| guiverc | bdmurray, lubuntu focal daily (2021-08-16) boots fine on sony_vaio_svp112a1cw; ubuntu zsync keeps falling over but i'll get there.. and I'll add entry on bug report in due course | 00:25 |
|---|---|---|
| mwhudson | vorlon: omg is it possibly shim-signed works on arm64 on all releases now? or do we still need it to migrate on bionic? | 01:15 |
| mardy | is there a way for a privileged process (such as snapd) to find the ID of all wayland windows belonging to a certain PID? | 08:02 |
| schopin | Regarding the freeze, what are the rules regarding the packages that are in -proposed at freeze time? Do they need a freeze exception ? | 09:34 |
| doko_ | no | 09:42 |
| === doko_ is now known as doko | ||
| slyon | schopin: This would best be asked in #ubuntu-release but from my understanding packages in -proposed are fine and can be migrated after FF (do not need a FFe), they just won't accept anything new into -proposed. | 10:08 |
| RikMills | mwhudson: qtwebengine fix looks good. it has been uploaded to unstable, and will get synced later | 12:30 |
| RikMills | sync will close the bug | 12:30 |
| rafaeldtinoco | sorry.. maybe this was discussed already here but what happened to ddebs.ubuntu.com/dists ? | 13:08 |
| rafaeldtinoco | asking this because all linux-image-XXX-dbgsym files are 15Kb | 13:13 |
| rafaeldtinoco | and dists directory was renamed to dists.old | 13:14 |
| xnox | ddstreet: laney: forwards the token to juliank..... | 13:15 |
| ddstreet | thanks, laney that's all you needed right? | 13:20 |
| juliank | We do not have docs for which token is which or how to udpate them | 13:21 |
| mclemenceau_ | o/ Hello, I'm looking for a sponsor for a simple merge for lshw (LP: #1939695) before feature freeze? anyone interested ? Thx :) | 13:22 |
| ubottu | Launchpad bug 1939695 in lshw (Ubuntu) "Please merge lshw-02.18.85-0.7 (main) from Debian unstable (main)" [Wishlist, New] https://launchpad.net/bugs/1939695 | 13:22 |
| juliank | nvm, I was on the wrong page | 13:23 |
| juliank | ERROR cannot deploy bundle: processing option "swift-web-credentials" for application "autopkgtest-web": resolving include "/srv/mojo/mojo-prod-proposed-migration/focal/production/local/swift-web-credentials.conf": include file "/srv/mojo/mojo-prod-proposed-migration/focal/production | 13:23 |
| juliank | /local/swift-web-credentials.conf" not found | 13:23 |
| juliank | horry | 13:23 |
| ddstreet | hmm, well i've never had access to the internals of autopkgtest-cloud deployment, i'm unlikely to be able to help with that | 13:24 |
| juliank | sil2100: ^ you added include-file for swift-web-credentials.conf to autopkgtest-cloud, but no such secret is provided on prod-proposed-migration | 13:27 |
| juliank | Also how did Laney deploy this if it's not deployable | 13:28 |
| sil2100 | juliank: we never deployed it in production so I guess! I know it has to be available on the staging instance, but I have no knowledge of how that ties to the creds for production | 13:34 |
| juliank | sil2100: in any case, it's broken right now, so no github systemd status thingy update | 13:39 |
| juliank | sil2100: I'll try to reconstruct it and hope everything works :D | 13:41 |
| slyon | mclemenceau_: o/ I can do that sponsoring. | 13:44 |
| sil2100 | Ah! So I see my private-testing branch got merged! Ok, so uh, I can try reading up on how the mojo creds are being stored, possibly moving those over from the staging instance could work | 13:44 |
| juliank | sil2100: Copying over doens't work, no, they're separate users | 13:45 |
| sil2100 | juliank: ah, you mean the swift user? Yeah, I wonder if we have a user for that for production, hmmm | 13:52 |
| juliank | I don't know what you did and how it related, certainly the password is different than what is stored in swift_password | 13:53 |
| sil2100 | Ok, so I see Laney only requested a new swift user for staging, let me request a new user from IS for production | 13:55 |
| mclemenceau_ | o/ slyon sure that would be great thank you! | 14:01 |
| === genii-core is now known as genii | ||
| Guest11 | I created a simple .deb package as per https://ubuntuforums.org/showthread.php?t=910717 .. I need to print some text as a banner while installing or removing the newly created .deb package, does DEBIAN/control supports this? | 14:09 |
| Guest11 | Is it possible to run a set of commands after post-installation or removal of the .deb package? | 14:10 |
| rbasak | Guest11: try #ubuntu-packaging. This channel is for the development of Ubuntu itself. | 14:15 |
| laney | sil2100: ohhhh I forgot that | 14:28 |
| laney | I thought I saw that mojo succeeded, what did I do | 14:28 |
| laney | nothing good, that's for sure | 14:29 |
| laney | sil2100: also this is the kind of thing that would be documented ;-) | 14:29 |
| sil2100 | laney: hah, no worries, I also forgot that we had the user only for staging, not production as well! Somehow in my mind I thought it's one for both | 14:30 |
| laney | juliank: just revert the charm to the previous version in the stable channel thing | 14:30 |
| laney | juliank: did you revise your opinion of not having docs? | 14:31 |
| juliank | docs are there | 14:31 |
| juliank | just in manage/admin thingy, not deploy, but that makes sense I guess | 14:31 |
| laney | nod | 14:32 |
| laney | interesting that xnox sent it to you and not me who replied to the thread and was CCed | 14:32 |
| laney | recipe for confusion there | 14:32 |
| sil2100 | laney: filled in RT: #132557, can you check if what I copy-pasted and edited there makes sense? | 14:35 |
| laney | sil2100: looks good, get it prioritised | 14:36 |
| cjwatson | rafaeldtinoco: Both dists and (to be ignored) dists.old are present. linux-image-*-dbgsym are probably not very interesting due to the signed/unsigned split; it looks like you typically want linux-image-unsigned-*-dbgsym | 14:37 |
| cjwatson | rafaeldtinoco: If that doesn't solve your problem, please be more specific | 14:37 |
| juliank | laney: I've blacked out on the whole charm revert thingy | 14:38 |
| rafaeldtinoco | cjwatson: it does indeed and thank you cjwatson. I thought signed and unsigned had diff debug packages. thx | 14:48 |
| laney | juliank: charm release takes a revision on the end | 14:52 |
| laney | juliank: the latest version is 109 so charm release cs:~ubuntu-release/autopkgtest-cloud-worker-108 | 14:53 |
| laney | might be an idea to document that one | 14:54 |
| juliank | laney: gotta look at -web | 15:00 |
| juliank | laney: tags for 55,56,57 are missing | 15:01 |
| laney | I just pushed 57 | 15:01 |
| laney | am I responsible for those others? :p | 15:02 |
| juliank | laney: I don't know | 15:02 |
| juliank | laney: neither revert helps anyway | 15:02 |
| laney | why, hook errors? | 15:03 |
| laney | let me look a second | 15:03 |
| juliank | laney: i'll try 54 | 15:03 |
| laney | hang on | 15:04 |
| laney | look in `juju debug-log`, the blooming disk is full | 15:04 |
| juliank | laney: still failing at processing that option / unlucde file | 15:05 |
| laney | where are you seeing this? | 15:05 |
| laney | when running mojo? | 15:07 |
| laney | juliank: ok resolved with a hack, removing the unused lxd snap to free up space | 15:13 |
| laney | mojo was unhappy because the reverted charm doesn't have that config item | 15:13 |
| juliank | laney: as usual, (mojo-prod-proposed-migration)prod-proposed-migration@wendigo:~$ | 15:13 |
| laney | probably want to reset the checkout there to correspond to 108/56 I guess | 15:14 |
| laney | we should add constraints to get bigger web machines, 10G is not good | 15:14 |
| laney | let me do that | 15:14 |
| laney | done | 15:15 |
| laney | now we need to ask IS to resize the existing instances to match that, let me do that too | 15:15 |
| schopin | doko: do you have any problem with me fixing the c++17-related errors in log4shib, including its public headers? You added a -std=g++14 flag but the headers cause failures in other packages :-( | 15:15 |
| juliank | laney: if we resize, we should also add a 2nd cpu | 15:17 |
| juliank | this should get rid of all spikes we have now | 15:17 |
| doko | schopin, well, I uploaded these other packages as well, so these should be fine now ... | 15:18 |
| laney | juliank: good idea, do you want to fix the constraints for that? | 15:18 |
| laney | see the commit I just made and the constraint syntax of juju | 15:18 |
| juliank | ok | 15:18 |
| * juliank in mth | 15:18 | |
| juliank | * meeting | 15:18 |
| laney | ack | 15:19 |
| doko | schopin, I didn't get to libcifpp and libpdb-redo yet, so that's c++17 stuff as well | 15:19 |
| doko | schopin, so maybe let the current uploads migrate first before touching them again | 15:20 |
| schopin | doko: you did xmltooling as well? | 15:21 |
| laney | juliank: what I don't know is if we `mojo run` now that we changed the constraints, will it notice and decide to add new machines ... | 15:21 |
| laney | well I guess that's not a bad way to do it really | 15:21 |
| laney | could hack in there, stop download-all-results, and copy the DB over | 15:22 |
| laney | second option if the resize route takes a long time | 15:22 |
| juliank | hmm | 15:22 |
| doko | schopin, are you subscribed to the impish-changes ML? | 15:24 |
| schopin | up until now I didn't see the point of it except drowning me in mail. I guess I'll start setup some filters then ;) | 15:25 |
| doko | https://lists.ubuntu.com/mailman/listinfo/impish-changes | 15:25 |
| vorlon | mwhudson: looks like shim-signed still needs to migrate on bionic... | 15:31 |
| xnox | laney: i lost track of things, sorry. | 15:44 |
| laney | ah well | 16:07 |
| === alan_g__ is now known as alan_g | ||
| laney | juliank: ah yay, IS did the migration quickly ♥ | 19:16 |
| laney | only one core still, sad | 19:17 |
| mwhudson | dbungert, vorlon: thanks for the livecd-rootfs review | 21:02 |
| dbungert | you bet | 21:03 |
| mwhudson | schopin, doko: all the autopkgtests still fail though because they are not using -std=c++14 | 21:41 |
| mwhudson | i think patching the source would be a more useful fix | 21:42 |
| bdmurray | mwhudson: The apport autopkgtests are failing with the new glibc fwict due to the gdb fields changing. https://autopkgtest.ubuntu.com/results/autopkgtest-impish/impish/amd64/a/apport/20210815_215748_e3421@/log.gz I've a fix for it https://paste.ubuntu.com/p/d6wPQSrdK4/ but wonder if that's best. | 21:50 |
| jawn-smith | Is any core dev available to retry https://autopkgtest.ubuntu.com/request.cgi?release=impish&arch=amd64&package=syncthing&trigger=glibc/2.34-0ubuntu1 | 22:01 |
| jawn-smith | It's one of the packages blocking glibc, but it appears to just be a flakey test based on some older test results | 22:02 |
| bdmurray | jawn-smith: I can do that | 22:04 |
| jawn-smith | thanks bdmurray! | 22:07 |
| mwhudson | bdmurray: yeah, i mean that patch looks like it would help but as usual i have to ask "what is this test testing"? | 22:14 |
| bdmurray | mwhudson: I'm not entirely sure but _validate_gdb_fields is used in almost every test | 22:15 |
| mwhudson | bdmurray: yeah so i've looked at the code and i'm still not really sure what it's trying to do | 22:54 |
| mwhudson | bdmurray: does anything else read the Stacktrace field or is it mostly for humans to read? | 22:55 |
| bdmurray | mwhudson: ah, that's a good question so apport generates a StacktraceTop and StacktraceAddressSignature from the Stacktrace | 23:00 |
| mwhudson | bdmurray: so viewing this as an integration test, it is kinda important that we find out if gdb changes output in a way that breaks crash_signature_addresses | 23:02 |
| mwhudson | bdmurray: but i'm not sure anything is really testing that? | 23:04 |
| mwhudson | all the tests of crash_signature_addresses i'm seeing use canned input | 23:04 |
| mwhudson | (which makes sense for a unit test, of course) | 23:04 |
| bdmurray | okay, I'll keep digging | 23:07 |
| mwhudson | thanks | 23:07 |
| mwhudson | i think if we can convince ourselves the signatures are still ok in the current state of impish-proposed, your patch is fine | 23:07 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!