#metabrainz

and then just catch them all (DatabaseException) and return an error

yepp

right. now, the stuff that Gentlecat is adding

see my other links above. we're being inconsistent in the two get methods when the data doesn't exist

because if job_id is not a UUID, postgres will start complaining

in the dataset get we raise an exception, in the job get we return None. We should do the same in both of them, so you should also change the get_job method to raise NoDataFoundException instead

Gentlecat: right. this is a common pattern. I wonder if we should have a decorator for it

for validating an argument?

can you do that? perhaps not

it might be a bit ugly because it's not going to be very obvious

and it will always have to be in the same position I think

anyway, at least some validation util

ah Okay

or not, if it's named

I ll add the NoDataFoundException

still a bit ugly. I think it's better to just have a function

let's leave it out of this PR for now, because it's bigger than just the get_job

I want to get this merged first

yeah, a method is probably fine

ruaok: OK. I'll take a look through it

thx

data flows nicely to Big Query. I'm collecting queries for use later.

also alastairp: Do you know some company related to music technology where I could apply for an internship for next summer?

many

do you have any preferred areas or countries?

umm Dj software, or something like the EchoNest Labs

regions I am fine with Europe or US

preferably US since I havent been there

:D

zas: what do I need to do to relay through gmail?

from which machine do you want to relay ?

and which domain?

whichever will host MeB.org when we deploy dockerized version

Project listenbrainz-server build #193: SUCCESS in 2 min 6 sec: https://ci.metabrainz.org/job/listenbrainz-serv...

Catgroove: do me a quick favor?

can you import into test.listenbrainz.org ?

Project listenbrainz-server build #194: SUCCESS in 1 min 40 sec: https://ci.metabrainz.org/job/listenbrainz-serv...

🍺

uh

well, it *is* after midday

ruaok: i cna do it now

thx

hmm how od i know if it's working?

is the dialog saying it is importing?

if so, open another tab and go to "your listens"

are they appearing?

the dialogue has just thel istenbrianz logo and no listens are apearing. gonan try with the othe firefox

works with the other firefox

k

it's gonna take a while though, it's on page 805 of 21279

yeah.

'ill be leaving at 15:40 some time before that..

actulaly I'll be leaving NOW since I haven't even figured out what to waer because weather is so unpredictable

i'll jsut leave it unning

running even

k, thx

still going?

quarter way about

I must go now so I can't reply to many more questions, sorry

no worries.

bitmap: ping

bitmap, ruaok, Gentlecat: i'm testing cAdvisor to monitor docker containers, reporting to influxdb, and grafana

http://dockertest1.brainz.io:8080/containers/ is the web output for cAdvisor

cAdvisor is collecting data for each container on the machine it is running on, it forwards metrics to our influxdb

zas: is it fine to now replace webstats and mrtg by grafana in the wiki doc?

yes, if links are broken

yes, done https://wiki.musicbrainz.org/index.php?title=Ho... (my first-ever wiki edit)

I am trying to create a listing for Festival Of Irish Music, Vol. 6 ( https://www.amazon.com/dp/B000006185 ). I found that this album was previously released under a different title and label number as "The Very Best Of Irish Music" ( https://musicbrainz.org/release/59d3cec4-34bd-4... ). Is it possible to get this album listed as a 'Release Duplicate' so that I don't need to manually import all the information

artist and songs that are already in the database? Thank you

CampSoup: #musicbrainz is better :) But you can click "Add Medium" on the tracklist page and search for the existing medium there

Ok, Ill try that

Thank you

I didn't think of that even though I had actually looked at that feature earlier...

Last question, do I have to submit the new listing before I could submit the disc ID for the second disc in the album?

Yeah :/

I think that's still how it works

Thats what I also thought. Thank you

zas: cool

bitmap: i created 2 scripts, one to install docker with docker-compose on 16.04, another to deploy cAdvisor container, see https://github.com/metabrainz/docker-helpers

i'm currently testing ways to clean old containers & images (because those leftover are using a lot of diskspace)

are we not using docker-machine stuff?

(I thought it installed docker for you)

dunno

docker run -ti -v /var/run/docker.sock:/var/run/docker.sock yelp/docker-custodian dcgc --max-image-age 1h --max-container-age 1h

works well

bitmap: i created a vm for docker tests

i'll add your key

I should run that locally...

great

ruaok: what are the limits for google cloud instances ? we need to create few vms for tests and wonder about that

how long do you want to run them for?

hours or weeks?

weeks for most, until we switch to NewHost

basically we'll need them to play with docker swarm and our containers and see if everything plays well together

how many vms?

and what size?

check current list, we may need one more, dockertest1 & dockertest2

guesses ok. :)