#metabrainz

/

      • ruaok
        db.data.get_lowlevel_id(rec_1[0], rec_1[1])
      • 2019-10-01 27433, 2019

      • alastairp
        yeah
      • 2019-10-01 27442, 2019

      • ruaok
        the second argument indicates which track of the possible duplicates to use.
      • 2019-10-01 27448, 2019

      • alastairp
        input is mbid,offset as in AB, but annoy uses rowid as an index
      • 2019-10-01 27408, 2019

      • ruaok
        k
      • 2019-10-01 27444, 2019

      • D4RK-PH0ENiX has quit
      • 2019-10-01 27434, 2019

      • thomasross has quit
      • 2019-10-01 27442, 2019

      • thomasross joined the channel
      • 2019-10-01 27406, 2019

      • D4RK-PH0ENiX joined the channel
      • 2019-10-01 27402, 2019

      • ssquare joined the channel
      • 2019-10-01 27406, 2019

      • bitmap
        reosarevok: if you rebase your PRs on master, tests should work again
      • 2019-10-01 27414, 2019

      • bitmap
        I'll fix the one I was messing with
      • 2019-10-01 27458, 2019

      • travis-ci joined the channel
      • 2019-10-01 27458, 2019

      • travis-ci
        metabrainz/picard#5020 (master - 8d08a16 : Laurent Monin): The build has errored.
      • 2019-10-01 27458, 2019

      • travis-ci
      • 2019-10-01 27458, 2019

      • travis-ci
      • 2019-10-01 27458, 2019

      • travis-ci has left the channel
      • 2019-10-01 27438, 2019

      • thomasross_ joined the channel
      • 2019-10-01 27438, 2019

      • thomasross has quit
      • 2019-10-01 27438, 2019

      • thomasross_ is now known as thomasross
      • 2019-10-01 27442, 2019

      • ruaok
        Mr_Monkey: making slow progress -- deploying on bono interestingly slow
      • 2019-10-01 27452, 2019

      • travis-ci joined the channel
      • 2019-10-01 27452, 2019

      • travis-ci
        metabrainz/picard#5020 (master - 8d08a16 : Laurent Monin): The build passed.
      • 2019-10-01 27452, 2019

      • travis-ci
      • 2019-10-01 27452, 2019

      • travis-ci
      • 2019-10-01 27452, 2019

      • travis-ci has left the channel
      • 2019-10-01 27418, 2019

      • ruaok copies the config to his own account to save the commits for debugging
      • 2019-10-01 27456, 2019

      • aidanlw17
        ruaok: deploying on bono is slow because of where the annoy indices are currently located
      • 2019-10-01 27413, 2019

      • ruaok
        do they need to be added to dockerignore?
      • 2019-10-01 27430, 2019

      • aidanlw17
        We need to copy them into /code every time
      • 2019-10-01 27404, 2019

      • aidanlw17
        alastairp can explain this better, but I think it’s that they need to be saved in a separate volume so they aren’t stored with the source code
      • 2019-10-01 27436, 2019

      • reosarevok
        bitmap: whoohoo
      • 2019-10-01 27444, 2019

      • ruaok
        I see, ya. that would help a lot.
      • 2019-10-01 27448, 2019

      • aidanlw17
        Yeah. It’s something we are meaning to change
      • 2019-10-01 27420, 2019

      • alastairp
        oh yeah, putting indexes in a volume would be much better
      • 2019-10-01 27443, 2019

      • ruaok
      • 2019-10-01 27406, 2019

      • ruaok
        I think I may have broken what was running on bono in an effort to get it setup in my account.
      • 2019-10-01 27422, 2019

      • alastairp
        mmm
      • 2019-10-01 27428, 2019

      • ruaok
        did I knock an annoy server offline?
      • 2019-10-01 27434, 2019

      • alastairp
        possibly
      • 2019-10-01 27414, 2019

      • alastairp
        see docker/docker-compose.bono.yml - it sets `image: acousticbrainz-server-path`
      • 2019-10-01 27423, 2019

      • alastairp
        and this is replicated in ./docker-compose.yml
      • 2019-10-01 27437, 2019

      • alastairp
        also in ./docker-compose.yml is that port definition (8088)
      • 2019-10-01 27413, 2019

      • ruaok
        are there containers running from more than one source tree?
      • 2019-10-01 27434, 2019

      • alastairp
        yeah, because I never planned on running more than 1 AB server on this server
      • 2019-10-01 27444, 2019

      • alastairp
        so each new one that we added we had to add more hacks
      • 2019-10-01 27453, 2019

      • ruaok
        ok.
      • 2019-10-01 27418, 2019

      • alastairp
        I can try and simplify it a bit more, but probably not today
      • 2019-10-01 27449, 2019

      • alastairp
        as I said, I set it up quickly to do testing on the main similarity stuff and work around not having consul
      • 2019-10-01 27452, 2019

      • ruaok
        got a moment to help bring this setup back?
      • 2019-10-01 27402, 2019

      • alastairp
        you put something in your home?
      • 2019-10-01 27408, 2019

      • ruaok
        > as I said, I set it up quickly to do testing on the main similarity stuff and work around not having consul
      • 2019-10-01 27411, 2019

      • ruaok
        no worries.
      • 2019-10-01 27422, 2019

      • alastairp
        what do you want to do? make the original one continue working, and then make yours work for dev?
      • 2019-10-01 27432, 2019

      • ruaok
        well, I gave up on that since I couldnt get it back to work.
      • 2019-10-01 27449, 2019

      • ruaok
        the goal is the get the similarity server running again.
      • 2019-10-01 27435, 2019

      • alastairp
        lol
      • 2019-10-01 27451, 2019

      • alastairp
        OK, I just put indices in a volume, so bulding is fast again
      • 2019-10-01 27408, 2019

      • aidanlw17
        !m alastairp
      • 2019-10-01 27408, 2019

      • BrainzBot
        You're doing good work, alastairp!
      • 2019-10-01 27404, 2019

      • travis-ci joined the channel
      • 2019-10-01 27404, 2019

      • travis-ci
        metabrainz/picard#5023 (master - 64c4153 : Philipp Wolfer): The build passed.
      • 2019-10-01 27404, 2019

      • travis-ci
      • 2019-10-01 27404, 2019

      • travis-ci
      • 2019-10-01 27404, 2019

      • travis-ci has left the channel
      • 2019-10-01 27438, 2019

      • alastairp
        ruaok: working again
      • 2019-10-01 27444, 2019

      • ruaok
        <3
      • 2019-10-01 27455, 2019

      • ruaok
        back to bad gateway. :) :)
      • 2019-10-01 27457, 2019

      • ruaok
        thanks!
      • 2019-10-01 27400, 2019

      • alastairp
        for some reason we didn't say that nginx depended on the webserver image in docker-compose, it had never failed before
      • 2019-10-01 27412, 2019

      • alastairp
        but this time it decided to start nginx before uwsgi, and so nginx crashed
      • 2019-10-01 27437, 2019

      • alastairp
        running it a few more times in a row would have probably eventually fixed it :-P
      • 2019-10-01 27408, 2019

      • alastairp
        I'm not 100% sure that the volume is correct, can someone hit our endpoint to ensure it works? aidanlw17?
      • 2019-10-01 27422, 2019

      • aidanlw17
        I can
      • 2019-10-01 27426, 2019

      • aidanlw17
        One sec
      • 2019-10-01 27426, 2019

      • aidanlw17
        Endpoint is working
      • 2019-10-01 27437, 2019

      • aidanlw17
        Endpoint is working
      • 2019-10-01 27411, 2019

      • alastairp
        it's working twice
      • 2019-10-01 27419, 2019

      • alastairp
        that's twice as good as ruaok's!
      • 2019-10-01 27433, 2019

      • ruaok
        solid logic!
      • 2019-10-01 27441, 2019

      • ruaok
        restarting
      • 2019-10-01 27412, 2019

      • aidanlw17
        oops lost connection in the middle of sending that! haha.
      • 2019-10-01 27437, 2019

      • aidanlw17
        ruaok: did you just re-deploy the server?
      • 2019-10-01 27443, 2019

      • ruaok
        yes
      • 2019-10-01 27448, 2019

      • aidanlw17
        ok makes sense
      • 2019-10-01 27450, 2019

      • ruaok
        doing it again, in fact.
      • 2019-10-01 27405, 2019

      • aidanlw17
        It worked 5 times, then I got `Empty reply from server`
      • 2019-10-01 27408, 2019

      • ruaok
        trying to iron out the last bits on the debug endpoint Mr_Monkey and I are building
      • 2019-10-01 27409, 2019

      • aidanlw17
        then it worked again after that
      • 2019-10-01 27414, 2019

      • ruaok
        seems right
      • 2019-10-01 27423, 2019

      • aidanlw17
        what is the debug endpoint?
      • 2019-10-01 27457, 2019

      • ruaok
        it basically returnes 100 nearest tracks for a given track, but for each metric.
      • 2019-10-01 27416, 2019

      • ruaok
        the odyssey server will then try and cobble more stuff together.
      • 2019-10-01 27444, 2019

      • ruaok
        wow, never got this one before: `OverflowError: Python int too large to convert to C long`
      • 2019-10-01 27426, 2019

      • aidanlw17
        never seen that before either
      • 2019-10-01 27413, 2019

      • ruaok
        seems that if you pass a uuid to nns function when it expects a 32 bits, you get that.
      • 2019-10-01 27419, 2019

      • ruaok is speshul
      • 2019-10-01 27422, 2019

      • ruaok
        too many small mistakes today. need maor sleep.
      • 2019-10-01 27430, 2019

      • aidanlw17
        everyone may need a long weekend
      • 2019-10-01 27440, 2019

      • aidanlw17
        do you have thanksgiving in Spain?
      • 2019-10-01 27445, 2019

      • ruaok has his planned
      • 2019-10-01 27453, 2019

      • ruaok
        no. yes.
      • 2019-10-01 27413, 2019

      • ruaok
        either my canadian or us friends will host some parties.
      • 2019-10-01 27444, 2019

      • ruaok
        but in general it isn't something celebrated here.
      • 2019-10-01 27412, 2019

      • aidanlw17
        if you have both canadian and us friends, then I guess you can celebrate it once now and once in November
      • 2019-10-01 27428, 2019

      • ruaok
        I think I am on track for that.
      • 2019-10-01 27413, 2019

      • ruaok
        omg omg omg, it didn't 502 this time.
      • 2019-10-01 27436, 2019

      • aidanlw17
        yipeee
      • 2019-10-01 27456, 2019

      • ruaok can finally continue
      • 2019-10-01 27419, 2019

      • Lotheric_ joined the channel
      • 2019-10-01 27437, 2019

      • Lotheric has quit
      • 2019-10-01 27438, 2019

      • outsidecontext
        yvanzo: are you around? do you know what happened to the Jira <-> Github integration?
      • 2019-10-01 27429, 2019

      • yvanzo
        outsidecontext: still in the train, it might be a cache issue that occurred yesterday, would require to rebuild it.
      • 2019-10-01 27459, 2019

      • outsidecontext
        yvanzo: Ok. I just noticed that links to commits / PRs shown for tickets which had previously already been linked to them are no longer displayed. Also the github hook to Jira fails. should I create an issue?
      • 2019-10-01 27438, 2019

      • yvanzo
        outsidecontext: if you have more info to report yes please, otherwise I will just update Jira tomorrow.
      • 2019-10-01 27440, 2019

      • ruaok
        Mr_Monkey:
      • 2019-10-01 27441, 2019

      • ruaok
      • 2019-10-01 27450, 2019

      • ruaok
        now to process that and return it to you.
      • 2019-10-01 27458, 2019

      • outsidecontext
        yvanzo: Ok, thanks. Good traveling
      • 2019-10-01 27406, 2019

      • Mr_Monkey
        Nice !
      • 2019-10-01 27418, 2019

      • yokel has quit
      • 2019-10-01 27449, 2019

      • yokel joined the channel
      • 2019-10-01 27447, 2019

      • aidanlw17
        ruaok: what is the distance key referring to in each of the inner metric dictionaries?
      • 2019-10-01 27441, 2019

      • ruaok
        we still provide two recordings.
      • 2019-10-01 27400, 2019

      • ruaok
        and this distance is the distance between the two recordings for that metric
      • 2019-10-01 27417, 2019

      • Mr_Monkey
        ruaok: I'm done with the improvements on the odyssey page.
      • 2019-10-01 27422, 2019

      • ruaok
        all the other listed distances are the distances from the starting point.
      • 2019-10-01 27433, 2019

      • ruaok
        Mr_Monkey: I'll have data for you in a few. good timing.
      • 2019-10-01 27436, 2019

      • ruaok puts head down
      • 2019-10-01 27429, 2019

      • aidanlw17
        Ah.
      • 2019-10-01 27439, 2019

      • aidanlw17
        Very nice
      • 2019-10-01 27404, 2019

      • ruaok
        our next immediate goal is to see if the the same tracks is returned for any two metrics.
      • 2019-10-01 27408, 2019

      • ruaok
        we wonder if that tells us something
      • 2019-10-01 27432, 2019

      • aidanlw17
        Cool, interested in seeing the outcome!
      • 2019-10-01 27414, 2019

      • thomasross has quit
      • 2019-10-01 27400, 2019

      • ssquare has quit
      • 2019-10-01 27442, 2019

      • ruaok
        Mr_Monkey: phew. finally:
      • 2019-10-01 27443, 2019

      • ruaok
      • 2019-10-01 27451, 2019

      • ruaok
        the metrics flag is superflous in this case.
      • 2019-10-01 27456, 2019

      • ruaok
      • 2019-10-01 27408, 2019

      • ruaok
        something is working. not sure what. :)
      • 2019-10-01 27415, 2019

      • alastairp
        better than "something isn't working"
      • 2019-10-01 27439, 2019

      • ruaok
        Mr_Monkey: additional_info should have dist and count keys.
      • 2019-10-01 27457, 2019

      • ssquare joined the channel
      • 2019-10-01 27441, 2019

      • ssquare
        Hello ther ,Shubham here from India.I want to be a part of this org.Ho should i start?Can anybody guide me
      • 2019-10-01 27421, 2019

      • ssquare has quit
      • 2019-10-01 27439, 2019

      • ssquare joined the channel
      • 2019-10-01 27408, 2019

      • ssquare has quit
      • 2019-10-01 27429, 2019

      • ssquare joined the channel
      • 2019-10-01 27434, 2019

      • BrainzGit
        [musicbrainz-server] reosarevok merged pull request #1206 (master…MBS-10353): MBS-10353: Enable ratings for release browse https://github.com/metabrainz/musicbrainz-server/…