#metabrainz

/

      • yvanzo[m]
        bitmap: How long did that take with your instance?
      • 2025-05-20 14019, 2025

      • bitmap[m]
        yvanzo[m]: not very long, under 10 minutes I think
      • 2025-05-20 14007, 2025

      • yvanzo[m]
        bitmap: Did you save your changes to the blog post? Git tags are still missing .0
      • 2025-05-20 14049, 2025

      • bitmap[m]
        I thought so, how about now?
      • 2025-05-20 14007, 2025

      • yvanzo[m]
        Good!!
      • 2025-05-20 14032, 2025

      • yvanzo[m]
        bitmap: For your suggested step 6, I got:... (full message at <https://matrix.chatbrainz.org/_matrix/media/v3/download/chatbrainz.org/sTFGmiRgedZrRFqZQPWyyREm>)
      • 2025-05-20 14012, 2025

      • yvanzo[m]
        (I updated the release notes without this step for now.)
      • 2025-05-20 14004, 2025

      • bitmap[m]
        the release_first_release_date message is expected, since it tries to build that if you specify recording_first_release_date (since the latter depends on the former)
      • 2025-05-20 14040, 2025

      • bitmap[m]
        the second message is not expected, since the schema 30 upgrade script does TRUNCATE recording_first_release_date;, so something must have went wrong
      • 2025-05-20 14010, 2025

      • yvanzo[m]
        It is building artist_release atm.
      • 2025-05-20 14042, 2025

      • bitmap[m]
        how many rows are in your recording_first_release_date table?
      • 2025-05-20 14010, 2025

      • yvanzo[m]
        I do have this in upgrade db logs though:... (full message at <https://matrix.chatbrainz.org/_matrix/media/v3/download/chatbrainz.org/kKEkDgfvVPOffVIEUlMdlEws>)
      • 2025-05-20 14020, 2025

      • bitmap[m]
        yeah, looks like the TRUNCATE was successful then...unless MAINTENANCE is not connecting to the same database, I'm quite confused
      • 2025-05-20 14054, 2025

      • yvanzo[m]
        446
      • 2025-05-20 14043, 2025

      • yvanzo[m]
      • 2025-05-20 14021, 2025

      • yvanzo[m]
        Might it be due to SIR triggers?
      • 2025-05-20 14031, 2025

      • Kladky joined the channel
      • 2025-05-20 14046, 2025

      • yvanzo[m]
        (There has been no change about these tables in SIR v4.)
      • 2025-05-20 14004, 2025

      • bitmap[m]
        weird, the only way data gets inserted into recording_first_release_date is via MBS triggers or the BuildMaterializedTables script
      • 2025-05-20 14044, 2025

      • bitmap[m]
        so unless there was some concurrent script running, like replication, I'm not sure where that data would have come from
      • 2025-05-20 14034, 2025

      • yvanzo[m]
        Yes, the previous step 5. is loading one replication packet.
      • 2025-05-20 14056, 2025

      • bitmap[m]
        🤦‍♂️
      • 2025-05-20 14015, 2025

      • bitmap[m]
        that's my bad, I could have sworn step 5 was running upgrade.sh, it should happen directly after that
      • 2025-05-20 14030, 2025

      • yvanzo[m]
        Can you please add it?
      • 2025-05-20 14024, 2025

      • _BrainzGit
        [musicbrainz-docker] 14yvanzo merged pull request #305 (03master…mbdb30-legacy-solr7): Upgrade MB DB schema version to 30 and downgrade Solr version to 7 https://github.com/metabrainz/musicbrainz-docker/…
      • 2025-05-20 14044, 2025

      • bitmap[m]
        ok, added it (correctly as step 5)
      • 2025-05-20 14004, 2025

      • _BrainzGit
        [musicbrainz-docker] release 03v-2025-05-20.0-mbdb30 has been published by 14yvanzo: https://github.com/metabrainz/musicbrainz-docker/…
      • 2025-05-20 14041, 2025

      • yvanzo[m]
        bitmap: Just added a P.S. about search before the instructions to the blog post.
      • 2025-05-20 14057, 2025

      • yvanzo[m]
        Also mentioned fetching EAA dump in mb-docker release notes.
      • 2025-05-20 14010, 2025

      • yvanzo[m]
        bitmap: Putting back in place the container for dumping (Solr 7) search indexes.
      • 2025-05-20 14029, 2025

      • yvanzo[m]
        (For Solr 9 backups, I will wait for full recording reindex.)
      • 2025-05-20 14005, 2025

      • yvanzo[m]
        bitmap: May I remove the file `aretha_mb_export.sh` and start services as usual?
      • 2025-05-20 14012, 2025

      • bitmap[m]
        yvanzo: I'm not sure what that file is 😄
      • 2025-05-20 14040, 2025

      • yvanzo[m]
        in aretha, starting sshd fullexport only.
      • 2025-05-20 14050, 2025

      • yvanzo[m]
        Back in place, search indexes dump will start as usual.
      • 2025-05-20 14030, 2025

      • bitmap[m]
        yvanzo: not sure who added that file but it's not useful to me at least
      • 2025-05-20 14040, 2025

      • yvanzo[m]
        Also replaced May 2024 with May 2025 in our blog post 🤦.
      • 2025-05-20 14001, 2025

      • bitmap[m]
        d'oh
      • 2025-05-20 14055, 2025

      • yvanzo[m]
        Can we specify “database” schema change? (As there at least two other schemas for MusicBrainz: XML & Solr.)
      • 2025-05-20 14043, 2025

      • bitmap[m]
        sure, can't hurt to be more specific
      • 2025-05-20 14026, 2025

      • yvanzo[m]
        bitmap: Ready to publish?
      • 2025-05-20 14046, 2025

      • bitmap[m] quickly proofreads it once more
      • 2025-05-20 14022, 2025

      • bitmap[m]
        yvanzo: ready :)
      • 2025-05-20 14049, 2025

      • bitmap[m] publishes it
      • 2025-05-20 14024, 2025

      • yvanzo[m] fixed the link from the release notes.
      • 2025-05-20 14012, 2025

      • zas[m]
        yvanzo: there are a lot of zombies on rakim, is that related to sir?
      • 2025-05-20 14058, 2025

      • zas[m] sent a code block: https://matrix.chatbrainz.org/_matrix/media/v3/download/chatbrainz.org/JeFNLStRhZpjZejrRarMPshc
      • 2025-05-20 14025, 2025

      • zas[m]
        Times matches issues you reported with connections/sir on rakim
      • 2025-05-20 14006, 2025

      • yvanzo[m]
        zas: Yes it is the ongoing reindex of the recording collection in Solr.
      • 2025-05-20 14018, 2025

      • zas[m]
        ok