| RAOF | Hm. Mesa doesn't seem to play nicely with debuginfod.ubuntu.com; I think the source map might not be quite right? | 07:24 |
|---|---|---|
| RAOF | gdb can't find any of the source files. | 07:24 |
| rbasak | RAOF: on which release? Source file indexing depends on a fairly recent patch AIUI. | 09:04 |
| rbasak | (and won't work on older releases) | 09:05 |
| RAOF | On mantic. Actually building the package, I *still* can't find the files it's referencing 🤷♀️ | 09:06 |
| RAOF | Like, gdb is looking for libegl.c; I can't find that file with that name in a built tree! | 09:09 |
| vinodwa | rbasak Hi, did you got your questions answered regarding LP#1996229 ? We need this reviewed and approved as soon as possible. Thanks | 13:23 |
| -ubottu:#ubuntu-devel- Launchpad bug 1996229 in python-magnumclient (Ubuntu Focal) "[SRU] magnum ui can not delete the coe cluster" [High, Triaged] https://launchpad.net/bugs/1996229 | 13:24 | |
| rbasak | freyes: to check we're not waiting on each other: you're working on answering my questions in bug 1996229, right? | 14:08 |
| -ubottu:#ubuntu-devel- Bug 1996229 in python-magnumclient (Ubuntu Focal) "[SRU] magnum ui can not delete the coe cluster" [High, Triaged] https://launchpad.net/bugs/1996229 | 14:08 | |
| freyes | rbasak, yes, I just posted the reply - https://bugs.launchpad.net/ubuntu/+source/magnum-ui/+bug/1996229/comments/5 | 14:08 |
| rbasak | ack, reading | 14:09 |
| rbasak | freyes: so another class of "Where problems could occur" would be users directly using python3-magnumclient with their own code, right? Has this been considered? | 14:13 |
| freyes | rbasak, yes, when I was going through the git log I didn't find code removed (e.g. deprecated API removed) | 14:14 |
| rbasak | Great, thanks | 14:15 |
| freyes | updated the bug description to include the impacted packages and this aspect of users running custom code with magnumclient as a library | 14:21 |
| rbasak | freyes: will the CLI for ListCluster (not sure what the command is exactly) now output more columns by default? | 14:24 |
| rbasak | I'm thinking about https://github.com/openstack/python-magnumclient/commit/74c5f22c6ffb23154f2e0412ba43371f3c02f3b7#diff-f055b476bf7193ad89f9512f8f9d13dc2fe6d32644ea5d8755294d45c8a23ea7 | 14:24 |
| -ubottu:#ubuntu-devel- Commit 74c5f22 in openstack/python-magnumclient "Support health_status on client side" | 14:24 | |
| freyes | rbasak, yes, it will | 14:26 |
| freyes | rbasak, this has been identified as a problem, not showing that column, since operators have no visibility on the cluster's state | 14:27 |
| freyes | rbasak, the general guideline has been to use the openstackclients snap | 14:27 |
| freyes | which ships newer versions of the clients (from upstream python tarballs) | 14:28 |
| freyes | although this change may break bash scripts :-/ I didn't think about it | 14:29 |
| rbasak | freyes: how configurable is the list CLI currently? Would it be feasible/trivial to make the default not output that column, for example, but make it possible with an option? Or would that be not useful anyway? | 14:34 |
| rbasak | If you want to deliberately change the output, then that wouldn't be unreasonable. We just need to be careful to think about the possible consequences. | 14:35 |
| rbasak | freyes: I'm also curious why you didn't only cherry-pick the necessary patches - looks like that would be feasible? | 14:35 |
| adrien | /lastlog -clear | 14:38 |
| adrien | woops :) | 14:38 |
| sergiodj | jbicha: hey, I'm about to start my patch pilot shift and I couldn't find your entry in the handoff doc. is there anything worth mentioning from yesterday's shift? | 14:40 |
| adrien | ginggs: you retried rust-globset on yesterday but I'm wondering if the triggers were right: more specifically I'm wondering if the needed ones are exactly rust-bstr/1.5.0-1 and rust-globset/0.4.10-1 , which would be that URI: https://autopkgtest.ubuntu.com/request.cgi?release=mantic&arch=armhf&package=rust-globset&trigger=rust-bstr%2F1.5.0-1&trigger=rust-globset%2F0.4.10-1 | 14:40 |
| ginggs | adrien: i think that combination was tried, but failed with 'websocket: bad handshake' -- let's try again... | 14:44 |
| freyes | rbasak, on the aspect of bash scripts, typically users run the openstack clients with flags (e.g. `openstack coe cluster list --format json | jq ...`), it's a very common practice by the complexity of the objects in OpenStack. Even users who don't want to deal with jq they do things like `openstack coe cluster list --format value -c ID` to filter the specific data they want to consume | 14:46 |
| freyes | rbasak, why not cherry-pick, this is because we went initially that route, and we kept finding incompatibility issues, so we were taking a path where the client would become more unstable than just bumping up the version with what was really tested upstream | 14:47 |
| adrien | ginggs: yup, that's why I think it's worth a new try; tbh I'm not certain that will be the solution but it seems plausible enough | 14:49 |
| freyes | rbasak, about patching the CLI to not display the health status, removing it from the list of default columns is a oneline change, I would need to check how large a change would be if to allow users to opt-in to the show the column, there is a "-c" flag that allows users to make the CLI print specific columns | 14:52 |
| === esembee_ is now known as esembee | ||
| sergiodj | @pilot in | 14:59 |
| === ChanServ changed the topic of #ubuntu-devel to: Archive: Mantic Open | Devel of Ubuntu (not support) | Build failures: http://qa.ubuntuwire.com/ftbfs/ | #ubuntu for support and discussion of Bionic-Lunar | Patch Pilots: sergiodj | ||
| adrien | I have a question about rust packages: for instance rust-jsonwebtoken; for 8.2.0, it says " crate directory not found: /usr/share/cargo/registry/jsonwebtoken-8.2.0" and that path comes from librust-jsonwebtoken-dev which is pulled at version 8.3.0-1 | 15:28 |
| adrien | I think the solution is to re-trigger with rust-jsonwebtoken too; is that analysis correct? | 15:29 |
| schopin | Probably, yes. | 15:29 |
| adrien | there's a followup comment but I need to get some stats first | 15:31 |
| adrien | I only get three packages affected by this at the moment but is there something to improve to prevent these from occuring again? | 15:33 |
| adrien | I think I can get the triggers for the affected packages fairly easily but only if I manage to do it before the children hord of children arrives | 15:34 |
| adrien | first one | 15:35 |
| adrien | % ./retry-autopkgtest-regressions --blocked-by-tests rust-jsonwebtoken | 15:35 |
| adrien | https://autopkgtest.ubuntu.com/request.cgi?release=mantic&arch=armhf&package=rust-jsonwebtoken&trigger=rust-pem%2F1.0.2-2 | 15:35 |
| adrien | rust-syn seems to be a slightly different issue: it searches for the 2.0.16 directory while testing 2.0.18 | 15:39 |
| adrien | the trigger is 2.0.18 and the -dev package is at version 2.0.18 | 15:40 |
| adrien | the only mention of 2.0.16 in https://autopkgtest.ubuntu.com/results/autopkgtest-mantic/mantic/amd64/r/rust-syn/20230621_233537_be14d@/log.gz is basically the failure itself | 15:41 |
| adrien | the only place I find 2.0.16 is in debian experimental and I'm not even sure that version was synced | 15:44 |
| adrien | I'll leave for now but I'm very interested in any idea people could have about this | 15:44 |
| ginggs | adrien: package=rust-jsonwebtoken&trigger=rust-pem%2F1.0.2-2 retried, but i think that combination has been retried many times | 15:57 |
| * ginggs tries package=rust-jsonwebtoken&trigger=rust-pem%2F1.0.2-2&trigger=rust-jsonwebtoken/8.3.0-1 | 16:00 | |
| ginggs | adrien: re rust-syn -- yes you are correct, the only mention of 2.0.16 is in the test itself | 16:18 |
| ginggs | https://salsa.debian.org/rust-team/debcargo-conf/-/blob/master/src/syn/debian/tests/control | 16:18 |
| ginggs | looks like that is the problem! | 16:18 |
| adrien | this looks like a good way to have version issues; I think the author was focused on a transition and specifically 2.0.16 | 16:21 |
| adrien | ginggs: oh, yes, I definitely wanted to have jsonwebtoken itself as part of the triggers; I think I completely misread the output of r-a-r and that's probably not the first time today; I think I was reading the package as part of the triggers | 16:22 |
| adrien | ah, right, almost 6 hours ago I pasted one URI which I had similarly misread (repeteadly) | 16:23 |
| rbasak | freyes: thanks. I think this is mostly on me now to finish my review, and then I might have some further discussion. I've been distracted by meetings and have a prior commitment this evening. I'm hoping to come back later though. | 16:52 |
| freyes | thanks, rbasak | 16:56 |
| jbicha | sergiodj: oh sorry I missed the doc. I didn't have anything to pass along | 17:17 |
| sergiodj | jbicha: ACK, np. thanks | 17:17 |
| ogayot | Hi, may I ask somebody to trigger this autopkgtest for me please? https://autopkgtest.ubuntu.com/request.cgi?release=mantic&arch=amd64&package=pyparted&ppa=ogayot%2Fmantic-proposed&trigger=pyparted%2F3.12.0-4ubuntu1~ppa2 Thanks! | 18:01 |
| sergiodj | ogayot: done | 18:03 |
| ogayot | sergiodj: thanks! | 18:03 |
| sergiodj | np | 18:03 |
| adrien | create an issue for rust-syn: https://salsa.debian.org/rust-team/debcargo-conf/-/issues/46 (I'll link it to LP and upd*ate-excuse on tomorrow) | 18:44 |
| -ubottu:#ubuntu-devel- Issue 46 in rust-team/debcargo-conf "syn: d/tests/control not synchronized with d/changelog" [Opened] | 18:44 | |
| sergiodj | RAOF: hey, https://bugs.launchpad.net/ubuntu/+source/thermald/+bug/1995606 has been updated and is in the sponsorship queue but I don't have enough time/knowledge to review it properly. it seems like you're driving the review whenever you have time, so I'll leave it to you | 18:48 |
| -ubottu:#ubuntu-devel- Launchpad bug 1995606 in thermald (Ubuntu Jammy) "Upgrade thermald to 2.5.1 for Jammy (22.04)" [Undecided, Incomplete] | 18:48 | |
| sergiodj | @pilot out | 19:09 |
| === ChanServ changed the topic of #ubuntu-devel to: Archive: Mantic Open | Devel of Ubuntu (not support) | Build failures: http://qa.ubuntuwire.com/ftbfs/ | #ubuntu for support and discussion of Bionic-Lunar | Patch Pilots: N/A | ||
| adrien | for rust-rustls-pemfile, I would like to run the rust-rustls tests on all arches again *but* with the additional rust-base64 trigger version 0.21.2-1 and I see no way to do that with retry-autopkgtest-regressions | 20:13 |
| adrien | for amd64 that would be https://autopkgtest.ubuntu.com/request.cgi?release=mantic&arch=amd64&package=rust-rustls&trigger=rust-rustls-pemfile%2F1.0.2-1&trigger=rust-base64%2F0.21.2-1 (I guess dropping arch=amd64 runs it on all arches?) | 20:15 |
| bdmurray | adrien: the documentation for retry-autopkgtest-regressions talks about pipeing the output to vipe and you could add a trigger in there | 20:33 |
| adrien | bdmurray: ok, thanks; I remember vorlon was asking for invocations of r-a-r rather than links to click on but I guess an invocation and a comment about what to change would work too | 20:57 |
| mwhudson | @pilot in | 22:21 |
| === ChanServ changed the topic of #ubuntu-devel to: Archive: Mantic Open | Devel of Ubuntu (not support) | Build failures: http://qa.ubuntuwire.com/ftbfs/ | #ubuntu for support and discussion of Bionic-Lunar | Patch Pilots: mwhudson | ||
| vorlon | bdmurray, adrien: I've been known to pipe to sed :) | 22:26 |
| bdmurray | that's what he said | 22:27 |
| vorlon | Right Sed Fred | 22:28 |
| bdmurray | lol | 22:28 |
| sarnold | lol | 23:41 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!