hello ? anybody have experience installing and searching the musicbrainz database here ?
2016-04-28 11908, 2016
Gentlecat
probably :)
2016-04-28 11929, 2016
minteria
hi!
2016-04-28 11931, 2016
minteria
i installed the musicbrainz database but seem to be having an awful time running queries on it... i was just curious if anybody knows what an ideal hardware setup might be
2016-04-28 11944, 2016
minteria
if anybody has any opinions
2016-04-28 11954, 2016
darwin
you may not have indexes?
2016-04-28 11909, 2016
darwin
how'd you install etc?
2016-04-28 11948, 2016
CallerNo6
also, are you querying directly (SQL), or using the web service?
2016-04-28 11924, 2016
minteria
its in a mysql database now
2016-04-28 11930, 2016
minteria
so directly
2016-04-28 11938, 2016
minteria
someone installed it for me
2016-04-28 11902, 2016
minteria
how would i know if there are indexes ?
2016-04-28 11929, 2016
minteria
could the problem be because he migrated the database to mysql ? ( i know on the website it says its postresql )
2016-04-28 11905, 2016
bitmap
most likely, but that's not a supported method of installing the database, so you'd have to ask that person what they did
it seems like sentry stopped receiving messages because of SNI :(
2016-04-28 11904, 2016
zas
huh?
2016-04-28 11914, 2016
zas
on sakura ?
2016-04-28 11937, 2016
Gentlecat
we should coordinate a python upgrade (preferably to the latest release of both 2 and 3)
2016-04-28 11947, 2016
Gentlecat
I created a ticket (MBH-438), sorry for delay
2016-04-28 11921, 2016
zas
Python 2.7.6
2016-04-28 11929, 2016
Gentlecat
in general. I've been checking some of our python apps on sakura and they have issues with certificate verification
2016-04-28 11902, 2016
Gentlecat
I believe SNI support has been added in 2.9.7
2016-04-28 11905, 2016
zas
Python 2.7.9+ is needed
2016-04-28 11913, 2016
Gentlecat
ideally 2.7.11, if we end up just building it or using some other source
2016-04-28 11949, 2016
Gentlecat
right, I meant 2.7.9 :)
2016-04-28 11921, 2016
zas
i will upgrade it on sakura, though it will be through a ppa, since ubuntu still provides 2.7.6
2016-04-28 11939, 2016
zas
it shouldn't break anything, but ...
2016-04-28 11948, 2016
Gentlecat
I will probably have to recreate virtual environments
2016-04-28 11901, 2016
zas
Ok, done, 2.7.11 is installed
2016-04-28 11916, 2016
zas
You may need to restart services
2016-04-28 11908, 2016
Gentlecat goes to check
2016-04-28 11919, 2016
kanha has quit
2016-04-28 11939, 2016
kanha joined the channel
2016-04-28 11924, 2016
pingupingu joined the channel
2016-04-28 11924, 2016
pingupingu has quit
2016-04-28 11924, 2016
pingupingu joined the channel
2016-04-28 11936, 2016
Slurpee has quit
2016-04-28 11949, 2016
Gentlecat
ok, everything is updated and seems to be working
2016-04-28 11954, 2016
Gentlecat
thanks, zas!
2016-04-28 11934, 2016
zas
np ;)
2016-04-28 11916, 2016
mihaitish has quit
2016-04-28 11902, 2016
zas
The "Block More" campaign is quite successful, the load on our servers significantly decreased, even though some are trying to circumvent blocks.
2016-04-28 11947, 2016
zas
I wonder if we couldn't use something more dynamic, like rate limit using iptables IPs detected as abusing the web service. Basically using rate limiter stats but instead of denying at nginx level, rate limit the number of tcp connections from reported IPs using iptables rules dynamically added and removed. If one IP is added more than one time, increases
Anyways, back to what I was saying, I know that as of the next schema change the database will track both merged and deleted MBIDs. But here are the things that I'm 100% sure on:
2016-04-28 11946, 2016
UmkaDK
Is is going to be possible to distinguish "this existed but has been deleted" from "we don't recognise this", ie: Will the error codes be updated to return a 410 instead of 404 for deleted MBIDs?
2016-04-28 11939, 2016
UmkaDK
Is there going to be a way to identify a reason for the MBID deletion (ie: it was a duplicate, error, etc)?
UmkaDK: I believe that the plan is that deleted and never-existed MBIDs should return different error codes.
2016-04-28 11909, 2016
D4RK-PH0ENiX has quit
2016-04-28 11923, 2016
Freso
Not sure if that is part of the schema change patch or not though.
2016-04-28 11935, 2016
D4RK-PH0ENiX joined the channel
2016-04-28 11939, 2016
Freso
If it's not, at least we'll have the data to eventually make it so. :)
2016-04-28 11935, 2016
Freso
UmkaDK: You'll have to ask ruaok about the bribe. :) (And possibly bitmap and/or Gentlecat, as it'd likely be one of those that'll end up coding it.)
2016-04-28 11951, 2016
UmkaDK
I see! :)
2016-04-28 11919, 2016
Freso
As far "Is there going to be a way to identify a reason for the MBID deletion"... yes, eventually :D
2016-04-28 11927, 2016
Freso
There's a ticket for it ;)
2016-04-28 11949, 2016
UmkaDK
Well, hopefully they'll be about later today ...
2016-04-28 11928, 2016
UmkaDK
I think the multiple MBIDs per entity caught the attention of the higher ups here... :)
2016-04-28 11933, 2016
Freso
(504'ing on Jira.)
2016-04-28 11950, 2016
Freso
UmkaDK: Where is it you work at again?
2016-04-28 11918, 2016
Freso
(I mean, a monetary donation to MetaBrainz is always welcome... and may make us like you enough that we can put something you really want to see happen a bit ahead in the queue of stuff to do. :D)
2016-04-28 11902, 2016
UmkaDK
There is interest in these features, but I don't hold the purse... :(
2016-04-28 11947, 2016
UmkaDK
... if it helps I could bounce this back upstairs and let "our people" speak to "your people" to work something out. :)
2016-04-28 11907, 2016
Freso
UmkaDK: Lets wait for ruaok to come around and see what he has to say, but that doesn't sound like a terrible idea. :)
2016-04-28 11908, 2016
UmkaDK
Glad to hear it :) ... and I do like the "[off]" feature!
but the last sentence ends with: ".... we'd just need to expose it somehow."
2016-04-28 11909, 2016
UmkaDK
Pretty much all of our use of MB data is via the API, so ...
2016-04-28 11932, 2016
UmkaDK
but at least we know that the data is there! :)
2016-04-28 11927, 2016
Nyanko-sensei has quit
2016-04-28 11916, 2016
Freso
Oh, yeah, edits can't (and most likely won't) be available via the web service, so there's not going to be a way to see why a given entity was removed via that route.
2016-04-28 11919, 2016
D4RK-PH0ENiX joined the channel
2016-04-28 11950, 2016
adhawkins
Seeing some problems trying to create an artist when adding a release. The window to create the new artist is locking up either when it first opens, or when I try to submit it. Known issue?
2016-04-28 11929, 2016
adhawkins
Browser is chrome if it makes any difference
2016-04-28 11932, 2016
Gentlecat
just met my gsoc student... at the supermarket :D
2016-04-28 11938, 2016
CJ_
Gentlecat, Were they on sale?
2016-04-28 11908, 2016
ruaok
UmkaDK: sorry, running around d London today. Email me with what needs doing and the proposed bribe? :-p
2016-04-28 11915, 2016
UmkaDK
Sure thing! :) Could you PM me your email address?
2016-04-28 11954, 2016
adhawkins
Seeing the same thing when trying to create a new label...
2016-04-28 11904, 2016
Freso
Gentlecat: :D
2016-04-28 11912, 2016
Freso
Gentlecat: It's a BCN native?
2016-04-28 11918, 2016
Freso
Or an exchange student?
2016-04-28 11900, 2016
Gentlecat
no, he's from another country
2016-04-28 11920, 2016
ruaok
Thanks freso
2016-04-28 11940, 2016
Freso
ruaok: 👍
2016-04-28 11900, 2016
ruaok
I just met with two music producers. My mind has officially been blown about the realities if creating music.
2016-04-28 11920, 2016
ruaok
Sausage might be easier to watch being made. :-)
2016-04-28 11933, 2016
ruaok runs off to the next meeting
2016-04-28 11941, 2016
adhawkins
I'm now trying to enter the edit for the new release, and when I click 'Enter edit' above the row of buttons I get '[object Object]' in red text...
