Ok, that seems to be all, since atj and zas have not appeared
2023-12-11 34543, 2023
reosarevok
So, thanks everyone for attending, and let's talk further next week!
2023-12-11 34501, 2023
reosarevok
(</OLUFSEN> and) </BANG>
2023-12-11 34507, 2023
ApeKattQuest
lmao
2023-12-11 34511, 2023
TOPIC: MetaBrainz Community and Development channel | MusicBrainz non-development: #musicbrainz | BookBrainz: #bookbrainz | Channel is logged; see https://musicbrainz.org/doc/IRC for details | Agenda: Reviews
2023-12-11 34516, 2023
monkey
Thanks!
2023-12-11 34545, 2023
aerozol
I waited until after the meeting for this - WTF, another gif tool from a MeB contributor?
2023-12-11 34501, 2023
reosarevok
Wait, there's more?
2023-12-11 34510, 2023
aerozol
Is Peek for linux outsidecontext? I use Licecap on Win which I think was from a MB editor as well
yvanzo: if you think that PR is now ok then we can put a beta out :)
2023-12-11 34537, 2023
reosarevok
(not the end of the world if you want to take a better look in the morning, mind!)
2023-12-11 34543, 2023
outsidecontext
aerozol: yes, for Linux. Licecap was one of its inspirations
2023-12-11 34528, 2023
bitmap
lucifer: hi, let me check
2023-12-11 34518, 2023
lucifer
mayhem: about choosing the release (hence recordings) for a release group, canonical releases do not take cover art into consideration so there may be cases where the cover art and the tracklist are from different releases.
2023-12-11 34505, 2023
bitmap
lucifer: iiuc it would have to be enabled on the primary and requires a server restart, but otherwise yes
2023-12-11 34510, 2023
bitmap
how many tables do you need?
2023-12-11 34534, 2023
lucifer
bitmap: ~15.
2023-12-11 34535, 2023
mayhem
lucifer: indeed. that is a tricky case to solve.
2023-12-11 34549, 2023
mayhem
but i think that isn't a dealbreaker.
2023-12-11 34503, 2023
aerozol
MeB contributors make everything that is good, it turns out
2023-12-11 34543, 2023
lucifer
bitmap: oh i see, i thought it would be possible to do it on standby which could possibly be restarted without a meb-wide downtime.
2023-12-11 34549, 2023
lucifer
mayhem: makes sense
2023-12-11 34521, 2023
bitmap
I think not, at least in version 12, but I'm not very familiar with the feature
2023-12-11 34521, 2023
lucifer
ah right, prod is using PG 12.
2023-12-11 34543, 2023
lucifer
makes sense, can we schedule a downtime for this?
2023-12-11 34506, 2023
bitmap
sure, let's coordinate with zas & atj when they are back
2023-12-11 34518, 2023
lucifer
sounds good, thanks!
2023-12-11 34541, 2023
bitmap
if all of these tables are (MB)replicated, you could also sync the changes from the dbmirror2 schema, but it won't be as seamless to setup as logical replication
2023-12-11 34506, 2023
lucifer
i think we need most of the bigger tables, track and recording. everything else is a magnitude smaller so doesn't really matter whether we copy it or not.
2023-12-11 34522, 2023
lucifer
only edit, edit_note, edit_data is larger and we do not need those tables.
2023-12-11 34509, 2023
lucifer
bitmap: do you suggest that we write a script in LB to download replication packets and apply those to our copy of data?
2023-12-11 34505, 2023
lucifer
i am not sure the existing scripts as in MB would work in creating a schema replica inside another database?
2023-12-11 34542, 2023
bitmap
it is possible but would probably require more effort and depends on how up-to-date you need the data to be
2023-12-11 34500, 2023
lucifer
i guess the data would be out of date for an hour or so which should be fine for now.
I guess if you want to go this route and need faster updates, you don't necessarily have to download the hourly packets - we can mirror the dbmirror2 schema somewhere and you can read directly from that
2023-12-11 34521, 2023
bitmap
logical replication will likely be easier to set up
2023-12-11 34545, 2023
lucifer
indeed indeed.
2023-12-11 34555, 2023
bitmap
my only concern would be that the docs mention it increases WAL volume but I'm not sure how significant that is
that's weird, it looks like it's trying to open "/dev/mapper/data-root" as the luks device, but that's the device for the contents of the luks device, it should be trying to unlock the physical partition device.
2023-12-11 34522, 2023
kepstin
either that or i'm misreading the error.
2023-12-11 34508, 2023
kepstin
hmm, this is luks on top of lvm?
2023-12-11 34522, 2023
mayhem
I think so.
2023-12-11 34540, 2023
kepstin
this might be the wrong uuid in the crypttab? crypttab needs the uuid of the luks device, while fstab needs the uuid of the filesystem inside the luks device.
