| opendevreview | Takashi Kajinami proposed openstack/ironic master: Make account prefix of Swift confgurable https://review.opendev.org/c/openstack/ironic/+/823913 | 00:42 |
|---|---|---|
| opendevreview | Takashi Kajinami proposed openstack/ironic master: Make account prefix of Swift confgurable https://review.opendev.org/c/openstack/ironic/+/823913 | 00:42 |
| opendevreview | Vanou Ishii proposed openstack/ironic master: Deal with iRMC virtual media incompatibility https://review.opendev.org/c/openstack/ironic/+/823790 | 07:05 |
| arne_wiebalck | Good morning janders and Ironic! | 07:22 |
| opendevreview | Arne Wiebalck proposed openstack/ironic-python-agent master: SoftwareRAID: Use efibootmgr (and drop grub2-install) https://review.opendev.org/c/openstack/ironic-python-agent/+/826026 | 07:54 |
| arne_wiebalck | dtantsur: ^^ nits addressed (except for the refactor which I suggest split into a followup) | 07:54 |
| arne_wiebalck | dtantsur: I also changed the reno to 'fixes', this should allow for more easy beackporting ... whether we do that depends on the whether we backport mgoddard's change (I think we should) | 07:56 |
| rpittau | good morning ironic! o/ | 08:14 |
| janders | good morning arne_wiebalck rpittau and Ironic o/ | 08:26 |
| rpittau | hey janders :) | 08:27 |
| arne_wiebalck | hey rpittau o/ | 08:31 |
| rpittau | hey arne_wiebalck :) | 08:37 |
| dtantsur | morning | 09:18 |
| dtantsur | arne_wiebalck: thanks! | 09:18 |
| rpittau | good morning dtantsur :) | 09:21 |
| janders | good morning dtantsur | 09:28 |
| opendevreview | Takashi Kajinami proposed openstack/ironic master: Make account prefix of Swift confgurable https://review.opendev.org/c/openstack/ironic/+/823913 | 10:28 |
| tkajinam | dtantsur, ^^^ sorry for updating the patch after your +2. I forgot to update the release note which is done in the latest patch | 10:32 |
| opendevreview | Arne Wiebalck proposed openstack/ironic-python-agent master: SoftwareRAID: Use efibootmgr (and drop grub2-install) https://review.opendev.org/c/openstack/ironic-python-agent/+/826026 | 10:48 |
| arne_wiebalck | dtantsur: ^^ zip objects have no len :-S | 10:49 |
| * arne_wiebalck deserves this since he pushed untested code :-D | 10:50 | |
| arne_wiebalck | dtantsur: zip objects have no len in py3, that is | 10:50 |
| iurygregory | good morning Ironic o/ | 10:51 |
| opendevreview | Riccardo Pittau proposed openstack/bifrost master: Enable extra repositories for CentOS Stream 9 https://review.opendev.org/c/openstack/bifrost/+/826442 | 10:54 |
| rpittau | hey iurygregory :) | 10:54 |
| iurygregory | rpittau, o/ | 10:55 |
| dtantsur | arne_wiebalck: mmm, right | 11:58 |
| opendevreview | Takashi Kajinami proposed openstack/ironic-lib master: Allow customizing roles to use JSON RPC https://review.opendev.org/c/openstack/ironic-lib/+/826456 | 12:28 |
| opendevreview | Takashi Kajinami proposed openstack/ironic-lib master: Allow customizing roles to use JSON RPC https://review.opendev.org/c/openstack/ironic-lib/+/826456 | 12:30 |
| janders | See you tomorrow Ironic o/ | 12:48 |
| opendevreview | Dmitry Tantsur proposed openstack/bifrost master: Use Type=notify in systemd units for services https://review.opendev.org/c/openstack/bifrost/+/826465 | 13:25 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: Return non-zero exit code on failures https://review.opendev.org/c/openstack/ironic/+/826467 | 13:40 |
| opendevreview | Arne Wiebalck proposed openstack/ironic-python-agent master: SoftwareRAID: Use efibootmgr (and drop grub2-install) https://review.opendev.org/c/openstack/ironic-python-agent/+/826026 | 13:46 |
| yoctozepto | dtantsur: simple q on ~> https://review.opendev.org/c/openstack/bifrost/+/825053 | 13:53 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: [WIP] Wait for conductor start before notifying systemd https://review.opendev.org/c/openstack/ironic/+/826470 | 13:59 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: [WIP] Wait for conductor start before notifying systemd https://review.opendev.org/c/openstack/ironic/+/826470 | 14:06 |
| dtantsur | yoctozepto: answered | 14:15 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: Wait for conductor start before notifying systemd https://review.opendev.org/c/openstack/ironic/+/826470 | 14:19 |
| opendevreview | Dmitry Tantsur proposed openstack/bifrost master: Use Type=notify in systemd units for services https://review.opendev.org/c/openstack/bifrost/+/826465 | 14:23 |
| *** dansmith is now known as Guest661 | 14:38 | |
| yoctozepto | dtantsur: thx | 14:59 |
| opendevreview | Radosław Piliszek proposed openstack/bifrost bugfix/10.0: [CI] Drop all config https://review.opendev.org/c/openstack/bifrost/+/825053 | 15:18 |
| opendevreview | Radosław Piliszek proposed openstack/bifrost bugfix/10.1: [CI] Drop all config https://review.opendev.org/c/openstack/bifrost/+/825054 | 15:20 |
| arne_wiebalck | rloo: TheJulia: dtantsur: wondering if you had any thoughts on https://storyboard.openstack.org/#!/story/2009805 | 16:01 |
| *** dansmith_ is now known as dansmith | 16:09 | |
| TheJulia | arne_wiebalck: I think that makes a lot of sense, or at least, pend/batch, or something | 16:38 |
| arne_wiebalck | TheJulia: it may also help with operators' "what if" questions at 3am :) | 16:40 |
| rpittau | bye everyone, see you tomorrow! o/ | 16:49 |
| TheJulia | Yeah | 16:53 |
| rloo | arne_wiebalck: we use ironic with nova, so cleaning (unless someone manually wants to clean a bunch of nodes) is triggered when nova-server-delete. one can't just 'clean' ironic nodes that are 'active'. wondering what the use case is (I mean, I'm fine with the rfe, but I don't think we have a use for it). | 17:06 |
| arne_wiebalck | rloo: thanks, that is a good point. we use ironic with nova as well, but we have seen cases where nova deletes its compute nodes (not the instances, though) for unknown reasons. this did not have any negative effect, apart from the need to manually restore the allocations. I am still worried that if nova "decides" to delete instances, it could, and ironic would trust it. | 17:10 |
| arne_wiebalck | rloo: and if someone or something gets access to the API to launch 'undeploy' in a loop, nothing would stop this (by the time we would get an alarm, nodes are already booting into the IPA image to clean); an additional safeguard _in the code_ to limit the concurrent cleaning could provide additional protection | 17:23 |
| arne_wiebalck | bye everyone, see you tomorrow o/ | 17:30 |
| TheJulia | arne_wiebalck: I know why compute nodes get deleted | 17:30 |
| TheJulia | I've actually started on a fix but have been going back and forth with nova folks | 17:30 |
| arne_wiebalck | TheJulia: oh, yeah? did I already ask for a link? | 17:30 |
| TheJulia | arne_wiebalck: basically https://review.opendev.org/c/openstack/nova/+/813897 | 17:31 |
| rloo | oh geez, rogue nova... should definitely fix that. | 17:32 |
| rloo | wondering if 'undeploy' is what needs to be guarded... | 17:33 |
| arne_wiebalck | TheJulia: would that make a nova-compute delete all its compute nodes? | 17:34 |
| arne_wiebalck | TheJulia: b/c this is what we have seen in 4 or 5 of our conductor groups | 17:34 |
| TheJulia | arne_wiebalck: the opposite, head off the race that would cause the deletion of the record by reconciling the DB upon rebalance | 17:34 |
| TheJulia | there is a second issue in that where nodes just disappear from the list during state transitions | 17:34 |
| TheJulia | stevebaker[m] had a WIP up for that | 17:35 |
| arne_wiebalck | TheJulia: what I meant was: does the problem your patch is addressing fit with the symptoms we see (n-c delete all nodes)? | 17:38 |
| arne_wiebalck | TheJulia: maybe there is yet another issue :-S | 17:39 |
| TheJulia | deletion of compute nodes only | 17:39 |
| TheJulia | the actual instances should not be deleted | 17:39 |
| arne_wiebalck | right, this is what we see | 17:39 |
| TheJulia | *but* it does reassociate them on rebalance | 17:39 |
| TheJulia | since they get orphaned | 17:39 |
| arne_wiebalck | but we see this for individual nodes sometimes, but we had an incident end of last year where *all* nodes were deleted in one go | 17:40 |
| arne_wiebalck | we were not able to understand what happened | 17:41 |
| arne_wiebalck | main (and only suspect) is a half-configured Ironic API node which I built for the upgrade and which may came online at some point | 17:42 |
| arne_wiebalck | anyway :) | 17:42 |
| arne_wiebalck | bye everyone o/ | 17:44 |
| *** sshnaidm is now known as sshnaidm|afk | 17:49 | |
| TheJulia | hmmm | 18:13 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Add additional ramdisk tests https://review.opendev.org/c/openstack/ironic/+/826506 | 18:17 |
| opendevreview | Julia Kreger proposed openstack/ironic stable/xena: Add additional ramdisk tests https://review.opendev.org/c/openstack/ironic/+/826507 | 18:21 |
| opendevreview | Julia Kreger proposed openstack/ironic stable/wallaby: Add additional ramdisk tests https://review.opendev.org/c/openstack/ironic/+/826508 | 18:23 |
| *** dansmith is now known as Guest688 | 19:01 | |
| stevebaker[m] | good morning | 19:28 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Add additional ramdisk tests https://review.opendev.org/c/openstack/ironic/+/826506 | 19:36 |
| *** Guest688 is now known as dansmith | 19:40 | |
| *** dansmith is now known as dansmith_ | 19:41 | |
| *** dansmith_ is now known as dansmith | 19:41 | |
| opendevreview | Julia Kreger proposed openstack/ironic stable/xena: Add additional ramdisk tests https://review.opendev.org/c/openstack/ironic/+/826507 | 19:46 |
| opendevreview | Julia Kreger proposed openstack/ironic stable/wallaby: Add additional ramdisk tests https://review.opendev.org/c/openstack/ironic/+/826508 | 19:47 |
| TheJulia | good morning stevebaker[m] | 20:00 |
| TheJulia | stevebaker[m]: do you happen to have some spare spoons? | 20:45 |
| stevebaker[m] | TheJulia: I do currently, yes | 20:46 |
| TheJulia | stevebaker[m]: I'd kind of <3 another set of eyes on https://bugzilla.redhat.com/show_bug.cgi?id=1980051 Specifically I posted https://review.opendev.org/c/openstack/ironic/+/826506 while digging into it, and it *looks* like it should work as expected, but it doesn't seem to and the parameter is empty. | 20:48 |
| TheJulia | what is weird, is they can pull the node info, and the field is populated as we would expect | 20:51 |
| TheJulia | just never makes it into the config file | 20:51 |
| stevebaker[m] | TheJulia: is the config file generated by a template? could something in the tooling be setting a different template? | 21:01 |
| TheJulia | it is a template, but that is a great question | 21:01 |
| stevebaker[m] | (something==puppet) | 21:05 |
| TheJulia | I believe it is default ipxe | 21:05 |
| hjensas | puppet-ironic, can override the ipxe template. But does not by default. I don't see anything in tripleo-heat-templates or tripleo-ansible that is overriding it. | 22:32 |
| TheJulia | Yeah, that is what I saw when I looked yesterday | 22:50 |
| janders | good morning Ironic o/ | 23:52 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!