16:01:08 <hiro> #startmeeting network-health 2024-09-30
16:01:08 <MeetBot> Meeting started Mon Sep 30 16:01:08 2024 UTC.  The chair is hiro. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:08 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:01:14 <morganava> <3
16:01:17 <juga> o/
16:01:21 <GeKo> hello!
16:01:23 <hiro> we have a new pad
16:01:31 <hiro> https://pad.riseup.net/p/tor-nethealthteam-2024-keep
16:03:52 <hiro> hello everyone
16:04:33 <hiro> Sarthik is running a little late... so this time is almost midnight for him, so I wanted to propose everyone if it was possible to move the meeting earlier
16:04:48 <hiro> like 2pm UTC or even 12pm
16:05:03 <GeKo> both work for me
16:05:52 <hiro> @juga? what do you think?
16:08:13 <hiro> I think we can try 12pm and if that is too early for ggus we can discuss when he is back
16:08:21 <GeKo> +1
16:08:24 <juga> wfm
16:08:34 <juga> both
16:08:37 <hiro> ok will update the  pad and wiki
16:09:20 <hiro> I do not have any bold items for this week... Just trying to solve performances on metricsdb-01: https://gitlab.torproject.org/tpo/tpa/team/-/issues/41779
16:10:40 <sarthikg> hey!
16:11:05 <GeKo> woot!
16:11:12 <GeKo> welcome!
16:11:22 <juga> hi sarthikg !
16:11:27 <GeKo> hiro: i guess giving the box more ram etc. can't hurt
16:11:40 <GeKo> even if you deal with thatr particular view
16:11:44 <GeKo> *that
16:11:55 <hiro> yep we are getting more ram and CPU @GeKo
16:12:05 <juga> i also agree with that
16:12:22 <hiro> https://gitlab.torproject.org/tpo/tpa/team/-/issues/41779#note_3085423
16:12:32 <hiro> anarcat is already doing that
16:12:38 <hiro> welcome @sarthikg !
16:13:07 <juga> \o/ re. #41779
16:13:09 <sarthikg> hey folks, sorry got a little late.
16:13:22 <hiro> but that view update needs 24minutes... so we either need to change the db schema or optimize the view or both
16:14:06 <hiro> no worries @sarthikg !!
16:14:13 <juga> in my experience
16:14:18 <juga> the tame a query takes
16:14:34 <juga> depends on how busy the cpu is with descriptor parser
16:14:47 <juga> that query has took only some mins some times
16:15:02 <hiro> descriptor parser runs that update at t he end of the cycle
16:15:07 <juga> though i think i haven't tried yet with latest changes
16:15:26 <hiro> loads all the data, then refreshes the materialized view
16:15:49 <juga> ah, so it was actually running without anything else running too
16:15:51 <juga> ?
16:15:58 <hiro> yep
16:16:02 <juga> ic
16:16:30 <anarcat> this afternoon, hopefully
16:16:37 <hiro> the parser adds the data to the db and then refreshes the view one by one
16:16:59 <hiro> s/view/views
16:17:00 <anarcat> wish i could join you now, but now is lunch here, ciao ciao! :)
16:17:08 <hiro> have a good luch!
16:17:09 <GeKo> no worries!
16:17:12 <GeKo> yeah
16:17:14 <hiro> *lunch
16:18:03 <juga> hiro: don't see an easy way to do this with descriptorParser, but it could be updated only once a day, not hour
16:18:29 <juga> maybe it's even worth to create a table, cause materialized views need to replace all the data again
16:18:36 <juga> instead of just adding the last data
16:18:44 <hiro> let's see maybe  we can structure the data in a way that is easier to query
16:18:45 <hiro> yep
16:18:51 <juga> again = when they're refereshed
16:18:59 <juga> ok
16:19:40 <hiro> @sarthikg how is your second week going for you?
16:20:38 <sarthikg> Going good. I've got a basic understanding of tagtor, got the local setup working, and have started working on raising MR's for the open issues.
16:20:50 <hiro> \o/
16:20:56 <GeKo> yay
16:21:35 <sarthikg> This week, I'm planning to further close issues, and start a gitlab issue for discussing the basic architecture for metrics2.0 repository.
16:22:06 <hiro> awesome! thanks!
16:22:19 <hiro> let us known if you need anything
16:22:43 <sarthikg> sure sure, will do!
16:23:53 <hiro> ok @GeKo do you need anything from me or anyone else for this week?
16:24:14 <GeKo> nope
16:24:36 <hiro> nice! if it is all good  then we can end the meeting
16:24:57 <juga> tomorrow p112 and Thursday p141, right?
16:25:25 <GeKo> yeah
16:25:31 <juga> kk
16:26:07 <hiro> yep... was checking why I do not have any invite and what calendar I should sync to see those
16:26:27 <juga> TPI Meetings, i think
16:26:36 <hiro> yep just realized
16:26:39 <hiro> thanks!
16:26:43 <juga> yw!
16:26:48 <hiro> anyways let's end the meeting, talk tomorrow
16:27:00 <hiro> #endmeeting