bitmap: hey, sorry, I don't remember the exact reason (should've put it in the commit), but I have a feeling the AB database wasn't showing up with that check, for some reason or the other.
Mr_Monkey
Moin
jmp_music joined the channel
prabal
Morningg
jmp_music
Morning!
Gazooo has quit
ZaphodBeeblebrox joined the channel
ZaphodBeeblebrox has quit
ZaphodBeeblebrox joined the channel
Gazooo joined the channel
CatQuest has quit
shivam-kapila
Morning
yvanzo
mo’’in’
diru1100
Mo''in' !
ruaok
'ooooin!
v6lur_ joined the channel
Mr_Monkey
ruaok: I just upgraded my local PG version, did the update kerfuffle, ran the tests and checked manually. No snag, everything working fine. I'mm adding a couple of steps to the migration document regarding BB
ruaok
very good. thanks for getting on that so quick. :)
do you have postgres functions in BB? one of the changes that affects us are functions now need to be fully qualified with schema name. that is the only thing that I see as a real snag.
Mr_Monkey
One question I have is regarding the DB update script that's going to be run. I have no visibility on it (no access to metabrainz/docker-postgres-cluster ?) so I just want to make sure there's nothing else I need to do to point to BB's database
I wouldn't even know what a function looks like, so I guess that's a no. i'll check.
ruaok
where are your DB creation scripts? gimme a link and I'll help.
regarding the update process -- from your perspective, the DB will go down and eventually come back running 12.3.
if no changes are necessary perhaps you need to do nothing at all.
Mr_Monkey
Right. Nothing to add to say "don't forgot database 'bookbrain' then. Perfect :)
and you've run through database creation and tests with PG12?
if all that passes, then I think you're good to go.
Mr_Monkey
Haven't created a DB from scratch yet on 12; that's my next step. Tests all pass and I've ensured the triggers do work. Not sure if it's any different if they're recreated from scratch
Gonna check all that now.
ruaok
the migration is scheduled for 5ish (CET) today. you should be at the ready to address issues as they come up. we're all pretty much going to be standing by and ready to jump if something goes wrong.
Mr_Monkey
Yep. I tweeted about planned downtime earlier.
ruaok
good thinking.
alastairp
this shouldn't affect AB, right?
do you need any help with LB/MsB/CB?
ruaok
I dont anticipate it affecting AB, since AB has its own DB. But its a very inter connected system, so hard to tell.
I think we're set on the other projects.
but you being around during release time would be good.
alastairp
I'll be here
ruaok
sweet
iliekcomputers
No meeting today then, right?
ruaok
depends. if all goes well, we might have one.
iliekcomputers
ishaanshah[m]: we should probably cancel or reschedule our meeting today. The channel will be busy.
> we should probably cancel or reschedule our meeting today. The channel will be busy.
iliekcomputers: Sure, does 7pm IST good for you?
yvanzo
metabrainz/base-image:latest is the same as metabrainz/base-image:v0.9.22-1 by a few log files (using container-diff), is there something to know about other v0.9.22-X tags?
TOPIC: MetaBrainz Community and Development channel | MusicBrainz non-development: #musicbrainz | GSoC students start here: https://bit.ly/2NUaxgb | Channel is logged; see https://musicbrainz.org/doc/IRC for details | Meeting agenda (today is "schema chance" day, might not be a meeting—stay tuned!): Reviews
TOPIC: MetaBrainz Community and Development channel | MusicBrainz non-development: #musicbrainz | GSoC students start here: https://bit.ly/2NUaxgb | Channel is logged; see https://musicbrainz.org/doc/IRC for details | Meeting agenda (today is "schema change" day, might not be a meeting—stay tuned!): Reviews
cloverfield has quit
cloverfield joined the channel
bitmap: the CAA needs to be taken down as well, right? I'm reviewing the migration doc...
bitmap
yeah, though I think we can leave the actual containers running - it'll come down at the gateway and the PG consul service will be down
ruaok
yeah, I'm looking at LB right now and we have a pile of containers and I think it might just be best to take them all down.
how long do you think the downtime will be?
iliekcomputers
ruaok: I would not take cron down
It's doing a full dump.
Happy to merge #862
bitmap
I'm hoping for less than 1.5 hours... depends on how long vacuuming & backups take
ruaok
are you sure that isn't going to hold any connections to PG, iliekcomputers ?
bitmap
the actual pg upgrade is pretty fast. MB has to recreate some indexes which takes a bit of time too
MajorLurker has quit
though I was testing this all on williams and it should be much faster on floyd, so
iliekcomputers
Can't say for certain. :/
Take it down, I guess. I'll trigger a full dump again today.
ruaok
let's try and keep it up. if a connection remains, let's take it down. I'll note that in the doc.