| *** osmanlicilegi is now known as Guest0 | 04:33 | |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: CI: Fix missing BASE_DISTRO for test-ovn.sh https://review.opendev.org/c/openstack/kolla-ansible/+/895132 | 05:31 |
|---|---|---|
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: CI: Fix missing BASE_DISTRO for test-ovn.sh https://review.opendev.org/c/openstack/kolla-ansible/+/895132 | 05:45 |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible stable/yoga: CI: Fix missing BASE_DISTRO for test-ovn.sh https://review.opendev.org/c/openstack/kolla-ansible/+/895024 | 05:52 |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible stable/yoga: CI: Fix missing BASE_DISTRO for test-ovn.sh https://review.opendev.org/c/openstack/kolla-ansible/+/895024 | 05:53 |
| mnasiadka | oh boy, seems Ubuntu has changed paths long ago to /var/run/ovn (just like RHEL does) | 07:01 |
| mnasiadka | let me fix that | 07:01 |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: ovn: Fix broken deployment/reconfig on Ubuntu https://review.opendev.org/c/openstack/kolla-ansible/+/895132 | 07:06 |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: ovn: Fix broken deployment/reconfig on Ubuntu https://review.opendev.org/c/openstack/kolla-ansible/+/895132 | 07:36 |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: ovn: Fix broken deployment/reconfig on Ubuntu https://review.opendev.org/c/openstack/kolla-ansible/+/895132 | 07:38 |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: CI: Change OVN jobs to voting https://review.opendev.org/c/openstack/kolla-ansible/+/894914 | 07:43 |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: ovn: Fix broken deployment/reconfig on Ubuntu https://review.opendev.org/c/openstack/kolla-ansible/+/895132 | 07:44 |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: CI: Change OVN jobs to voting https://review.opendev.org/c/openstack/kolla-ansible/+/894914 | 07:56 |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: CI: Change OVN jobs to voting https://review.opendev.org/c/openstack/kolla-ansible/+/894914 | 07:56 |
| kevko | Regarding base_distro in kolla-ansible | 08:06 |
| kevko | We should drop this and instead of that we should prepare images to be "same" | 08:06 |
| kevko | Symlinks, edit scripts etc ... | 08:07 |
| kevko | Or propose base_distro on service level | 08:07 |
| mnasiadka | We've been discussing that before, and we agreed it's rather a PTG topic | 08:07 |
| mnasiadka | because symlinking python3.something to just python is a bit dramatic I guess | 08:07 |
| hrw | kevko: ha ha ha, funny idea ;D | 08:08 |
| mnasiadka | and we really need that only for horizon I think | 08:08 |
| kevko | These days I am fighting with this | 08:08 |
| hrw | debian vs rhel ways of doing things, naming dirs etc | 08:08 |
| kevko | Fluent for example - td-agent vs fluent user | 08:08 |
| mnasiadka | we're extremely good at ranting | 08:08 |
| kevko | Cron var spool vs I don't remember ..and similar | 08:09 |
| kevko | mnasiadka btw, I tried to redeploy ovn ..restarted nova/neutron and I have not working stack :D | 08:09 |
| kevko | I can't get ovn working again ..db is empty | 08:10 |
| kevko | Some advices ? :/ | 08:10 |
| mnasiadka | run the sync util from neutron | 08:10 |
| kevko | I've already done | 08:11 |
| kevko | Nothing | 08:11 |
| mnasiadka | neutron-ovn-db-sync-util | 08:11 |
| kevko | Yes ...no change | 08:11 |
| mnasiadka | and NB is still empty? | 08:11 |
| kevko | Yes :( | 08:11 |
| mnasiadka | So check sync util logs | 08:11 |
| kevko | I will paste here in few minutes | 08:12 |
| kevko | Now the morning coffee is more important | 08:12 |
| mnasiadka | kevko: I have a feeeling after the refactor we're not very good on catching kolla-build errors: https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_475/894948/1/check/kolla-build-rocky9/4759c08/kolla/build/000_FAILED_fluentd.log | 08:17 |
| mnasiadka | in the past we could understand what really happened ;-) | 08:17 |
| kevko | mnasiadka: this is the same error i've got with podman patchset !!!! | 08:21 |
| opendevreview | Rafal Lewandowski proposed openstack/kolla-ansible master: Added OVS precheck for neutron_plugin_agent setting https://review.opendev.org/c/openstack/kolla-ansible/+/895143 | 08:30 |
| opendevreview | Will Szumski proposed openstack/kolla-ansible master: Support exposing prometheus_server externally https://review.opendev.org/c/openstack/kolla-ansible/+/828695 | 08:46 |
| opendevreview | Christian Berendt proposed openstack/kolla-ansible master: ceilometer: process custom event_pipeline.yaml with merge_yaml https://review.opendev.org/c/openstack/kolla-ansible/+/894182 | 08:53 |
| opendevreview | Rafal Lewandowski proposed openstack/kolla-ansible master: Added OVS precheck for neutron_plugin_agent setting https://review.opendev.org/c/openstack/kolla-ansible/+/895143 | 09:01 |
| kevko | mnasiadka: i don't think it's refactor ...it is something very weird because if you checkout change for which you posted a log ..it will pass on localhost i think | 09:02 |
| opendevreview | Sven Kieske proposed openstack/kolla master: bump td-agent lts from v4 to v5 https://review.opendev.org/c/openstack/kolla/+/894948 | 09:16 |
| opendevreview | Pierre Riteau proposed openstack/kayobe stable/2023.1: Complete Skydive removal and Skyline addition https://review.opendev.org/c/openstack/kayobe/+/895025 | 09:23 |
| SvenKieske | looks like the td-agent renaming stuff could be not that much fallout, after all (I guess now I jinxed it oO) | 09:36 |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: ovn: Fix broken deployment/reconfig on Ubuntu https://review.opendev.org/c/openstack/kolla-ansible/+/895132 | 09:41 |
| opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: CI: Change OVN jobs to voting https://review.opendev.org/c/openstack/kolla-ansible/+/894914 | 09:43 |
| opendevreview | Rafal Lewandowski proposed openstack/kolla-ansible master: Added OVS precheck for neutron_plugin_agent setting https://review.opendev.org/c/openstack/kolla-ansible/+/895143 | 09:51 |
| mmalchuk | good morning Kolla o/ | 10:11 |
| mmalchuk | cores, please review and merge backports: https://review.opendev.org/q/I169e51c0f5c691518ada1837990b5bdd2a3d1481 | 10:11 |
| mmalchuk | mnasiadka could you please look at: https://review.opendev.org/q/Ice13759c406ff7cec91c9d740b0e0c43eda99510 | 10:12 |
| opendevreview | Rafal Lewandowski proposed openstack/kolla-ansible master: Added OVS precheck for neutron_plugin_agent setting https://review.opendev.org/c/openstack/kolla-ansible/+/895143 | 10:12 |
| mmalchuk | mnasiadka RP-1 needed here https://review.opendev.org/c/openstack/kolla-ansible/+/893765 ? | 10:13 |
| mnasiadka | it's failing :) | 10:14 |
| mmalchuk | zuul gives +1 | 10:15 |
| mmalchuk | anyway genconfig not affects CI | 10:16 |
| opendevreview | Michal Arbet proposed openstack/kolla master: Fix stderr logging when image build fail https://review.opendev.org/c/openstack/kolla/+/895157 | 10:31 |
| opendevreview | Michal Arbet proposed openstack/kolla master: Add podman for image building https://review.opendev.org/c/openstack/kolla/+/825791 | 10:33 |
| kevko | ha ! now i will be not blind ! | 10:33 |
| kevko | show me your secrets Zuul ! | 10:33 |
| * kevko laughs terrifyingly | 10:34 | |
| SvenKieske | :D | 10:34 |
| mmalchuk | lol | 10:42 |
| mnasiadka | kevko: nice | 10:44 |
| kevko | i knew it | 10:55 |
| kevko | podman is not problem | 10:55 |
| kevko | we have problem with eventlet | 10:55 |
| kevko | https://da1d82983230150eebb3-89d69bea015aaa7a4b0cf0e0890d5ac2.ssl.cf2.rackcdn.com/825791/21/check/kolla-ansible-rocky9-upgrade/f683cab/primary/logs/build/000_FAILED_glance-base.log | 10:55 |
| Vii | hi why is there no "collectd" for ubuntu version? why https://quay.io/repository/openstack.kolla/collectd?tab=tags. (quay.io/openstack.kolla/collectd:2023.1-ubuntu-jammy not found) | 10:55 |
| mmalchuk | becouse missing collectd-core package in Ubuntu | 10:57 |
| Vii | Ahh ok, thx | 10:57 |
| mmalchuk | the same issue with telegraf container | 10:57 |
| mnasiadka | https://review.opendev.org/c/openstack/kolla/+/884001 - seems to be buildable in B | 11:00 |
| mnasiadka | should we backport? let's try | 11:00 |
| opendevreview | Michal Nasiadka proposed openstack/kolla stable/2023.1: ubuntu: mark collectd and telegraf as buildable https://review.opendev.org/c/openstack/kolla/+/895026 | 11:00 |
| opendevreview | Rafal Lewandowski proposed openstack/kolla-ansible master: Added OVS precheck for neutron_plugin_agent setting https://review.opendev.org/c/openstack/kolla-ansible/+/895143 | 11:01 |
| mmalchuk | mnasiadka buildable | 11:01 |
| mnasiadka | Vii: should be fixed when 895026 merges | 11:01 |
| opendevreview | Sven Kieske proposed openstack/kolla master: bump td-agent lts from v4 to v5 https://review.opendev.org/c/openstack/kolla/+/894948 | 11:34 |
| opendevreview | Rafal Lewandowski proposed openstack/kolla-ansible master: Added ML2/OVN and ML2/OVS setting checks for neutron https://review.opendev.org/c/openstack/kolla-ansible/+/895143 | 11:38 |
| opendevreview | Jan Gutter proposed openstack/kolla-ansible master: debian: Remove kolla-ansible-debian-zun job https://review.opendev.org/c/openstack/kolla-ansible/+/895163 | 11:45 |
| SvenKieske | mhm, someone familiar with td-agent|fluentd-package internals? the "/var/lib/" path and the username has changed as well.. | 11:45 |
| jangutter | mnasiadka: happy with just disabling the zun-debian job for now, and next week we have one with a release note? https://review.opendev.org/c/openstack/kolla-ansible/+/895163 | 11:46 |
| kevko | mnasiadka: okay, i've used sync parameter repair and now working | 12:04 |
| andreboc | Hi Kollas, I am glad to see that you all are developing support for podman, however, I have a question regarding future support. Will podman and docker be supported equally when the final path is released, or will one solution remain recommended? | 12:08 |
| mnasiadka | equally | 12:11 |
| mnasiadka | at least for now | 12:11 |
| kevko | mnasiadka: why kolla [branch master] downloads antelope upper constraints and install service from master ? | 12:19 |
| opendevreview | Merged openstack/kolla stable/zed: Improve kolla-toolbox offline support. https://review.opendev.org/c/openstack/kolla/+/893851 | 12:19 |
| mnasiadka | kevko: where? | 12:21 |
| mnasiadka | it should use OPENSTACK_RELEASE variable, which is master on master branch if we're talking about building images | 12:22 |
| kevko | mnasiadka: here https://da1d82983230150eebb3-89d69bea015aaa7a4b0cf0e0890d5ac2.ssl.cf2.rackcdn.com/825791/21/check/kolla-ansible-rocky9-upgrade/f683cab/primary/logs/build/000_FAILED_glance-base.log | 12:26 |
| kevko | Check the glance version ..check requirements and check upper constraints | 12:26 |
| andreboc | mnasiadka: "at least for now" means you consider switching to podman (in long-term planning)? | 12:37 |
| mnasiadka | andreboc: no, we are not | 12:37 |
| opendevreview | Merged openstack/kolla stable/yoga: Improve kolla-toolbox offline support. https://review.opendev.org/c/openstack/kolla/+/893852 | 12:40 |
| opendevreview | Merged openstack/kolla stable/2023.1: Improve kolla-toolbox offline support. https://review.opendev.org/c/openstack/kolla/+/893849 | 12:40 |
| mnasiadka | kevko: well, I can see that we're using /requirements/upper-constraints.txt - which are from base, why do you think these are from antelope? | 12:40 |
| mnasiadka | seems we're taking them from zuul checkout of requirements | 12:42 |
| mnasiadka | kevko: maybe something wrong with depends-on on the podman change to an abandoned requirements patch? | 12:45 |
| kevko | mnasiadka: i dropped depends-on | 12:45 |
| kevko | now | 12:45 |
| kevko | will see | 12:45 |
| opendevreview | Michal Arbet proposed openstack/kolla master: Add podman for image building https://review.opendev.org/c/openstack/kolla/+/825791 | 12:45 |
| kevko | mnasiadka: but i don't think so | 12:46 |
| kevko | master requirements | 12:46 |
| kevko | 48d56ac82 (OpenStack Proposal Bot 2023-01-28 08:28:07 +0000 162) eventlet===0.33.3 | 12:46 |
| kevko | it is there from january | 12:46 |
| opendevreview | Merged openstack/kayobe stable/2023.1: Complete Skydive removal and Skyline addition https://review.opendev.org/c/openstack/kayobe/+/895025 | 12:46 |
| kevko | but log is saying | 12:47 |
| kevko | [0m | 12:47 |
| kevko | ERROR:kolla.common.utils.glance-base:The conflict is caused by: | 12:47 |
| kevko | glance 27.0.0.0b3.dev51 depends on eventlet>=0.33.3 | 12:47 |
| kevko | The user requested (constraint) eventlet===0.33.1 | 12:47 |
| kevko | To fix this you could try to: | 12:47 |
| kevko | 1. loosen the range of package versions you've specified | 12:47 |
| kevko | 2. remove package versions to allow pip attempt to solve the dependency conflict | 12:47 |
| kevko | glance 27 is master | 12:47 |
| kevko | The user requested (constraint) eventlet===0.33.1 >> which is antelope -> https://github.com/openstack/requirements/blob/2710414b4663b85b48ecbbba21f4f0ea42dc3df1/upper-constraints.txt#L158 | 12:47 |
| kevko | but master is https://github.com/openstack/requirements/blob/b7cf680233409b977c52ee3ba076ea67771ecf1a/upper-constraints.txt#L162 0.33.3 | 12:48 |
| kevko | that's why i think it is from antelope | 12:48 |
| kevko | this one https://tarballs.opendev.org/openstack/requirements/requirements-stable-2023.1.tar.gz | 12:54 |
| mnasiadka | well, it's the upgrade job - so we might be checking out antelope requirements I guess | 12:55 |
| kevko | mnasiadka: so why installing glance 27 ? | 12:56 |
| kevko | which is master ? | 12:56 |
| mnasiadka | because of https://github.com/openstack/kolla-ansible/blob/428acfe97fdfbc89fd4dea577160192647e6f6b5/tests/run.yml#L109 | 12:56 |
| mnasiadka | we use kolla-ansible ansible-collection-kolla and requirements from earlier branch | 12:56 |
| mnasiadka | and then we're using it for building master images | 12:56 |
| mnasiadka | question why it does not fail on other patchsets | 12:57 |
| mnasiadka | https://github.com/openstack/kolla/blob/64098b209cfab47b1cd23351c09e5de3ed275c46/roles/kolla-build-config/templates/kolla-build.conf.j2#L10 - that's the code to blame for this | 12:58 |
| mnasiadka | so it might be the problem with depends-on | 12:58 |
| mnasiadka | :) | 12:58 |
| kevko | what ? why we are installing master pip packages with requirements of earlier branch ? | 12:58 |
| opendevreview | Rafal Lewandowski proposed openstack/kolla-ansible master: Added ML2/OVN and ML2/OVS setting checks for neutron https://review.opendev.org/c/openstack/kolla-ansible/+/895143 | 12:59 |
| mnasiadka | because normally we're not using requirements that are checked out by Zuul | 12:59 |
| mnasiadka | but with depends-on - you made it to use them | 12:59 |
| mnasiadka | and that's most probably why it failed | 12:59 |
| mnasiadka | a bit of a corner case | 12:59 |
| kevko | let's see ..depends on removed | 13:03 |
| kevko | mnasiadka: the good thing is that I was forced to fix the error logging | 13:03 |
| kevko | :D | 13:03 |
| mnasiadka | kevko: see, there's always some bright side ;) | 13:03 |
| kevko | haha | 13:03 |
| kevko | mnasiadka: zuul 27 minutes running ...so it looks like it was the issue | 13:13 |
| kevko | mnasiadka: but this edge case should be fixed i think | 13:13 |
| mnasiadka | SvenKieske: seems you'll need to do a similar dance that we did in the past with changed fluentd binary name (https://github.com/openstack/kolla/commit/23f6d2ab011a1d664976ced95310d0dbc49d0d00) | 13:22 |
| opendevreview | Juan Pablo Suazo proposed openstack/kolla master: Adds TAAS Neutron plugin to support OVS port mirrors https://review.opendev.org/c/openstack/kolla/+/885151 | 13:23 |
| kevko | mnasiadka: SvenKieske: why we can't fix in image ? | 13:28 |
| mnasiadka | kevko: td-agent changed binary name from td-agent to fluentd - do you want a symbolic link? ;) | 13:29 |
| kevko | which td-agent , which fluentd -> zero or non-zero ...etc | 13:29 |
| kevko | mnasiadka: no, but we have startup script ..why we not check the situation and depends on situation do additional steps ? | 13:29 |
| mnasiadka | well, startup script is one thing, but the command that we're running is in kolla-ansible | 13:30 |
| mnasiadka | so we would need to call an intermediate script | 13:30 |
| SvenKieske | mnasiadka: thanks for the linl | 13:30 |
| SvenKieske | link* | 13:30 |
| mnasiadka | it might be easier to use the container label and leave it forever, since they change every major version :D | 13:30 |
| mnasiadka | I'm fine with either of those solutions, to be clear :) | 13:31 |
| SvenKieske | wasn't aware this changed in the past | 13:31 |
| SvenKieske | I'l think about it for a bit. | 13:31 |
| kevko | or, you can use script kolla-run-fluentd forever in kolla-ansible ..and forewer change only script | 13:31 |
| kevko | i just want to say that kolla-ansible should orchestrate things ..but "API" is the kolla side | 13:32 |
| SvenKieske | might be worth it, if they change this stuff every few releases.. | 13:41 |
| opendevreview | Jake Hutchinson proposed openstack/kolla-ansible master: Ironic parameter rework and default NTP server https://review.opendev.org/c/openstack/kolla-ansible/+/893031 | 13:53 |
| opendevreview | Juan Pablo Suazo proposed openstack/kolla-ansible master: Deploy Glance and Cinder Backup with S3 backend support https://review.opendev.org/c/openstack/kolla-ansible/+/844614 | 14:06 |
| opendevreview | Jake Hutchinson proposed openstack/kolla-ansible master: Ironic parameter rework and default NTP server https://review.opendev.org/c/openstack/kolla-ansible/+/893031 | 14:12 |
| opendevreview | Jake Hutchinson proposed openstack/kolla-ansible master: Ironic parameter rework and default NTP server https://review.opendev.org/c/openstack/kolla-ansible/+/893031 | 14:23 |
| opendevreview | Juan Pablo Suazo proposed openstack/kolla-ansible master: Deploy Glance and Cinder Backup with S3 backend support https://review.opendev.org/c/openstack/kolla-ansible/+/844614 | 14:54 |
| simondodsley | general question - how does kolla deal with unique iSCSI IQNs for each nova-compute or cinder-volume container, and how does this persist over container restarts? | 15:13 |
| simondodsley | I ask as we are seeing these sorts of issues with NMVe NQN values for the same containers, ie nova-compute an dcinder-volume | 15:13 |
| SvenKieske | what sort of issue are you referring to with "these"? I see no issue mentioned in your question. :) | 15:20 |
| kevko | frickler, bbezak, hrw ? >> https://review.opendev.org/c/openstack/kolla/+/895157/1 | 15:20 |
| SvenKieske | smondodsley: when I consult https://wiki.archlinux.org/title/Open-iSCSI#ISCSI_Qualified_Name_(IQN) (I don't use iSCSI myself) I think this stuff should be either persisted in /etc/ or in /var/lib/iscsi/ so should be no problem with regards to container reboots, at least from a cursory glance? | 15:23 |
| simondodsley | SvenKieske: we are seeing kolla images coming up with the same hostnqn file, so it has to be deleted and a new one generated for each container. The question then comes as to how we can persist the hostnqn file over container restarts as I would assume the restarted container would revert to the original hostnqn that comes from the image | 15:23 |
| simondodsley | the NVMe file is in the `/etc/nvme` directory. I guess the first part of the question is more important. Can the kolla images have the hostnqn and hostid files removed so all containers using the image don't end up with the same NQN value | 15:25 |
| simondodsley | These files are created as part of the `nvme` cli package being installed | 15:26 |
| simondodsley | works the same way as the iSCSI `inititatorname` file | 15:26 |
| simondodsley | check out https://bugs.launchpad.net/os-brick/+bug/2035606 | 15:28 |
| kevko | simondodsley: modify your kolla-ansible to map your hostqn file ? | 15:29 |
| kevko | into container ? | 15:29 |
| simondodsley | each nova-compute and cinder-volume require unique NQN values, so mapping a common file won't work | 15:30 |
| simondodsley | just like iscsi - which is why i was asking how iscsi is handled by kolla | 15:31 |
| SvenKieske | there's a dedicated iscsi role in k-a, but I don't see IQN mentioned there | 15:32 |
| kevko | simondodsley: not common file, jinja rendered template ..so you will have different id for different host ? | 15:33 |
| simondodsley | ok - i'll look at that option, but really whatever iSCSI does in kolla should be applied to nvme given it uses the same logic. | 15:34 |
| kevko | simondodsley: send a link what kolla does for iSCSI | 15:35 |
| simondodsley | kevko: i don't have access to that - i'm a vendor asking on behalf of a customer with this issue | 15:37 |
| SvenKieske | simondodsley: sure you have, this is open source software, anybody can take a look | 15:38 |
| simondodsley | true - i'll see if i can find it and get back to you | 15:38 |
| SvenKieske | here is the iscsi role, but I don't know if this really all applies to nvme's: https://opendev.org/openstack/kolla-ansible/src/branch/master/ansible/roles/iscsi | 15:39 |
| kevko | simondodsley: this one ? https://github.com/openstack/kolla/blob/master/docker/iscsid/extend_start.sh | 15:40 |
| SvenKieske | mhm, that might be something similar. | 15:40 |
| SvenKieske | that's the change: https://review.opendev.org/c/openstack/kolla/+/353483 | 15:42 |
| simondodsley | kevko: that process would be what needs to occur to generate a new NQN, but apparently the base image you currentry has already has an NQN created as part of the nvmecli installation. This means all the containers are getting the same NQN. The original image hostnqn, etc need to be deleted and a new one created similar to this | 15:43 |
| kevko | and this https://github.com/openstack/kolla/blob/64098b209cfab47b1cd23351c09e5de3ed275c46/docker/iscsid/Dockerfile.j2#L28-L31 | 15:43 |
| SvenKieske | simondodsley: how do you build your containers, or which containers do you consume, if you don't build them yourself? | 15:44 |
| SvenKieske | or your customer | 15:44 |
| kevko | ^^ simondodsley | 15:44 |
| simondodsley | i'll have to ask the customer | 15:44 |
| kevko | simondodsley: hmm, but this ID should persist between upgrade no ? | 15:45 |
| simondodsley | yes | 15:46 |
| simondodsley | otherwise connected volumes would get disconnected | 15:46 |
| simondodsley | as with the iscsi IQN | 15:46 |
| simondodsley | its all the same logic | 15:46 |
| kevko | simondodsley: hmm, so we have also bug in iscsi | 15:46 |
| kevko | in iscsi is this problem in image | 15:47 |
| kevko | root@ceph0:~# for i in $(seq 1 5); do iscsi-iname ; done | 15:47 |
| kevko | iqn.2016-04.com.open-iscsi:d9efe19afb7f | 15:47 |
| kevko | iqn.2016-04.com.open-iscsi:7f7a56d8602c | 15:47 |
| kevko | iqn.2016-04.com.open-iscsi:277a24fc6ff1 | 15:47 |
| kevko | iqn.2016-04.com.open-iscsi:e71943d8f83 | 15:47 |
| kevko | iqn.2016-04.com.open-iscsi:d8a98b38b18 | 15:47 |
| kevko | kolla removed ID and added generation to extend start | 15:48 |
| kevko | but if iscsci icontainer is upgraded ... there is new image with removed id and new extend start from new image will jump into process and again generate new ID | 15:48 |
| SvenKieske | that should imho be a proper container volume and the data needs to be persisted and injected into the container | 15:52 |
| SvenKieske | maybe with a rebuild command, to trigger new device discovery | 15:52 |
| SvenKieske | but who has time for that? :D worth writing a bugreport on launchpad.net imho | 15:52 |
| kevko | SvenKieske: aaa, it is ... https://github.com/openstack/kolla-ansible/blob/428acfe97fdfbc89fd4dea577160192647e6f6b5/ansible/roles/iscsi/defaults/main.yml#L48C1-L48C28 | 15:53 |
| kevko | so we are OK | 15:53 |
| kevko | nvme stuff should be fixed in the same way | 15:53 |
| SvenKieske | right, phew. | 15:53 |
| frickler | kevko: is there an example of the logging fix in action somewhere? sadly(?) the podman patch stopped failing | 15:58 |
| simondodsley | i think that is the correct approach for nvme. I guess it needs an nvme role | 15:58 |
| SvenKieske | simondosley: could you open a bugreport/feature request for this on launchpad? that helps keeping track of things and increases the chance of this getting implemented :) | 16:00 |
| simondodsley | I don't think you need to worry about ironic as I don't think they support nvme boot yet | 16:00 |
| simondodsley | SvenKieske: yes - I'll do that | 16:00 |
| SvenKieske | thank you :) | 16:00 |
| kevko | frickler: issue wasn't in podman patch but in Depends-On on requirements patch which results into incompatible glance installation ..so upgrade failed ..i just didn't know why ... when i implemented logging fix ... https://review.opendev.org/c/openstack/kolla/+/895157/1 i just found the issue and removed Depends-On | 16:04 |
| frickler | kevko: yes, I just would prefer to see an example of the error log before approving. but maybe we can go without it | 16:05 |
| kevko | frickler here it is https://da1d82983230150eebb3-89d69bea015aaa7a4b0cf0e0890d5ac2.ssl.cf2.rackcdn.com/825791/21/check/kolla-ansible-rocky9-upgrade/f683cab/primary/logs/build/000_FAILED_glance-base.log | 16:06 |
| kevko | ERROR:kolla.common.utils.glance-base:The conflict is caused by: | 16:07 |
| kevko | glance 27.0.0.0b3.dev51 depends on eventlet>=0.33.3 | 16:07 |
| kevko | The user requested (constraint) eventlet===0.33.1 | 16:07 |
| kevko | ibecause of depends-on the requirements repo for antelope was checked out | 16:07 |
| frickler | kevko: ah, cool, that's exactly what I wanted to see, thx | 16:10 |
| kevko | frickler: it would be nice to merge today because i have a fix for podman kolla-ansible aarch64 | 16:13 |
| kevko | frickler: i will fix other comments until evening | 16:14 |
| SvenKieske | nice | 16:14 |
| kevko | probably ... but i can't verify it until it will be merged | 16:15 |
| SvenKieske | maybe the time draws near, where podman is merged :) | 16:15 |
| kevko | that's the reason why i am doing it :D | 16:15 |
| kevko | and also to gain some reviewers time for letsencrypt :D | 16:16 |
| kevko | frickler: log fix can be merged after and i will add also podman error log fix | 16:17 |
| frickler | kevko: I'm confused now, I approved the log fix. I'm pretty sure we won't merge the podman build patch today | 16:19 |
| kevko | ah, ok | 16:19 |
| kevko | frickler: ok, nevermind, i will add log fix for podman after podman will be merged | 16:20 |
| kevko | frickler: podman is quite simple patch and i am not owner ..so i can add my +2 when i adress your comments | 16:20 |
| opendevreview | Jake Hutchinson proposed openstack/kayobe master: Add NTP parameter configuration https://review.opendev.org/c/openstack/kayobe/+/895199 | 16:32 |
| -opendevstatus- NOTICE: The lists.airshipit.org and lists.katacontainers.io sites will be offline briefly for migration to a new server | 16:47 | |
| opendevreview | Merged openstack/kolla master: Fix stderr logging when image build fail https://review.opendev.org/c/openstack/kolla/+/895157 | 17:39 |
| opendevreview | Michal Nasiadka proposed openstack/kolla stable/2023.1: Fix stderr logging when image build fail https://review.opendev.org/c/openstack/kolla/+/895030 | 17:57 |
| *** EugenMayer4404 is now known as EugenMayer440 | 18:13 | |
| opendevreview | Michal Arbet proposed openstack/kolla master: Add podman for image building https://review.opendev.org/c/openstack/kolla/+/825791 | 18:51 |
| opendevreview | Michal Arbet proposed openstack/kolla master: Add podman for image building https://review.opendev.org/c/openstack/kolla/+/825791 | 18:52 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!