| TheJulia | continuity: how does it hang? | 00:07 |
|---|---|---|
| TheJulia | I'd use dib, but I don't know. arne_wiebalck are y'all using rocky at all? | 00:08 |
| TheJulia | stepping away shortly for the evening | 00:08 |
| TheJulia | and arne is the EU, so expect delays | 00:08 |
| TheJulia | :) | 00:08 |
| continuity | yeah im also in the EU, and should be asleep :D | 00:13 |
| continuity | it gets part way through the boot then throws "dracut initqueue timeout" | 00:14 |
| continuity | I think to be honest its time to get some sleep and come at it frsh | 00:14 |
| opendevreview | Merged openstack/bifrost master: Simplify set_fact for ssh_public_key https://review.opendev.org/c/openstack/bifrost/+/874857 | 00:44 |
| opendevreview | Iury Gregory Melo Ferreira proposed openstack/ironic master: Add DB API for Firmware https://review.opendev.org/c/openstack/ironic/+/883062 | 06:24 |
| rpittau | good morning ironic! o/ | 06:54 |
| rpittau | FYI after some investigation we've found the root cause of metal3-integration job failures and a fix has been proposed https://github.com/metal3-io/metal3-dev-env/pull/1238 | 07:58 |
| rpittau | we should recheck and merge https://review.opendev.org/c/openstack/ironic/+/882073 as soon as the fix merges | 07:59 |
| dtantsur | iurygregory, TheJulia, our habit of calling flush() is questionable enough, I don't think calling it several times per transaction is a reasonable thing to do. | 09:47 |
| dtantsur | Until the transaction is over, there are no guarantees that a duplicate error won't be raised. | 09:49 |
| iurygregory | good morning Ironic, happy friday | 12:13 |
| * iurygregory trying to be awake to answer things | 12:13 | |
| iurygregory | oh nice I got -1 because of the docs :D (normally I don't run tox -e docs if I'm not doing a change with releasenote or rst file .-.) | 12:25 |
| iurygregory | dtantsur, even if we are doing one at a time? | 12:30 |
| dtantsur | iurygregory: what are you trying to achieve? | 12:47 |
| dtantsur | I'm not sure flush even does what you think it should do. And it definitely leaves a space for races until COMMIT is issued. | 12:48 |
| iurygregory | I've changed the approach a little bit, instead of trying to create the list of firmware components | 12:49 |
| iurygregory | I will create one at a time | 12:49 |
| dtantsur | iurygregory: be careful to not fall in the same trap as Mahnoor just discovered happened with inventory API, when we have a create() call while we need a create_or_update() | 12:50 |
| dtantsur | i.e. think what you actually want to do with duplicates | 12:51 |
| rpittau | fix for metal3-integration worked, I think we can merge https://review.opendev.org/c/openstack/ironic/+/882073 | 12:51 |
| iurygregory | dtantsur, do you have a link? | 12:54 |
| dtantsur | iurygregory: not yet | 12:54 |
| dtantsur | you need to decide for your case though | 12:55 |
| iurygregory | so it's more of a question of design? | 12:55 |
| dtantsur | it's a question of which pattern you actually need to use | 12:56 |
| dtantsur | so, you see, when a user creates a port, it has to be unique | 12:56 |
| iurygregory | yup | 12:56 |
| dtantsur | when inspection creates an inventory, it has to either be unique or replace the old one | 12:56 |
| dtantsur | a semantic difference here | 12:56 |
| iurygregory | I would say it's more like port creation | 12:58 |
| iurygregory | we have unique constraint for the node_id + component | 13:00 |
| dtantsur | it's an orthogonal question | 13:01 |
| dtantsur | it's unique, sure, but what happens if a record already exists? | 13:01 |
| iurygregory | if the request is to create it would fail | 13:02 |
| dtantsur | and what happens with the failure? | 13:02 |
| dtantsur | I'm leading to the thought: it's possible you need a create-or-update semantics since there is no case of "a user is creating a duplicate" for firmware upgrades (or not inventory) | 13:03 |
| iurygregory | the failure would be raised to the caller | 13:03 |
| iurygregory | and it would depend how we would handle I would say | 13:04 |
| dtantsur | the caller is you :) | 13:04 |
| dtantsur | you can, of course, do try.. except, but you can just as well to DELETE+INSERT on the database layer | 13:05 |
| iurygregory | not sure if would be a good idea ... | 13:05 |
| iurygregory | because like, we have initial_version it would be the first one we saw | 13:05 |
| iurygregory | other fields would be updated | 13:06 |
| iurygregory | but if we decide that we can just re-create we could tell that the initial version was "v2.0.0" instead of the "v1.0.0" that was before... | 13:06 |
| dtantsur | yeah, possibly. i'm just making sure you have this in mind | 13:06 |
| dtantsur | (for node inventory there is no point to have a separate create, but we do..) | 13:07 |
| opendevreview | Iury Gregory Melo Ferreira proposed openstack/ironic master: Add DB API for Firmware https://review.opendev.org/c/openstack/ironic/+/883062 | 13:33 |
| iurygregory | finally fixed the docs, I need more coffee | 13:33 |
| iurygregory | dtantsur, we can probably discuss with the community if you think it would be interesting what we could do about the habit of calling flush in every method also (but monday is holiday in EU for most countries) | 13:35 |
| TheJulia | dtantsur: I talked to iury last night and I think he was going to end up dropping it in favor of retooling the approach since update will be generally more common | 13:38 |
| * TheJulia tries to wake up | 13:38 | |
| opendevreview | Mohammed Boukhalfa proposed openstack/sushy-tools master: Add fake_ipa inspection, lookup and heartbeater to fake system https://review.opendev.org/c/openstack/sushy-tools/+/884502 | 13:40 |
| * iurygregory wondering if the world is on fire .. | 13:41 | |
| rpittau | iurygregory: not yet | 13:42 |
| iurygregory | rpittau, hahah XD | 13:42 |
| iurygregory | is just because TheJulia showed up :D | 13:43 |
| TheJulia | eh, I didn't sleep well and I was at least going to check | 13:47 |
| iurygregory | ouch =( sorry to hear that | 13:47 |
| TheJulia | The hockey game last night was awesome | 13:47 |
| iurygregory | I had 6hrs of sleep only, trying to update the patch XD | 13:48 |
| iurygregory | I never saw a hockey game (not even on TV) | 13:48 |
| TheJulia | heh | 13:54 |
| TheJulia | I blame JayF for introducing us to hockey | 13:54 |
| iurygregory | ++ | 13:59 |
| TheJulia | https://www.youtube.com/watch?v=1MQFZ8_UKD4 <-- game recap video | 14:01 |
| iurygregory | I like Firebirds in the name of the team | 14:16 |
| opendevreview | Julia Kreger proposed openstack/ironic stable/xena: Fix Cinder Integration fallout from CVE-2023-2088 https://review.opendev.org/c/openstack/ironic/+/883580 | 14:32 |
| rpittau | bye everyone, see you on tuesday, have a great weekend! o/ | 14:38 |
| TheJulia | I suspect the idea came up because it is so hot here you can burst in to flames during the summer time | 14:38 |
| TheJulia | (not really, but feels like it | 14:38 |
| iurygregory | oh wow | 14:42 |
| iurygregory | bye rpittau, enjoy the long weekend! | 14:42 |
| opendevreview | Mohammed Boukhalfa proposed openstack/sushy-tools master: Add fake_ipa inspection, lookup and heartbeater to fake system https://review.opendev.org/c/openstack/sushy-tools/+/875366 | 14:47 |
| opendevreview | Merged openstack/ironic master: Make metal3 job voting https://review.opendev.org/c/openstack/ironic/+/882073 | 14:57 |
| clarkb | re ^ what did you do for handling the github api limits? | 14:58 |
| iurygregory | clarkb, tbh I'm not sure if hit this issue yet | 15:01 |
| iurygregory | but the patches rpittau worked on was: https://review.opendev.org/c/openstack/project-config/+/878223 https://review.opendev.org/c/openstack/ironic/+/877600 https://review.opendev.org/c/openstack/ironic/+/863873 | 15:02 |
| * TheJulia goes and gets breakfast | 15:11 | |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: Migrate the inspector's /continue API https://review.opendev.org/c/openstack/ironic/+/875944 | 15:19 |
| dtantsur | I think this is ready ^^ | 15:19 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: [DNM] Test new inspection code path https://review.opendev.org/c/openstack/ironic/+/881492 | 15:19 |
| dtantsur | have a nice weekend o/ | 15:20 |
| iurygregory | enjoy the long weekend dtantsur o/ | 15:21 |
| opendevreview | Julia Kreger proposed openstack/ironic stable/wallaby: Fix Cinder Integration fallout from CVE-2023-2088 https://review.opendev.org/c/openstack/ironic/+/883581 | 17:14 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!