a non-admin user who wants to create a tls listener. | 18:07 |
| nrado | Is the neutron or octvaia user the one who is responsible for the listener as a consumer? I'm asking because it might be that I have some problems with the RBAC of barbican. | 18:07 |
| nrado | That's why I need to know what is happening in the background | 18:07 |
| nrado | I was able to to create a TLS listener with an admin user, but as soon as I use a normal user, it's forbidden | 18:08 |
| johnsom | Right. The issue is the new ACL model in barbican. When you store your secrets in barbican they are under your project_id. However, when using them for a listener in octavia/lbaas either neturon or octavia will need access to those secrets containers. | 18:10 |
| johnsom | If everything is admin, we are fine. I.e. the container is created as admin and lbaas/octavia are running as admin we can access the containers in barbican. | 18:10 |
| johnsom | However, when the containers are created under a different project_id we have to authorize the account lbaas/octavia is running under to have access to those containers in barbican and the contents of the containers. | 18:11 |
| johnsom | This is the ACL lines in that example. | 18:12 |
| johnsom | Over time, the way barbican handles this has changed, thus the wiki docs being a bit old. | 18:12 |
| johnsom | In the future we are working to allow the listener create API to do this authorization on behalf of the requesting user, but we are not there yet. | 18:13 |
| johnsom | Does that help? | 18:14 |
| nrado | Yeah, sounds fine to me. So I need to run the "openstack acl user add" commands in order to authorize the neutron or octavia user to have access to the containers | 18:16 |
| *** bank_ has joined #openstack-lbaas | 18:17 | |
| nrado | I'll give it a try now | 18:17 |
| johnsom | Yep | 18:17 |
| nrado | Hope it will work, thanks for helping me | 18:17 |
| johnsom | No problem, sorry this has been a bit of a moving target for us. Maybe in Pike we can get this cleaned up | 18:18 |
| *** bana_k has quit IRC | 18:19 | |
| *** eezhova has quit IRC | 18:23 | |
| *** bank_ has quit IRC | 18:29 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia: Remove an erroneous MarkHealthMonitorActiveInDB task https://review.openstack.org/409403 | 18:36 |
| *** cody-somerville has quit IRC | 18:48 | |
| johnsom | Hmm, this subunit failure is going to be tricky to track down. I just ran a test on 457 testr output files I had local and none of them failed like the gate.... | 18:50 |
| *** eezhova has joined #openstack-lbaas | 18:54 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia: Do not merge: Save original testr output https://review.openstack.org/420907 | 18:55 |
| *** cody-somerville has joined #openstack-lbaas | 19:10 | |
| *** cody-somerville has quit IRC | 19:10 | |
| *** cody-somerville has joined #openstack-lbaas | 19:10 | |
| *** bana_k has joined #openstack-lbaas | 19:11 | |
| *** ducttape_ has joined #openstack-lbaas | 19:15 | |
| *** ducttape_ has quit IRC | 19:19 | |
| *** cody-somerville has quit IRC | 19:20 | |
| *** ducttape_ has joined #openstack-lbaas | 19:25 | |
| *** ducttape_ has quit IRC | 19:30 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia: Do not merge: Save original testr output https://review.openstack.org/420907 | 19:52 |
| openstackgerrit | Michael Johnson proposed openstack/octavia: Do not merge: Save original testr output https://review.openstack.org/420907 | 19:54 |
| *** jerrygb has joined #openstack-lbaas | 19:57 | |
| nmagnezi | johnsom, hey | 19:58 |
| johnsom | Hi | 19:59 |
| nmagnezi | johnsom, re ^^ (static routes), i think this is what i tried | 19:59 |
| nmagnezi | that didn't work for me | 19:59 |
| nmagnezi | i added a static route to a subnet | 19:59 |
| nmagnezi | and created a loadbalancer in that subnet | 19:59 |
| nmagnezi | created a listener and checked the cfg files | 19:59 |
| nmagnezi | didn't find any static route | 19:59 |
| johnsom | Hmmm, that is how I tested it. Let me see if I can try again | 20:00 |
| johnsom | amp is booting | 20:04 |
| johnsom | nmagnezi where are you checking? | 20:05 |
| *** bana_k has quit IRC | 20:07 | |
| nmagnezi | johnsom, devstack master branch | 20:07 |
| nmagnezi | johnsom, i mean octavia master :) | 20:07 |
| nmagnezi | johnsom, ubuntu amp | 20:07 |
| johnsom | Yeah, where were you checking in the amp? | 20:07 |
| nmagnezi | johnsom, oh, sec | 20:07 |
| nmagnezi | johnsom, http://paste.openstack.org/show/595102/ | 20:08 |
| *** bana_k has joined #openstack-lbaas | 20:10 | |
| *** nrado has left #openstack-lbaas | 20:11 | |
| johnsom | nmagnezi Hmmm, well, on my system I'm getting another error I will need to look into, but I do have the route in /etc/netns/amphora-haproxy/network/interfaces.d/eth1.cfg | 20:11 |
| johnsom | http://paste.openstack.org/show/TPYDUj6UcWT4X0V8Nl3q/ | 20:12 |
| nmagnezi | johnsom, can you please paste the cmd you used to create the subnet? maybe i'm doing something wrong there | 20:12 |
| johnsom | neutron subnet-update --host-route destination=99.99.99.0/24,nexthop=172.21.21.244 private-subnet | 20:12 |
| johnsom | neutron lbaas-loadbalancer-create --name lb1 private-subnet | 20:12 |
| johnsom | That is what I did | 20:13 |
| nmagnezi | johnsom, thanks! I'll check and let you know :) | 20:13 |
| johnsom | Now, to figure out what went wrong with the agent... (though I'm not sure what code I have checked out at the moment on this devstack) | 20:14 |
| nmagnezi | johnsom, that's the fun part :) | 20:17 |
| *** cody-somerville has joined #openstack-lbaas | 20:27 | |
| *** cody-somerville has quit IRC | 20:27 | |
| *** cody-somerville has joined #openstack-lbaas | 20:27 | |
| *** woodster_ has joined #openstack-lbaas | 20:32 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia: Do not merge: Save original testr output https://review.openstack.org/420907 | 20:32 |
| openstackgerrit | Michael Johnson proposed openstack/octavia: Do not merge: Save original testr output https://review.openstack.org/420907 | 20:33 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/neutron-lbaas: Updated from global requirements https://review.openstack.org/420843 | 20:34 |
| *** ducttape_ has joined #openstack-lbaas | 20:45 | |
| *** ducttape_ has quit IRC | 20:49 | |
| *** gcheresh_ has joined #openstack-lbaas | 20:53 | |
| *** eezhova has quit IRC | 21:29 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia: Do not merge: Save original testr output https://review.openstack.org/420907 | 21:30 |
| johnsom | Of course, with the log archiving the test won't fail... | 21:31 |
| *** ducttape_ has joined #openstack-lbaas | 21:35 | |
| *** frippe75 has quit IRC | 21:36 | |
| *** ducttape_ has quit IRC | 21:43 | |
| *** ducttape_ has joined #openstack-lbaas | 21:44 | |
| *** ducttape_ has quit IRC | 21:44 | |
| *** eezhova has joined #openstack-lbaas | 21:46 | |
| *** eezhova has quit IRC | 21:56 | |
| *** csomerville has joined #openstack-lbaas | 22:01 | |
| *** cody-somerville has quit IRC | 22:04 | |
| *** cody-somerville has joined #openstack-lbaas | 22:05 | |
| *** saju_m has quit IRC | 22:08 | |
| *** csomerville has quit IRC | 22:08 | |
| *** saju_m has joined #openstack-lbaas | 22:10 | |
| *** gcheresh_ has quit IRC | 22:13 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia: Archive the raw testrepository log https://review.openstack.org/420907 | 22:15 |
| *** yamamoto has joined #openstack-lbaas | 22:15 | |
| nmagnezi | johnsom, just tested it | 22:15 |
| johnsom | Well, I'm giving up and just adding that log archive to our gates. | 22:15 |
| nmagnezi | johnsom, not sure what happens but the loadbalancer ends up with ERROR state and the amphora gets deleted | 22:16 |
| johnsom | nmagnezi How did it go? I found my issue. I was passing in a bogus host route, so the plug was failing | 22:16 |
| nmagnezi | i should set a break point so I can debug | 22:16 |
| johnsom | nmagnezi Yeah, could be the same issue. If the route doesn't have a valid gw it fails. | 22:17 |
| nmagnezi | johnsom, i should try this stuff with valid routes :) | 22:17 |
| johnsom | nmagnezi FYI, you can stop the revert from running and cleaning up the amp. | 22:17 |
| nmagnezi | johnsom, anyhow I trust what you just pasted. I'll take this as example | 22:17 |
| nmagnezi | johnsom, by stopping the hk and hm services? | 22:17 |
| johnsom | nmagnezi To stop reverts, edit octavia/common/base_taskflow.py | 22:18 |
| johnsom | nmagnezi then add "never_resolve=True," to the tf_engines.load() | 22:18 |
| * nmagnezi takes notes | 22:19 | |
| johnsom | That makes taskflow just stop the flow if an error occurs that would cause a revert | 22:19 |
| nmagnezi | johnsom, that's a nifty trick :) | 22:19 |
| johnsom | I asked for it special from the taskflow folks | 22:19 |
| nmagnezi | johnsom, till now I just placed breakpoints (pdb) on specific location | 22:19 |
| johnsom | The name is a bit cryptic, but it works | 22:19 |
| nmagnezi | johnsom, for example after the amp instance is created | 22:20 |
| johnsom | I should really add a config setting for that.... | 22:20 |
| nmagnezi | johnsom, so I can use rsync and sync the latest amphora-agent code to the instance and restart the agent | 22:20 |
| nmagnezi | johnsom, a bit tricky, but it works | 22:20 |
| nmagnezi | johnsom, http://paste.openstack.org/show/595110/ :) | 22:22 |
| nmagnezi | johnsom, centos amp | 22:22 |
| johnsom | Nice!!!! | 22:22 |
| nmagnezi | johnsom, indeed! I just need to finish the jinja2 templates to include those static routes and ipv6 | 22:23 |
| nmagnezi | johnsom, doing my best to finish this week.. | 22:23 |
| johnsom | Please do. It would great to get this in | 22:23 |
| nmagnezi | johnsom, if you have some spare cycles (you don't) you can already have a look at the patch and give me an initial feedback | 22:24 |
| johnsom | I will make time | 22:24 |
| nmagnezi | johnsom, it would. I really want to make it. I'll keep you posted on this | 22:24 |
| johnsom | Ok, sounds good | 22:24 |
| nmagnezi | johnsom, okay, /me going to sleep it passed midnight | 22:25 |
| johnsom | Good night | 22:26 |
| nmagnezi | johnsom, good night | 22:26 |
| nmagnezi | johnsom, oh before i go, one thing i wanted to ask you about | 22:27 |
| nmagnezi | johnsom, https://review.openstack.org/#/c/331841/41/elements/amphora-agent/init-scripts/systemd/amphora-agent.service | 22:27 |
| nmagnezi | johnsom, so we either need to have separate files for ubuntu and centos | 22:27 |
| nmagnezi | johnsom, or we can omit the full patch and use | 22:28 |
| johnsom | Ah, the install path is different | 22:28 |
| nmagnezi | johnsom, or we can omit the full patch and use: amphora-image to trigger the alias | 22:28 |
| nmagnezi | yup | 22:28 |
| johnsom | nmagnezi Well, systemd requires the full path. It will fail without. | 22:28 |
| nmagnezi | i changed it just for my dev needs | 22:28 |
| nmagnezi | johnsom, ack, so we'll need additional file | 22:29 |
| johnsom | nmagnezi I think there is a way to force the install to /usr/local/bin. Would that be ok if I can pull up the way to override that? | 22:29 |
| nmagnezi | johnsom, mmm.. i dunno, i never tried such a thing. if it works i guess it is okay to have that | 22:31 |
| nmagnezi | johnsom, it might be a silly suggestion but maybe we can create a symlink ? | 22:31 |
| johnsom | nmagnezi Ok, I will research and push a patch while you sleep | 22:31 |
| nmagnezi | johnsom, thanks a lot :) | 22:32 |
| johnsom | Yeah, a sym link would probably work too. I will investigate | 22:32 |
| nmagnezi | johnsom, aye. good night | 22:33 |
| *** jerrygb has quit IRC | 22:34 | |
| *** ducttape_ has joined #openstack-lbaas | 22:45 | |
| *** ducttape_ has quit IRC | 22:49 | |
| *** cody-somerville has quit IRC | 22:59 | |
| johnsom | Six runs with the testr archive and no failures... ugh. Well, we can just merge the archive code and use it when it fails again in the future | 23:00 |
| *** cody-somerville has joined #openstack-lbaas | 23:18 | |
| *** cody-somerville has quit IRC | 23:18 | |
| *** cody-somerville has joined #openstack-lbaas | 23:18 | |
| *** jerrygb has joined #openstack-lbaas | 23:22 | |
| *** jerrygb has quit IRC | 23:26 | |
| *** guest has joined #openstack-lbaas | 23:29 | |
| *** jerrygb has joined #openstack-lbaas | 23:34 | |
| *** ducttape_ has joined #openstack-lbaas | 23:35 | |
| openstackgerrit | Merged openstack/octavia: Updated from global requirements https://review.openstack.org/420845 | 23:36 |
| *** csomerville has joined #openstack-lbaas | 23:39 | |
| *** cody-somerville has quit IRC | 23:42 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia: WIP - Fix the amphora-agent support for RH based Linux flavors https://review.openstack.org/331841 | 23:54 |
| *** ankur-gupta-f2 has left #openstack-lbaas | 23:57 | |
| *** fnaval has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!