[listenbrainz-android] 14dependabot[bot] opened pull request #412 (03dev…dependabot/gradle/dev/androidx.lifecycle-lifecycle-livedata-ktx-2.8.0): Bump androidx.lifecycle:lifecycle-livedata-ktx from 2.7.0 to 2.8.0 https://github.com/metabrainz/listenbrainz-andr...
[listenbrainz-android] 14dependabot[bot] opened pull request #413 (03dev…dependabot/gradle/dev/androidx.lifecycle-lifecycle-viewmodel-ktx-2.8.0): Bump androidx.lifecycle:lifecycle-viewmodel-ktx from 2.7.0 to 2.8.0 https://github.com/metabrainz/listenbrainz-andr...
[listenbrainz-android] 14dependabot[bot] opened pull request #414 (03dev…dependabot/gradle/dev/paging_version-3.3.0): Bump paging_version from 3.2.1 to 3.3.0 https://github.com/metabrainz/listenbrainz-andr...
[listenbrainz-android] 14dependabot[bot] opened pull request #415 (03dev…dependabot/gradle/dev/androidx.lifecycle-lifecycle-runtime-ktx-2.8.0): Bump androidx.lifecycle:lifecycle-runtime-ktx from 2.7.0 to 2.8.0 https://github.com/metabrainz/listenbrainz-andr...
[listenbrainz-android] 14dependabot[bot] opened pull request #416 (03dev…dependabot/gradle/dev/androidx.lifecycle-lifecycle-runtime-compose-2.8.0): Bump androidx.lifecycle:lifecycle-runtime-compose from 2.7.0 to 2.8.0 https://github.com/metabrainz/listenbrainz-andr...
[listenbrainz-android] 14dependabot[bot] opened pull request #417 (03dev…dependabot/gradle/dev/androidx.lifecycle-lifecycle-viewmodel-compose-2.8.0): Bump androidx.lifecycle:lifecycle-viewmodel-compose from 2.7.0 to 2.8.0 https://github.com/metabrainz/listenbrainz-andr...
adhawkins: Never happened during tests, did you skip any step(s)? The script `upgrade-to-postgres16` is also rebuilding images for containers at step 4.
reosarevok
aerozol: I can check
aerozol: so grindcore :p
aerozol: seems good, made a couple tiny changes
SigHunter has quit
SigHunter joined the channel
aerozol[m] joined the channel
aerozol[m]
Thants reosarevok! I just realized what you meant with grindcore haha - I'm quite pleased with that intro 😛
MonkeyPython has an instrument post coming, was thinking of publishing it tomorrow, does that seem good?
SigHunter has quit
MonkeyPython
aerozol: 'd like to get a confirmation on this so I can squedule the blogpost asap. rest of the week is gonna be reall hecktic for me so I want to be finished with this :D
reosarevok
MonkeyPython: just schedule it for now, we can change it a bit if needed :)
MonkeyPython
👍
KassOtsimine is now known as ahvalmissaamine
SigHunter joined the channel
aerozol[m]
Reosarevok: MonkeyPython: no hurry with the other blog post, go for it!
reosarevok: screenshot re pull 3256 looks fine to me
reosarevok
Thanks
ahvalmissaamine
wait so publish it right now or???
i'm fine with that, or tomorrow
SigHunter has quit
d4rkie has quit
d4rkie joined the channel
adhawkins
yvanzo: This is my history from the session (not sure why I did the fetch and checkout multiple times!): https://paste.debian.net/1317004/
bitmap: git rev-parse HEAD shows f362446959a9f0984f0a85be1b20a96aa73872ea
reosarevok
ahvalmissaamine: right now seems fine too I guess :)
MonkeyPython
oh!
yvanzo
adhawkins: nothing wrong there.
same for your HEAD commit.
adhawkins
yvanzo: Just re-ran the 'upgrade-db-schema' and it's giving the same error, looks like it's running 20211203-mbs-11312.sql which is years old...
yvanzo
adhawkins: Does `docker-compose exec musicbrainz grep MUSICBRAINZ_DB_SCHEMA_SEQUENCE /usr/local/bin/upgrade-db-schema.sh`give you 28?
adhawkins
27
yvanzo
You probably have old images. Did the upgrade to Postgres 16 work?
adhawkins
I don't recall any errors there. Let me re-run that step.
yvanzo
I’m not sure it is reentrant, just check the PG version: `docker-compose exec db psql --version`
adhawkins
16.3
If I re-run the db upgrade script it says 'The data directory was initialized by PostgreSQL version 16, which is not compatible with this version 12.18 (Debian 12.18-1.pgdg120+2).'
yvanzo
And has the database content been upgraded too? `docker-compose exec db psql -U musicbrainz -d musicbrainz_db -tA -P pager=off -c 'SHOW server_version_num'`
SigHunter joined the channel
adhawkins
"psql: error: connection to server on socket "/var/run/postgresql/.s.PGSQL.5432" failed: No such file or directory Is the server running locally and accepting connections on that socket?"
yvanzo
So the PG upgrade probably failed for some reason.
adhawkins
I did re-run the pg upgrade script, which may have been a mistake.
yvanzo
Ok
adhawkins
Sadly, I forgot to take a snapshot of the VM before starting this process too. I have a snapshot from 27 days ago, which I think is the previous version but before I applied Debian updates. I could start from that again?
yvanzo
Yes but we can try a bit more beforehand.
`docker-compose stop; docker-compose rm db; docker-compose build; docker-compose up -d`
reosarevok
adhawkins, yvanzo: fwiw my psql16 started running in 5433 since psql12 was in 5432
yvanzo
Then the above `docker-compose exec musicbrainz grep MUSICBRAINZ_DB_SCHEMA_SEQUENCE /usr/local/bin/upgrade-db-schema.sh`should give you 28.
reosarevok
I had to change the conf by hand
(not docker tho(
adhawkins
Ok, I now get schema 28. Re-run the schema upgrade?
yvanzo
adhawkins: yes!
adhawkins
Ok, it's vaccuming.
yvanzo
reosarevok: Probably because you were running both servers on the same host?
reosarevok
Yeah, I guess :)
kellnerd
aerozol[m]: Blog post looks good to me, very nice!
adhawkins
Ok, so looks like it failed to update the db container?
Ok, doesn't look like I have a log from that command. I'll try to remember to start recording the whole session in future.
Would you like me to revert and re-run it to see if it happens again?
yvanzo
It seems to be a unique event, so it should be fine, I will just add troubleshooting notes just in case someone is getting the same problem.
adhawkins
I'm happy to redo if it might help. Up to you.
The local mirror only gets uses overnight to run my tagger against my whole music collection.
atj
that's what I call dedication
adhawkins
Me atj? I'm a bit anal like that :)
yvanzo
The `screen` is very useful for that, with the options `-L -Logfile screen.that.will.blow.up.%n.log`
adhawkins
Ok. Shall I do that then yvanzo? No trouble really.
atj
adhawkins: I think it's safe to stay you're among friends ;)
adhawkins
:) atj
atj
s/stay/say/
yvanzo
atj: That’s why we are always pointing to the same song by Rick Astley ;)
adhawkins
still getting collation mismatches: The collation in the database was created using version 153.14, but the operating system provides version 153.120.
aerozol[m]
Thank you kellnerd
(for extra blog post checking)
yvanzo
adhawkins: VACUUM ended successfully already?
adhawkins
Not yet. Could be hours couldn't it?
reosarevok
Yep. Efficiency abhors a VACUUM
yvanzo
How can you see the collation mismatches then?
SigHunter has quit
It took 2h+ last time I tested it.
I mentioned your initial error in the release notes just in case, thanks for having reported it.
If the collation warnings persist after having following all the release instructions, please let us know.
adhawkins
Will do yvanzo. I'll let this complete, then revert to my snapshot and run it again. I'll provide the entire log from screen when done.
Ah, vacuum done. Will run the rest of the steps now.
Ok, it's complete. Did get the collation message during that initial replication run (the one limited to 1)
yvanzo
ansh, lucifer: The MB mirror on `wolf` is up again. Feel free to restart your containers at any time.
adhawkins: Interesting, maybe it is due to the undetermined failure during the Postgres 16 upgrade step.
adhawkins
Ok. I'll revert to my 18th April snapshot, run the Debian updates, allow it to replicate up to the schema change, then re-run the steps logging everything.
I think I've been getting that collation error for ages.
s/error/warning/
yvanzo
I‘ve been getting it for a long time too, but it vanished after upgrading to Postgres 16.
bitmap: I’m actually seeing the AMQP warning sometimes in replication logs, not just for the first replication packet in schema 29.
adhawkins
yvanzo: Ok, I've got the collation warning in the replication I'm doing on the old version applying all the packets from April 18th :)
SigHunter joined the channel
d4rkie has quit
d4rkie joined the channel
mayhem
well, writing a shitty blog post seemed to have helped with SSL.com -- either that or it was total coincidence... but they are finally refunding our money.
zas
\o/
SigHunter has quit
SigHunter_ joined the channel
SigHunter_ has quit
SigHunter joined the channel
reosarevok
"Delivery Attempted - On the Way to a Local UPS Access Point™"
SigHunter has quit
lol UPS. Sure you did, been home all day - apparently I get to pick up my expensive package from... the CBD store 🙄
The magic of delivery
mayhem
well, kill two birds with one stone. :)
reosarevok
But the birds are already stoned... oh, wait, CBD
mayhem
are they chill??
besides, birds are not real.
SigHunter joined the channel
SigHunter has quit
SigHunter joined the channel
reosarevok: on latest support email
reosarevok
heh. Ok
mayhem
of course he now claims to not have threatended legal action. he's reported us to our "office".
mayhem notes that no-one appeared at the office.
reosarevok
I like "TO: WHOMEVER"
mayhem
and the notice titled "LEGAL NOTICE" that isn't a legal threat. clever.
reosarevok
Well, it's not a threat, it's just he noticed something... legal?