#metabrainz

/

      • bitmap
        please do

        • 2015-10-26 29906, 2015

      • bitmap
        welp, dunno what's up MBJenkins, usually he comes back by now

        • 2015-10-26 29919, 2015

      • Gentlecat
        hmm https://gist.github.com/gentlecat/9fb2633daa801c3…

        • 2015-10-26 29950, 2015

      • Gentlecat
        I don't think I touched any of these submodules

        • 2015-10-26 29902, 2015

      • bitmap
        huh...well rm -r and re-fetch the submodules is usually the fastest solution

        • 2015-10-26 29916, 2015

      • Gentlecat
        git submodule foreach --recursive git checkout .

        • 2015-10-26 29931, 2015

      • Gentlecat
        this works too it seems

        • 2015-10-26 29902, 2015

      • bitmap
        looks smarter

        • 2015-10-26 29918, 2015

      • bitmap
        > There is no field-applied_edits

        • 2015-10-26 29926, 2015

      • bitmap
        oops :)

        • 2015-10-26 29900, 2015

      • Gentlecat
        wait what. why did that stop working?

        • 2015-10-26 29914, 2015

      • Gentlecat
        oh, I forgot to add something

        • 2015-10-26 29914, 2015

      • LordSputnik joined the channel

        • 2015-10-26 29908, 2015

      • LordSputnik
        bitmap: do you have time to run "sudo make install" for the GCC 4.9 build on rika? It's in "/home/bookbrainz/gcc_builddir"

        • 2015-10-26 29924, 2015

      • bitmap
        sure

        • 2015-10-26 29933, 2015

      • LordSputnik
        (although might have to uninstall gcc 4.4 first)

        • 2015-10-26 29951, 2015

      • LordSputnik
        great, thank you

        • 2015-10-26 29900, 2015

      • bitmap
        did you configure it to install to /usr/local?

        • 2015-10-26 29919, 2015

      • LordSputnik
        I didn't change the config, let me check

        • 2015-10-26 29941, 2015

      • LordSputnik
        Yes, the prefix is /usr/local

        • 2015-10-26 29950, 2015

      • bitmap
        ok

        • 2015-10-26 29915, 2015

      • bitmap
        it's done

        • 2015-10-26 29916, 2015

      • LordSputnik
        sampsyo: hey :) Just wondering if you know if anyone is working on fixing the lastimport plugin?

        • 2015-10-26 29920, 2015

      • LordSputnik
        bitmap: thanks!

        • 2015-10-26 29944, 2015

      • sampsyo
        LordSputnik: alas no -- as far as I can tell, the last.fm API is broken

        • 2015-10-26 29949, 2015

      • sampsyo
        and there may not be anything we can do about it

        • 2015-10-26 29912, 2015

      • LordSputnik
        sampsyo: OK, I'm going to see if I can come up with something

        • 2015-10-26 29926, 2015

      • sampsyo
        awesome. thanks for looking into it!

        • 2015-10-26 29939, 2015

      • LordSputnik
        I'll probably also integrate ListenBrainz, if there's support for querying user play counts by MBID

        • 2015-10-26 29908, 2015

      • sampsyo
        ListenBrainz integration! living on the cutting edge

        • 2015-10-26 29921, 2015

      • sampsyo
        sounds awesome

        • 2015-10-26 29937, 2015

      • JonnyJD has quit

        • 2015-10-26 29936, 2015

      • LordSputnik
        sampsyo: Oh nope, Last.fm no longer accepting API registration. Guess it'll be a ListenBrainz-only plugin

        • 2015-10-26 29954, 2015

      • Gentlecat
        bitmap: https://beta.musicbrainz.org/search/edits?auto_ed… \o/

        • 2015-10-26 29931, 2015

      • sampsyo
        :(

        • 2015-10-26 29939, 2015

      • Gentlecat
        hmm, fails when it's "with less than 1"

        • 2015-10-26 29904, 2015

      • bitmap
        Gentlecat: nice :D

        • 2015-10-26 29916, 2015

      • LordSputnik
        Gentlecat: do you happen to know if the ListenBrainz docs are published anywhere yet?

        • 2015-10-26 29917, 2015

      • Gentlecat
        "is 0" is a bit long as well

        • 2015-10-26 29946, 2015

      • LordSputnik
        Oh, I've found http://listenbrainz.readthedocs.org/en/latest/

        • 2015-10-26 29901, 2015

      • Gentlecat
        listenbrainz.rtfd.org, but "IT'S NOT OFFICIAL YET!!!"

        • 2015-10-26 29955, 2015

      • bitmap
        Gentlecat: well, edit search is going to be slow in general

        • 2015-10-26 29915, 2015

      • bitmap
        since it touches the edit table...

        • 2015-10-26 29939, 2015

      • Gentlecat
        I used it a couple of times before, so I don't know how well it works usually

        • 2015-10-26 29947, 2015

      • LordSputnik
        will the MSID match the MBID if the original track was tagged with an MBID?

        • 2015-10-26 29903, 2015

      • Gentlecat
        I don't think there's any matching at this point

        • 2015-10-26 29944, 2015

      • Nyanko-sensei has quit

        • 2015-10-26 29949, 2015

      • github joined the channel

        • 2015-10-26 29949, 2015

      • github
        [listenbrainz-server] LordSputnik opened pull request #66: Fix query parameter name in api.py documentation (master...patch-1) https://github.com/metabrainz/listenbrainz-server…

        • 2015-10-26 29949, 2015

      • github has left the channel

        • 2015-10-26 29917, 2015

      • LordSputnik
        so it's not possible to get the recording MBID of a listen?

        • 2015-10-26 29918, 2015

      • Gentlecat
        https://listenbrainz.rtfd.org/en/latest/dev/json.…

        • 2015-10-26 29916, 2015

      • LordSputnik
        sampsyo: hmm, guess we can't do it with ListenBrainz either yet :/

        • 2015-10-26 29950, 2015

      • sampsyo
        :(

        • 2015-10-26 29951, 2015

      • Nyanko-sensei joined the channel

        • 2015-10-26 29957, 2015

      • Nyanko-sensei has quit

        • 2015-10-26 29924, 2015

      • Nyanko-sensei joined the channel

        • 2015-10-26 29944, 2015

      • ruaok has quit

        • 2015-10-26 29943, 2015

      • LordSputnik has quit

        • 2015-10-26 29955, 2015

      • Freso has quit

        • 2015-10-26 29913, 2015

      • Freso joined the channel

        • 2015-10-26 29941, 2015

      • Freso has quit

        • 2015-10-26 29904, 2015

      • Freso joined the channel

        • 2015-10-26 29918, 2015

      • JesseW joined the channel

        • 2015-10-26 29913, 2015

      • chirlu`
        No fish.

        • 2015-10-26 29919, 2015

      • ariscop joined the channel

        • 2015-10-26 29914, 2015

      • JesseW has quit

        • 2015-10-26 29953, 2015

      • JesseW joined the channel

        • 2015-10-26 29926, 2015

      • diana_olhovik joined the channel

        • 2015-10-26 29931, 2015

      • diana_olhovik has quit

        • 2015-10-26 29949, 2015

      • Gentlecat has quit

        • 2015-10-26 29959, 2015

      • JesseW has quit

        • 2015-10-26 29933, 2015

      • UmkaDK joined the channel

        • 2015-10-26 29902, 2015

      • ariscop has quit

        • 2015-10-26 29901, 2015

      • UmkaDK has quit

        • 2015-10-26 29941, 2015

      • ruaok joined the channel

        • 2015-10-26 29911, 2015

      • yeeeargh joined the channel

        • 2015-10-26 29910, 2015

      • emitattuo has quit

        • 2015-10-26 29923, 2015

      • emitattuo joined the channel

        • 2015-10-26 29925, 2015

      • JonnyJD joined the channel

        • 2015-10-26 29900, 2015

      • ariscop joined the channel

        • 2015-10-26 29905, 2015

      • ariscop has quit

        • 2015-10-26 29926, 2015

      • ariscop joined the channel

        • 2015-10-26 29951, 2015

      • samphippen joined the channel

        • 2015-10-26 29915, 2015

      • ariscop has quit

        • 2015-10-26 29918, 2015

      • ariscop joined the channel

        • 2015-10-26 29920, 2015

      • ariscop has quit

        • 2015-10-26 29915, 2015

      • ariscop joined the channel

        • 2015-10-26 29945, 2015

      • UmkaDK joined the channel

        • 2015-10-26 29948, 2015

      • Gentlecat joined the channel

        • 2015-10-26 29939, 2015

      • zas
        ruaok: search servers look happy now, i propose to let them run as is until next issue

        • 2015-10-26 29949, 2015

      • zas
        btw, they eat memory far too fast imho, but this is something only devs can solve, who was in charge of this code ?

        • 2015-10-26 29902, 2015

      • ruaok_ joined the channel

        • 2015-10-26 29941, 2015

      • github joined the channel

        • 2015-10-26 29941, 2015

      • github
        [listenbrainz-server] mayhem pushed 2 new commits to master: https://github.com/metabrainz/listenbrainz-server…

        • 2015-10-26 29941, 2015

      • github
        listenbrainz-server/master 7d7b8b6 Ben Ockmore: Fix query parameter name in api.py documentation...

        • 2015-10-26 29941, 2015

      • github
        listenbrainz-server/master ec5a21c Robert Kaye: Merge pull request #66 from LordSputnik/patch-1...

        • 2015-10-26 29941, 2015

      • github has left the channel

        • 2015-10-26 29926, 2015

      • ruaok__ joined the channel

        • 2015-10-26 29949, 2015

      • alastairp
        hmm, weird

        • 2015-10-26 29942, 2015

      • UmkaDK has quit

        • 2015-10-26 29928, 2015

      • UmkaDK joined the channel

        • 2015-10-26 29949, 2015

      • UmkaDK_ joined the channel

        • 2015-10-26 29946, 2015

      • UmkaDK has quit

        • 2015-10-26 29908, 2015

      • ruaok has quit

        • 2015-10-26 29939, 2015

      • samphippen has quit

        • 2015-10-26 29951, 2015

      • ariscop has quit

        • 2015-10-26 29950, 2015

      • emitattuo has quit

        • 2015-10-26 29946, 2015

      • ariscop joined the channel

        • 2015-10-26 29951, 2015

      • ruaok__ has quit

        • 2015-10-26 29920, 2015

      • ruaok_
        zas: strong work on beating the search servers into submission.

        • 2015-10-26 29932, 2015

      • zas
        ;)

        • 2015-10-26 29935, 2015

      • ruaok_
        I still don't understand how the monit restarts are not causing problems now.

        • 2015-10-26 29950, 2015

      • zas
        because they don't restart at the same time

        • 2015-10-26 29913, 2015

      • zas
        to my understanding, this is the main cause: one server down, load balancer puts all load on the other, the other drop, no server, first one starts take all load during startup, can't handle it, second server starts, load balancer sends all load to it, can't handle it, etc....

        • 2015-10-26 29908, 2015

      • zas
        nginx recent versions has a slowstart feature, that could help in this imho (not sure which version has it though)

        • 2015-10-26 29926, 2015

      • ruaok_
        that would be perfect!!

        • 2015-10-26 29929, 2015

      • zas
        btw, it can still happen, if we are unlucky

        • 2015-10-26 29929, 2015

      • ruaok_
        (slow start)

        • 2015-10-26 29938, 2015

      • reosarevok joined the channel

        • 2015-10-26 29957, 2015

      • russss
        this is elasticsearch?

        • 2015-10-26 29909, 2015

      • ruaok_
        heavens no.

        • 2015-10-26 29925, 2015

      • ruaok_
        it is our own system on top of lucene, hosted in tomcat.

        • 2015-10-26 29933, 2015

      • russss
        I wrote one of those once.

        • 2015-10-26 29953, 2015

      • russss
        I only say that because ES has a way of pre-warming indexes before it comes online https://www.elastic.co/guide/en/elasticsearch/ref…

        • 2015-10-26 29924, 2015

      • russss
        if you like json.

        • 2015-10-26 29912, 2015

      • russss
        but given that it's actually working, I suspect that your search load doesn't really need caches warming

        • 2015-10-26 29907, 2015

      • zas
        well, the core issue is that we need to restart those damn search servers far too often (2 days ago it was every 10-15 minutes), mostly because they eat memory as hell, and because of indexes pushs (about every hour), i reduced the number of restarts, which reduced the "chance" for them to go mad, but it is more a workaround that a fix.

        • 2015-10-26 29914, 2015

      • ariscop has quit

        • 2015-10-26 29908, 2015

      • zas
        ruaok_ said someone was working on a replacement, but i don't know anything about it

        • 2015-10-26 29922, 2015

      • zas
        nginx plus is the non-free version ? according to docs, the slow start is an nginx plus feature

        • 2015-10-26 29937, 2015

      • ariscop joined the channel

        • 2015-10-26 29959, 2015

      • ariscop has quit

        • 2015-10-26 29945, 2015

      • samphippen joined the channel

        • 2015-10-26 29928, 2015

      • ariscop joined the channel

        • 2015-10-26 29956, 2015

      • Freso
        LordSputnik: MSID != MBID, and I guess an MSID→MBID matching can change over time.

        • 2015-10-26 29904, 2015

      • Leftmost
        MSID?

        • 2015-10-26 29945, 2015

      • russss
        messybrainz ID

        • 2015-10-26 29910, 2015

      • ruaok_ has quit

        • 2015-10-26 29949, 2015

      • ruaok joined the channel

        • 2015-10-26 29919, 2015

      • Nyanko-sensei has quit

        • 2015-10-26 29908, 2015

      • ruaok
        zas: I'm going through and adjusting nagios warnings for loads and total procs as they arrive.

        • 2015-10-26 29917, 2015

      • ruaok
        trying to quiet things down a little.

        • 2015-10-26 29924, 2015

      • alastairp
        Freso: we hope they won’t change

        • 2015-10-26 29933, 2015

      • alastairp
        well, they may if it’s wrong

        • 2015-10-26 29952, 2015

      • alastairp
        but, e.g. I don’t think we should change if 2 other mbids are merged

        • 2015-10-26 29935, 2015

      • Freso
        alastairp: I was thinking that initial mapping might prove wrong.

        • 2015-10-26 29941, 2015

      • alastairp
        right

        • 2015-10-26 29959, 2015

      • alastairp
        my hope is that we get it right the first time - that is, don’t make a mapping unless you’re sure

        • 2015-10-26 29907, 2015

      • alastairp
        but you’re right, it’s a potential problem