I'm guessing the double spikes are time-separated index rotations.
2016-11-30 33558, 2016
ruaok
but why do the number of ops spike during that time? I would expect time to spike, not ops
2016-11-30 33538, 2016
zas
nope, reload
2016-11-30 33507, 2016
zas
number of requests ending in 408 per minute per upstream
2016-11-30 33524, 2016
zas
i can sum up both upstreams if you prefer
2016-11-30 33552, 2016
ruaok
ah, got it.
2016-11-30 33546, 2016
drsaunder has quit
2016-11-30 33556, 2016
lazka joined the channel
2016-11-30 33557, 2016
drsaunder joined the channel
2016-11-30 33541, 2016
ruaok
ok, what is going on then?
2016-11-30 33549, 2016
dboys_ joined the channel
2016-11-30 33558, 2016
ruaok
there seems to be a period of rising 408s before the index rotation.
2016-11-30 33529, 2016
ruaok
what causes that? IO contention during index copy?
2016-11-30 33533, 2016
ruaok
vodoo?
2016-11-30 33513, 2016
dpmittal has quit
2016-11-30 33528, 2016
CatQuest
probably voodoo
2016-11-30 33559, 2016
zas
ruaok: reload, i added rx bytes/s on each search server to the dashboard
2016-11-30 33548, 2016
Freso
Oh, crud. I forgot that Jira isn't sending out e-mails.
2016-11-30 33526, 2016
Freso wonders how many Jira accounts he's made that never got their password e-mails :|
2016-11-30 33504, 2016
zas
ruaok: i added cpu % (moving avg)
2016-11-30 33516, 2016
legoktm_ is now known as legoktm
2016-11-30 33532, 2016
ruaok
thx!
2016-11-30 33553, 2016
ruaok
at 18:05, cpu on prince and boingo is really out of whack.
2016-11-30 33513, 2016
ruaok
the CPU tends to fluctuate quite a lot between successive top screens.
2016-11-30 33526, 2016
ruaok
much more so than at oldhost.
2016-11-30 33555, 2016
ruaok
I'll look again tomorrow with more data. right now that is just wtf.
2016-11-30 33509, 2016
Freso
ruaok: What is the latest I should expect a package to arrive at?
2016-11-30 33517, 2016
Freso is starting to feel rather chill...
2016-11-30 33530, 2016
ruaok
next thursday?
2016-11-30 33548, 2016
ruaok
I usually gave up by 7pm.
2016-11-30 33558, 2016
ruaok
which mean that they would show up at 9am the next day.
2016-11-30 33506, 2016
ruaok
hence me shipping things to drop of locations now.
2016-11-30 33500, 2016
Freso
Mhm.
2016-11-30 33510, 2016
Freso
MRW has a place nicely close to the office.
2016-11-30 33520, 2016
Freso picked up something there on his way here today
2016-11-30 33556, 2016
CatQuest has quit
2016-11-30 33552, 2016
Freso
...
2016-11-30 33503, 2016
Freso
I think I just locked myself out of my Jira account.
2016-11-30 33517, 2016
ruaok
👏
2016-11-30 33527, 2016
Freso
yep.
2016-11-30 33529, 2016
Freso
🙌
2016-11-30 33528, 2016
Freso
Anyone have any idea if the Jira not-sending-email issue is going to be solved soon? If not, can someone with admin access (that isn't me…) set my password to something and let me know?
2016-11-30 33503, 2016
ruaok has no idea
2016-11-30 33517, 2016
ruaok
still too many other things screaming. stupid business shit mostly.
2016-11-30 33546, 2016
ruaok partly hopes that a jira volunteer will step up
2016-11-30 33556, 2016
ruaok
check your pm's freso
2016-11-30 33534, 2016
anshuman73
How does one install Picard once one has added the PPA for ubuntu ?
2016-11-30 33532, 2016
anshuman73 realizes it was as simple as `sudo apt-get install picard`
2016-11-30 33537, 2016
anshuman73
I guess that could be added to the launchpad link, just in case...
2016-11-30 33555, 2016
Freso
anshuman73: If you're developing on it, you might want to just run it from the git checkout. :)
2016-11-30 33526, 2016
anshuman73
Ahh, not yet developing on it, but yes, that surely is a better idea considering I may need it
2016-11-30 33511, 2016
anshuman73
Freso, is there a way to add the official websites of artists in their relationships ?
2016-11-30 33507, 2016
Freso
anshuman73: Yes. :) Just add the official website to the list of URLs and mark it as "official homepage".
2016-11-30 33536, 2016
anshuman73
ooh ! Alright !
2016-11-30 33511, 2016
anshuman73
My First edit on Musicbrainz ! \o/
2016-11-30 33534, 2016
Freso
🙌
2016-11-30 33521, 2016
MajorLurker has quit
2016-11-30 33547, 2016
MajorLurker joined the channel
2016-11-30 33537, 2016
CallerNo6
oh, hey, anshuman73 , I asked about release types. It sounds like you can ignore 'type' and look only at 'primary' and 'secondary'. The WS 'type' is derived from primary/secondary.
2016-11-30 33501, 2016
anshuman73
Hey CallerNo6, Thanks a lot. This sure solves confusions !
2016-11-30 33513, 2016
anshuman73
though, what does WS stand for ?
2016-11-30 33522, 2016
CallerNo6
web service
2016-11-30 33539, 2016
anshuman73
oh
2016-11-30 33553, 2016
anshuman73
sampsyo: ping !
2016-11-30 33551, 2016
glennren joined the channel
2016-11-30 33559, 2016
CallerNo6
I say that to distinguish it from the 'type' column in the 'release_group' table, which is not the same thing.
The Artist Mellisa Hayes doesn't exist on Musicbrainz it seems
2016-11-30 33522, 2016
Freso
Then you should add her.
2016-11-30 33539, 2016
CallerNo6
That's correct. The 'artist' field(s) aren't just text. they need to be matched to an artist in the db. when you've selected a match, the field will turn green.
It turns out, that is not the same as the 'type' returned by the web service.
2016-11-30 33545, 2016
CallerNo6
(so, that was part of /my/ confusion)
2016-11-30 33533, 2016
anshuman73
Yeah, according to that code, type should always end up being associated with the primary type, but doesn't end up being that way
2016-11-30 33543, 2016
CallerNo6
The important thing for your purpose is to know that the 'type' returned by the WS is a legacy field calculated f rom the newer primary/secondary values.
2016-11-30 33511, 2016
anshuman73
yes !
2016-11-30 33545, 2016
glennren has quit
2016-11-30 33551, 2016
CallerNo6
(so you can ignore it)
2016-11-30 33530, 2016
reosarevok
Freso: do you still need that jira stuff?
2016-11-30 33541, 2016
Freso
reosarevok: No, ruaok sorted me out. Thanks though.
2016-11-30 33544, 2016
reosarevok
Ok :)
2016-11-30 33523, 2016
anshuman73
CallerNo6: I think that's sampsyo's call ;) As of now we've put the new types obtained from the primarytype field and the secondarytype fields into the logs for later use. Bu this clarification should definitely help us get moving.
2016-11-30 33549, 2016
Lord_of_Codes has quit
2016-11-30 33523, 2016
sampsyo
Yes, this helps! Thanks to both of you, anshuman73 and CallerNo6.
2016-11-30 33538, 2016
Lord_of_Codes joined the channel
2016-11-30 33540, 2016
anshuman73
CallerNo6: P.s. I came across the issue while working on beets, if you're wondering where sampsyo get's involved
2016-11-30 33547, 2016
sampsyo
I think we'll end up probably keeping all three "type" values—one for backwards compatibility, and then both the primary and secondary as well.
2016-11-30 33505, 2016
anshuman73
sampsyo: great ! Does that mean I revert again in the PR ?
2016-11-30 33527, 2016
sampsyo
Nope, let's keep it as is.
2016-11-30 33530, 2016
ruaok
chirlu: see the footer of this page.
2016-11-30 33502, 2016
sampsyo
The logging approach is still the right first milestone—we'll save the values in the next phase, once some more infrastructure is in place.
2016-11-30 33505, 2016
anshuman73
sampsyo: sure ! Can I continue working on this and take up the next task related to it as well ?
2016-11-30 33530, 2016
sampsyo
Yeah; that would be awesome!!
2016-11-30 33543, 2016
sampsyo
Do you see the task I'm referring to, or can I help by looking it up for you?
