#metabrainz

/

      • reosarevok
        But probably a bit better for us to debug at least?
      • 2023-02-02 03304, 2023

      • reosarevok
        aerozol: oh, I guess you're not an account admin in MB yet? We should change that, I guess :) (mayhem, I assume that's ok?)
      • 2023-02-02 03321, 2023

      • mayhem
        on mb.org ? fine.
      • 2023-02-02 03344, 2023

      • aerozol
        Uh oh, moe admin, moe problems
      • 2023-02-02 03344, 2023

      • Aqib joined the channel
      • 2023-02-02 03307, 2023

      • reosarevok
        aerozol: you should be able to see people's emails now
      • 2023-02-02 03336, 2023

      • reosarevok
        (also delete them, mark them as spammers, etc - we can talk about all those things when needed :) )
      • 2023-02-02 03358, 2023

      • aerozol
        Thanks reo, I wont touch any of that unless asked to I think
      • 2023-02-02 03316, 2023

      • reosarevok
        Well, if you find a spammer it's safe to mark them as such anyway
      • 2023-02-02 03340, 2023

      • aerozol
        I will now start emailing CB and LB problem users straight from that email account - speaking of which, I think you said I had access, but how?
      • 2023-02-02 03356, 2023

      • aerozol
        Hmm okay, I might use the spammer button in that case
      • 2023-02-02 03300, 2023

      • reosarevok
        Oh, I don't email *from* the account
      • 2023-02-02 03312, 2023

      • reosarevok
        I email from mine, but cc the account so there's a paper trail
      • 2023-02-02 03327, 2023

      • reosarevok
        Ideally we'd also set the account as our reply-to, so their replies also go automatically to everyone
      • 2023-02-02 03339, 2023

      • monkey
        Hi, sorry I missed the database party. Anything I need to do?
      • 2023-02-02 03332, 2023

      • mayhem
        nope, all good.
      • 2023-02-02 03347, 2023

      • monkey
        👍
      • 2023-02-02 03303, 2023

      • aerozol
        okay reo, what’s the account I’m ccing in? ‘musicbrainz-account-admin’?
      • 2023-02-02 03330, 2023

      • reosarevok
        Yup
      • 2023-02-02 03332, 2023

      • reosarevok
        @meb
      • 2023-02-02 03319, 2023

      • yvanzo
        ...for how long has the database party been ended?
      • 2023-02-02 03323, 2023

      • yvanzo
        did the SSH key for floyd change?
      • 2023-02-02 03352, 2023

      • yvanzo
        (host key)
      • 2023-02-02 03323, 2023

      • reosarevok
        yvanzo: I understand floyd is down for repairs?
      • 2023-02-02 03331, 2023

      • yvanzo
        thank you
      • 2023-02-02 03319, 2023

      • yvanzo
        MB search indexer is still off atm.
      • 2023-02-02 03318, 2023

      • YashP joined the channel
      • 2023-02-02 03326, 2023

      • reosarevok
        Can you point it to pink? I'm not sure how long floyd'll be down, but it might be a while?
      • 2023-02-02 03308, 2023

      • YashP has quit
      • 2023-02-02 03336, 2023

      • yvanzo
        It was running on floyd.
      • 2023-02-02 03325, 2023

      • BrainzGit
        [musicbrainz-server] 14mwiencek opened pull request #2819 (03master…mbs-12863): MBS-12863: Tooltips too far to the left https://github.com/metabrainz/musicbrainz-server/…
      • 2023-02-02 03344, 2023

      • reosarevok
        Oh
      • 2023-02-02 03352, 2023

      • reosarevok
        Then I guess we need to put it up elsewhere?
      • 2023-02-02 03357, 2023

      • reosarevok
        zas: ^
      • 2023-02-02 03342, 2023

      • BrainzGit
        [listenbrainz-server] 14amCap1712 opened pull request #2352 (03master…artist-credit-similarity): Use artist credits for artist similarity https://github.com/metabrainz/listenbrainz-server…
      • 2023-02-02 03309, 2023

      • lucifer
      • 2023-02-02 03354, 2023

      • lucifer
        $150 for 500 API request 🤨
      • 2023-02-02 03319, 2023

      • Lotheric has quit
      • 2023-02-02 03333, 2023

      • zas
        yes, floyd is down for now, no ETA from Hetzner
      • 2023-02-02 03352, 2023

      • aerozol
        aww rest up little CritiqueBrainz
      • 2023-02-02 03335, 2023

      • Lotheric joined the channel
      • 2023-02-02 03344, 2023

      • yvanzo
        lucifer: feel lucky they didn’t go for the opposite: $500 for 150 req 🤦
      • 2023-02-02 03358, 2023

      • aerozol
        A lot of funny replies on that twitter thread haah
      • 2023-02-02 03313, 2023

      • yvanzo
        does it still compete with SMS?
      • 2023-02-02 03300, 2023

      • reosarevok
        Are they selling API or NFTs?
      • 2023-02-02 03346, 2023

      • yvanzo
        zas: can I make SIR run on another server (e.g. ludwig) in the meantime?
      • 2023-02-02 03301, 2023

      • reosarevok
        bitmap: heh, good catch with my missing that condition for the checkbox
      • 2023-02-02 03323, 2023

      • Aqib has quit
      • 2023-02-02 03307, 2023

      • zas
        yvanzo: sure
      • 2023-02-02 03329, 2023

      • atj
        i don't get why hetzner don't proactively replace fans, it's obviously a known issue
      • 2023-02-02 03331, 2023

      • atj
      • 2023-02-02 03338, 2023

      • atj
      • 2023-02-02 03305, 2023

      • pbryan
        Does production or beta BookBrainz expose an API endpoint?
      • 2023-02-02 03346, 2023

      • serialata joined the channel
      • 2023-02-02 03311, 2023

      • Zhele has quit
      • 2023-02-02 03341, 2023

      • Zhele joined the channel
      • 2023-02-02 03322, 2023

      • zas
        floyd is back, and after extensive testing, Hetzner didn't find any issue, so I guess that's a kernel bug causing incorrect report of excessive temperature, there are reports in this sense. Unclear what's causing it though, some threads indicate a firmware issue
      • 2023-02-02 03322, 2023

      • zas
        bitmap: I did few extra checks, including cpu stress tests, and no message in logs anymore... on reboot, new kernel + new cpu microcode so, may be, the issue doesn't exist anymore. It should be noted that temperatures reported by sensors were correct, only the repeated message in kern logs was indicating a high temp issue. A search on the web shows that some people got this message while no overheating did actually happen.
      • 2023-02-02 03303, 2023

      • zas
        I just did a test setting all cpus (16) on 100% for 2 minutes, no problem
      • 2023-02-02 03330, 2023

      • zas
        so I guess we can move services back to floyd (or make floyd secondary as you prefer)
      • 2023-02-02 03359, 2023

      • zas
        atj: ^^ read above about floyd / cpu temp
      • 2023-02-02 03333, 2023

      • atj
        zas: thanks, v weird bug
      • 2023-02-02 03305, 2023

      • zas
        that said, setting all cpus on 100% for 2 minutes generated an alert, but this one is based on sensors, cpu temp 100°C
      • 2023-02-02 03338, 2023

      • zas
        though no 'Core temperature is above threshold' in logs...
      • 2023-02-02 03333, 2023

      • zas
        this alert on temp when stressing cpus seem to show there's a cooling problem, because 100°C should never happen with a proper cooling...
      • 2023-02-02 03344, 2023

      • zas
        let me check intel recommendations on this
      • 2023-02-02 03344, 2023

      • atj
        i think hetzner run their DCs hot
      • 2023-02-02 03337, 2023

      • zas
        I run the test again, for 4 minutes
      • 2023-02-02 03359, 2023

      • zas
        if the cpu protection triggers we should see a message about throttling
      • 2023-02-02 03305, 2023

      • zas
        after 4 minutes at 100% on all cpus, temp went to 101°C, slowly increasing, but no throttling
      • 2023-02-02 03344, 2023

      • zas
        I'm checking actual cpu frequencies, and nope, no throttling yet, I'll do a more violent test
      • 2023-02-02 03302, 2023

      • zas
        right now one can cook on floyd's cpu
      • 2023-02-02 03311, 2023

      • zas
        100°C for 4 minutes then for 5 minutes, 100% cpu on all cores, no throttling, max cpu freq the whole time https://usercontent.irccloud-cdn.com/file/XEb1dp2…
      • 2023-02-02 03349, 2023

      • zas
        and temperature drops as soon load decrease
      • 2023-02-02 03326, 2023

      • zas
        so I guess we can conclude there's no real issue
      • 2023-02-02 03334, 2023

      • trolley has quit
      • 2023-02-02 03321, 2023

      • pbryan
        Anyone know if there's a BB prod/beta API endpoint?
      • 2023-02-02 03329, 2023

      • trolley joined the channel
      • 2023-02-02 03354, 2023

      • zas
        bitmap, yvanzo: floyd is back to work
      • 2023-02-02 03344, 2023

      • bitmap
        zas: thanks, I'll work on restoring the standby service soon, but we can keep pink as primary for now if it won't cause any issues. (at least to avoid further downtime)
      • 2023-02-02 03354, 2023

      • bitmap
        pbryan: is this what you're looking for? https://api.test.bookbrainz.org/1/docs/
      • 2023-02-02 03318, 2023

      • bitmap
        oh, that's test, sorry.
      • 2023-02-02 03335, 2023

      • pbryan
        Yeah, database is there is a bit thin on the ground.
      • 2023-02-02 03341, 2023

      • pbryan
        Looking for access to real data.
      • 2023-02-02 03330, 2023

      • bitmap
        I can't find one documented but monkey would know for sure
      • 2023-02-02 03337, 2023

      • bitmap peeked in docker-server-configs and only sees a test instance deployed, so probably not
      • 2023-02-02 03328, 2023

      • aerozol
        bitmap: looks like that forum thread is resolved? A no from reo on one change, a thumbs up from jesus2099 on the rest?
      • 2023-02-02 03336, 2023

      • aerozol
        gg
      • 2023-02-02 03359, 2023

      • bitmap
        aerozol: yeah I think it's probably resolved for now, thanks for checking
      • 2023-02-02 03342, 2023

      • pbryan
        Thanks, bitmap.