!vxTmkuJzhGPsMdkAOc:transformierende-gesellschaft.org

NixOS Matrix Subsystem

150 Members
Coordination and discussion about the matrix subsystem in NixOS - https://wiki.nixos.org/wiki/Matrix74 Servers

Load older messages


SenderMessageTime
19 Feb 2022
@f0x:pixie.townf0x
In reply to @hexa:lossy.network
but what would you people actually be using if both were available?
the vast majority just needs some fixed worker types
01:23:02
@dandellion:dodsorf.asDandellioni'd use IP in that case01:23:02
@hexa:lossy.networkhexabut how would we generalize towards the user?01:23:04
@f0x:pixie.townf0xto split the important parts into a few processes01:23:11
@f0x:pixie.townf0xtbqh most people might not even need multiple of the difficult to load balance types01:23:52
@hexa:lossy.networkhexaanyway, if you have ideas for how to solve these problems, dump them into the document01:27:18
@hexa:lossy.networkhexa I'm off for the night \o 01:27:31
@dandellion:dodsorf.asDandellionIt's 2 am for me as well, I'll share my thoughts tomorrow :)01:28:06
21 Feb 2022
@cw:kernelpanic.cafeChinchilla Optional changed their display name from Chuck Winter to Chinchilla Wetreat.00:47:14
@janejasperous:one.ems.host@janejasperous:one.ems.host joined the room.04:28:35
@hexa:lossy.networkhexa ma27: what's your schedule? 🙂 23:00:00
22 Feb 2022
@hexa:lossy.networkhexaadded two ideas on how to delay worker starts to the pad16:03:43
@hexa:lossy.networkhexa one uses systemctl is-active, the other uses the /health endpoint on the main instance 16:04:18
@hexa:lossy.networkhexa Also Dandellion's worker units should probably use Requires=matrix-synapse.service, so the workers shut down when th main instance does 16:07:03
@hexa:lossy.networkhexa * Also Dandellion's worker units should probably use Requires=matrix-synapse.service, so the workers shut down when the main instance does 16:07:13
@dandellion:dodsorf.asDandellionI'll add it to the requires if enableMainSynapse is true ig17:32:41
@philipp:xndr.dephilipp Minor bump in mtxclient that fixes a bug for nheko: https://github.com/NixOS/nixpkgs/pull/161403 17:50:16
23 Feb 2022
@hexa:lossy.networkhexa even with a bunch of workers joining #neovim:matrix.org doesn't just work 15:02:51
@hexa:lossy.networkhexathe main instance goes up to eleven for a minute until the request times out15:03:30
@hexa:lossy.networkhexaand >90% of the log entries in the matrix-synapse.service point to presence15:04:17
@dandellion:dodsorf.asDandellionpresence was historically so bad you basically couldnt have it enabled for any mid sized instance. I think it's gotten better in the past year, but I still haven't enabled it15:06:15
@hexa:lossy.networkhexaalso I was joined to the bridged libera room on the first join, it's just element hasn't joined yet 😀15:09:52
@hexa:lossy.networkhexahm ok15:10:02
@dandellion:dodsorf.asDandellion * presence was historically so bad you basically couldnt have it enabled for any mid sized instance. I think it's gotten better in the past yeara, but I still haven't enabled it15:10:34
@hexa:lossy.networkhexalooks like it is default enabled in the example config https://github.com/matrix-org/synapse/blob/a711ae78a8f8ba406ff122035c8bf096fac9a26c/docs/sample_config.yaml#L119-L12315:10:40
@dandellion:dodsorf.asDandellionyeah, it is. I'm not sure if it's supposed to be actually problematic anymore though15:12:45
@f0x:pixie.townf0xafaik it still slogs cpu15:14:49
@f0x:pixie.townf0xpresence stuff over federation are just EDU's though so I don't think you can filter them to a specific worker15:16:52
@hexa:lossy.networkhexamy synapse host has ~45 Mbit/s ingress while joining the neovim room15:17:50
@f0x:pixie.townf0xim not sure why presence would be much of an issue specifically during joining though15:20:06

Show newer messages


Back to Room ListRoom Version: 4