| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2020-01-06 | |||
| 22:39:31 | sean-k-mooney | i ckeckt the kolla ansible source and i dont see a change there | |
| 22:39:53 | born2bake | so why its not default then | |
| 22:40:11 | sean-k-mooney | there have been a few chagnes for ipv6 but the they all kept backward compatabity | |
| 22:41:50 | sean-k-mooney | the url format is defiend here in kolla ansible https://github.com/openstack/kolla-ansible/blob/stable/train/ansible/roles/nova/defaults/main.yml#L137-L139 | |
| 22:42:54 | sean-k-mooney | could you have modifed nova_api_port to be 8774/compute ? | |
| 22:43:05 | sean-k-mooney | or could someone have? | |
| 22:43:10 | born2bake | no I didnt | |
| 22:43:20 | born2bake | if you check answers in bug report | |
| 22:44:41 | born2bake | https://docs.openstack.org/api-ref/compute/ | |
| 22:44:55 | born2bake | Service URLs - For instance, if the service url is http://mycompute.pvt/compute/v2.1 then the full API call for /servers is http://mycompute.pvt/compute/v2.1/servers. | |
| 22:45:46 | sean-k-mooney | yes so in your case the comptue service api is registered in keystone as http://10.0.225.254:8774/v2.1 | |
| 22:45:58 | sean-k-mooney | so the full url would be http://10.0.225.254:8774/v2.1/servers | |
| 22:46:39 | sean-k-mooney | which appers to be the endpoint rancher is trying to reach | |
| 22:49:13 | sean-k-mooney | there is no requirement to have compute in the url and we did not change that in train | |
| 22:49:54 | born2bake | yes, rancher is trying to reach http://10.0.225.254:8774/v2.1/servers | |
| 22:50:01 | born2bake | when I am trying to open it using curl | |
| 22:50:04 | born2bake | 404 error | |
| 22:50:16 | born2bake | if I am trying to curl http://10.0.225.254:8774/compute/v2.1/servers | |
| 22:50:18 | born2bake | it works | |
| 22:51:01 | sean-k-mooney | it sound like this is an issue with the haproxy config or apache config | |
| 22:51:21 | born2bake | so then it is related to kolla-ansible then? O_o | |
| 22:55:04 | sean-k-mooney | i belive so | |
| 22:56:08 | sean-k-mooney | given you can apperatly invoke the api fine at the other location then this is a configuration issue rather then and issue with the service | |
| 22:59:04 | born2bake | in bug report they mentioned that its related to nova | |
| 22:59:14 | born2bake | nova said that its related to rancher :) | |
| 23:00:50 | sean-k-mooney | nova does not know the api endpoint its api is exposed at so its not a nova issue | |
| 23:01:06 | sean-k-mooney | that is stored in keystone with is configured by the installation tool | |
| 23:01:52 | sean-k-mooney | my haproxy config for my kolla-ansible train deploy ment looks liek this http://paste.openstack.org/show/788097/ | |
| 23:01:56 | sean-k-mooney | is yours the same | |
| 23:04:16 | born2bake | yup same | |
| 23:06:08 | sean-k-mooney | ok well its late for me so im going to call it a night. if i think of anything else ill update the bug | |
| 23:06:19 | sean-k-mooney | does the openstack cli work by the way | |
| 23:06:31 | sean-k-mooney | e.g. if you just do an openstack server list or soemthing? | |
| 23:07:10 | sean-k-mooney | or "openstack keypair list" | |
| 23:08:38 | born2bake | everything works smoothly | |
| 23:08:50 | sean-k-mooney | if it does you can add --debug e.g. "openstack --debug keypair list" and it will print out the raw data and effectivly thec curl command | |
| 23:08:58 | sean-k-mooney | ok can you try with debug | |
| 23:09:16 | sean-k-mooney | i just need the REQ: line | |
| 23:09:32 | sean-k-mooney | it will look like this "REQ: curl -g -i -X GET http://192.168.1.254:8774/v2.1/os-keypairs -H "Accept: application/json" -H "User-Agent: python-novaclient" -H "X-Auth-Token: {SHA256}ac329731a91c5c72b3a7ca81c0579cbe5ecc8d20028518e9d94735efe05121ba" -H "X-OpenStack-Nova-API-Version: 2.1"" | |
| 23:09:47 | sean-k-mooney | thats not really the auth token by the way just a hash of it | |
| 23:10:30 | born2bake | REQ: curl -g -i -X GET http://10.0.225.254:8774/v2.1/os-keypairs -H "Accept: application/json" -H "User-Agent: python-novaclient" -H "X-Auth-Token: {SHA256}152dcee94db8af1a3f509137abb21fec6d7bb484433729f9b192989274bb6d95" -H "X-OpenStack-Nova-API-Version: 2.1" | |
| 23:10:31 | born2bake | Starting new HTTP connection (1): 10.0.225.254:8774 | |
| 23:10:58 | sean-k-mooney | ya so the client is using the same url | |
| 23:11:44 | sean-k-mooney | so i would guess the 403 is likely because the keystone auth token expired or did not have the correct permissions | |
| 23:11:55 | sean-k-mooney | when rancher invoked the endpoint | |
| 23:12:25 | born2bake | it worked fine with stein | |
| 23:12:40 | born2bake | i installed train 1.5 weeks ago | |
| 23:14:15 | sean-k-mooney | yes but given the openstack client work fine and what you have already told me it looks like the openstack in stall is working correctly | |
| 23:15:49 | sean-k-mooney | if it was a nova issue i would expect it to also break the openstack client which seams to be working fine | |
| #openstack-nova - 2020-01-07 | |||
| 00:38:15 | openstackgerrit | Dustin Cowles proposed openstack/nova master: WIP: Provider Config File: Enable loading and merging of provider configs https://review.opendev.org/693460 | |
| 00:43:44 | openstackgerrit | Victor Morales proposed openstack/nova master: Update Testing NUMA documentation https://review.opendev.org/701301 | |
| 00:51:30 | openstackgerrit | Merged openstack/nova master: docs: Add note about an image signature validation limitation when using rbd https://review.opendev.org/700152 | |
| 01:30:25 | openstackgerrit | Merged openstack/nova master: [api-ref] Fix the incorrect link https://review.opendev.org/700362 | |
| 01:30:30 | openstackgerrit | Merged openstack/nova master: Fix an invalid assertIsNotNone statement https://review.opendev.org/700603 | |
| 05:29:26 | openstackgerrit | Merged openstack/nova master: Report trait 'COMPUTE_IMAGE_TYPE_PLOOP' https://review.opendev.org/698132 | |
| 05:39:03 | openstackgerrit | Merged openstack/nova master: Ensure source service is up before resizing/migrating https://review.opendev.org/700062 | |
| 05:39:10 | openstackgerrit | Merged openstack/nova master: Add recreate test for bug 1855927 https://review.opendev.org/699045 | |
| 05:39:10 | openstack | bug 1855927 in OpenStack Compute (nova) "_poll_unconfirmed_resizes may not retry later if confirm_resize fails in API" [Low,In progress] https://launchpad.net/bugs/1855927 - Assigned to Matt Riedemann (mriedem) | |
| 07:18:56 | openstackgerrit | jichenjc proposed openstack/nova master: downgrade when host does not support capabilities https://review.opendev.org/697228 | |
| 07:19:07 | openstackgerrit | jichenjc proposed openstack/nova master: libvirt: avoid cpu check at s390x arch https://review.opendev.org/696228 | |
| 08:18:05 | gibi | eandersson: I don't know about tools in nova to emit fake notifications but we have the sample notifications stored so you can emit those with your small tool you linked | |
| 08:40:14 | openstackgerrit | waleed mousa proposed openstack/os-vif master: [Follow Up] OVS DPDK port representors support https://review.opendev.org/665965 | |
| 10:26:18 | bauzas | gibi: hola, hope you had good vacations | |
| 10:26:30 | bauzas | (and happy new year !) | |
| 10:27:52 | gibi | bauzas: hi! happy new year! yes, I had a good one. I hope the same for you | |
| 10:28:13 | bauzas | oh yeah, took a week in a loving resort :) | |
| 10:28:32 | bauzas | gibi: if you have needing reviews, tell me which ones | |
| 10:29:23 | gibi | bauzas: the current bottom of the qos + live migration patch series is still being worked on by me right now I will ping you when it is ready | |
| 10:29:30 | bauzas | cool | |
| 10:29:41 | bauzas | I'm also working on the placement audit method | |
| 10:29:49 | bauzas | and then, I'll work on some specs | |
| 10:30:34 | gibi | bauzas: there is a rename / refactor patch that is ready I think https://review.opendev.org/#/c/698735/ | |
| 10:30:55 | gibi | bauzas: if you need my eyes on the placement audit patch then let me know | |
| 10:31:14 | bauzas | thanks | |
| 10:31:30 | bauzas | gibi: and looking at your change today afternoon | |
| 10:31:53 | gibi | cool | |
| 10:31:54 | gibi | thanks | |
| 10:46:36 | openstackgerrit | Shilpa Devharakar proposed openstack/nova master: Handle new is_volume_backend join column query https://review.opendev.org/694462 | |
| 10:46:37 | openstackgerrit | Shilpa Devharakar proposed openstack/nova master: Instance object changes for the new 'is_volume_backed' expected_attr https://review.opendev.org/694463 | |
| 10:46:37 | openstackgerrit | Shilpa Devharakar proposed openstack/nova master: Ignore root_gb if instance is booted from volume https://review.opendev.org/612626 | |
| 10:59:56 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Func test for failed and aborted live migration https://review.opendev.org/699065 | |
| 10:59:57 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Use common server create function for qos func tests https://review.opendev.org/701353 | |
| 11:03:35 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Enable live migration with qos ports https://review.opendev.org/699066 | |
| 11:12:16 | openstackgerrit | Merged openstack/nova-specs master: Amend cross-cell-resize spec https://review.opendev.org/696197 | |
| 11:43:16 | stephenfin | bauzas: Can we cold migrate an instance with vGPU? | |
| 11:43:33 | stephenfin | I think so (since it's a rebuild) but want to be sure before I put it in writing :) | |
| 12:09:53 | openstackgerrit | Lee Yarwood proposed openstack/os-traits master: Add COMPUTE_BFV_RESCUE trait https://review.opendev.org/694033 | |
| 12:25:18 | luyao | stephenfin, artom, sean-k-mooney : FYI, spec ‘support live migration with vpmem’ is updated and still waiting for review, looking forward to your comments. :) | |
| 12:26:12 | luyao | https://review.opendev.org/#/c/695863/ | |
| 12:26:56 | sean-k-mooney | luyao: thanks we are just gettting back but i guess now is a good a time as any to take a look | |
| 12:28:03 | luyao | sean-k-mooney: Hah, thanks. :D | |
| 12:28:39 | sean-k-mooney | stephenfin: yes you can but ther is an issue with it | |
| 12:29:03 | stephenfin | sean-k-mooney: tell me more | |
| 12:29:08 | sean-k-mooney | stephenfin: you can cold mighrate but then you have to either rebuild or hard reboot it | |
| 12:29:24 | sean-k-mooney | a cold migration does not set up the mdev properly | |
| 12:29:44 | sean-k-mooney | there is a note in the docs and bauzas is aware of the issue | |
| 12:29:53 | sean-k-mooney | i dont think he has had time to fix it | |
| 12:30:01 | sean-k-mooney | bauzas: ^ is correct right | |
| 12:30:22 | sean-k-mooney | stephenfin: i rememebr talking about this with bauzas just before the break but i could be mistaken | |
| 12:31:14 | stephenfin | sean-k-mooney: Ah, indeed. "Cold migrating an instance to another host will have the same problem as resize. If you want to migrate an instance, make sure to rebuild it after the migration." | |