[01:46:11] <MirahezeRelay>	 <blankeclair> any reason why we can't set https://github.com/miraheze/puppet/blob/7c18a17189fedf66f7ffe5e486b395c8a6eb2e51/modules/mediawiki/templates/mediawiki.conf.erb#L106-L153 to `server_name *`?
[02:42:47] <MirahezeRelay>	 <agentisai, replying to blankeclair> I actually swapped that with the first block so that MediaWiki was served by default but it broke some check I think and no one had the time to fix it so we reverted
[02:43:02] <MirahezeRelay>	 <blankeclair> ah :p
[02:43:34] <MirahezeRelay>	 <agentisai> https://github.com/miraheze/puppet/commit/82d893c9559ed844462b8299edb2ed0b39f83189
[02:43:41] <MirahezeRelay>	 <agentisai> seems mwdeploy broke somehow
[06:20:19] <MirahezeRelay>	 <rhinosf1> @agentisai that shouldn't have broken deployments unless something else was wrong with the commit
[06:20:41] <MirahezeRelay>	 <rhinosf1> I'd like to try it again at some point
[06:55:43] <MirahezeRelay>	 <posix_memalign, replying to raidarr> [1/3] I figured out a better solution for non-vector-2022 skins. If Extension:DarkMode is activated, the stylesheet would not apply any dark mode styles and let the extension handle everything. This change fixes the link colors, so everything should look good on the new version of the main page except for the "Start a wiki" button, which already looks bad on the
[06:55:43] <MirahezeRelay>	 <posix_memalign, replying to raidarr> [2/3]  current main page due to the inversion.
[06:55:43] <MirahezeRelay>	 <posix_memalign, replying to raidarr> [3/3] https://cdn.discordapp.com/attachments/1006789349498699827/1376091342236680222/image.png?ex=6834106e&is=6832beee&hm=9068b9dfb7d547a69df0045f4143973fed435dc7ef4c99c65800eabbb8016003&
[07:22:14] <MirahezeRelay>	 <posix_memalign> [1/2] Nevermind. I fixed the button color too. I think all changes compared to the original are strict improvements at this point.
[07:22:15] <MirahezeRelay>	 <posix_memalign> [2/2] https://cdn.discordapp.com/attachments/1006789349498699827/1376098018662682675/image.png?ex=683416a6&is=6832c526&hm=12b5dabfe833243f55e216def5f305f031c51e0a027c86db0bb2ae0da5dbb5d9&
[08:22:53] <MirahezeRelay>	 <blankeclair> [1/3] anyone happen to know if there is a proper way to add custom buttons to graylog messages? i wanna add one to "Copy formatted exception"
[08:22:53] <MirahezeRelay>	 <blankeclair> [2/3] (this message is not an exception, but shh)
[08:22:53] <MirahezeRelay>	 <blankeclair> [3/3] https://cdn.discordapp.com/attachments/1006789349498699827/1376113278882812005/1748161372178.png?ex=683424dc&is=6832d35c&hm=596d371e86a89fa3f614791cac0a5dab32b5021a4e3ef5594d45bf1b59fe2170&
[08:42:50] <MirahezeRelay>	 <blankeclair> the answer appears to be yes: https://github.com/Graylog2/graylog2-server/blob/3ef65b755a309558985e8139bcd6ad8071bc7a2f/graylog2-web-interface/src/views/components/messagelist/MessageActions.tsx#L68-L74
[10:02:46] <MirahezeRelay>	 <rhinosf1, replying to blankeclair> We could do a wmf and add a button on Phab to auto pull the exception
[10:03:04] <MirahezeRelay>	 <blankeclair> still needs to be manually vetted :p
[10:03:13] <MirahezeRelay>	 <blankeclair> alternatively, i could try to make a userscript
[10:03:32] <MirahezeRelay>	 <rhinosf1> @blankeclair yes
[10:03:36] <MirahezeRelay>	 <rhinosf1> It doesn't auto save
[10:03:44] <MirahezeRelay>	 <blankeclair> oh?
[10:03:57] <MirahezeRelay>	 <rhinosf1> But they have button so someone with Logstash access can use the reqID field on their exception form
[10:04:04] <MirahezeRelay>	 <rhinosf1> To auto open a logstash search
[10:04:16] <MirahezeRelay>	 <rhinosf1> And then there must be a way to auto copy to the exception fields
[10:04:19] <MirahezeRelay>	 <rhinosf1> Never used it
[10:04:48] <MirahezeRelay>	 <blankeclair, replying to blankeclair> opinyans on this route?
[10:04:50] <MirahezeRelay>	 <rhinosf1> https://wikitech.wikimedia.org/wiki/Phatality
[10:04:56] <MirahezeRelay>	 <rhinosf1, replying to blankeclair> Go for it
[10:05:00] <MirahezeRelay>	 <blankeclair> oki~
[11:10:45] <MirahezeRelay>	 <aeywoo> @blankeclair Cosmic must be off doing something else, did you want to review my PR for adding MobileFrontend settings? I already tested them on test151/mirabeta and if you want to check the settings then you can.
[11:11:11] <MirahezeRelay>	 <blankeclair, replying to aeywoo> linkie~
[11:11:40] <MirahezeRelay>	 <aeywoo, replying to blankeclair> https://github.com/miraheze/mw-config/pull/5928
[11:11:57] <MirahezeRelay>	 <aeywoo> Oh wait, there's some settings I think got sunset, I need to make a list rq.
[11:12:06] <MirahezeRelay>	 <blankeclair> hate it when my laptop decides to make usb stop working
[11:12:44] <MirahezeRelay>	 <aeywoo> Some wikis had settings that are now available through ManageWiki with this PR set through `LocalSettings.php`, so you might need to manually set the settings for them.
[11:14:25] <MirahezeRelay>	 <blankeclair> crap, can't really review, i gotta debug this usb issue ^^;
[11:14:42] <MirahezeRelay>	 <aeywoo> oh well, tanks anyway
[11:17:04] <MirahezeRelay>	 <blankeclair> [1/7] ```
[11:17:04] <MirahezeRelay>	 <blankeclair> [2/7] > lsusb
[11:17:04] <MirahezeRelay>	 <blankeclair> [3/7] Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
[11:17:05] <MirahezeRelay>	 <blankeclair> [4/7] Bus 002 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
[11:17:05] <MirahezeRelay>	 <blankeclair> [5/7] Bus 003 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
[11:17:05] <MirahezeRelay>	 <blankeclair> [6/7] Bus 004 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
[11:17:05] <MirahezeRelay>	 <blankeclair> [7/7] ```
[11:17:09] <MirahezeRelay>	 <blankeclair> well, usb really is dead huh
[11:19:59] <MirahezeRelay>	 <blankeclair> [1/10] nothing like turning it off and on again
[11:19:59] <MirahezeRelay>	 <blankeclair> [2/10] ```
[11:19:59] <MirahezeRelay>	 <blankeclair> [3/10] Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
[11:20:00] <MirahezeRelay>	 <blankeclair> [4/10] Bus 002 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
[11:20:00] <MirahezeRelay>	 <blankeclair> [5/10] Bus 003 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
[11:20:00] <MirahezeRelay>	 <blankeclair> [6/10] Bus 003 Device 002: ID 0c45:6740 Microdia Integrated_Webcam_FHD
[11:20:01] <MirahezeRelay>	 <blankeclair> [7/10] Bus 003 Device 003: ID 0a5c:5843 Broadcom Corp. BCM58200 ControlVault 3 (FingerPrint sensor + Contacted SmartCard)
[11:20:01] <MirahezeRelay>	 <blankeclair> [8/10] Bus 003 Device 004: ID 1050:0407 Yubico.com Yubikey 4/5 OTP+U2F+CCID
[11:20:01] <MirahezeRelay>	 <blankeclair> [9/10] Bus 004 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
[11:20:01] <MirahezeRelay>	 <blankeclair> [10/10] ```
[11:25:19] <MirahezeRelay>	 <aeywoo> nice
[11:42:02] <MirahezeRelay>	 <awing_ding, replying to blankeclair> Reboot
[11:42:08] <MirahezeRelay>	 <awing_ding> It happens when you update kernel
[11:42:21] <MirahezeRelay>	 <awing_ding> And don't reboot afterwards
[11:42:22] <MirahezeRelay>	 <blankeclair, replying to awing_ding> i didn't update the kernel though
[11:42:35] <MirahezeRelay>	 <blankeclair> and internal usb devices disappeared ^^;
[11:42:41] <MirahezeRelay>	 <awing_ding> Wtf
[11:43:24] <MirahezeRelay>	 <awing_ding> Ig kernel just freaked out for some reason
[13:11:49] <MirahezeRelay>	 <originalauthority, replying to rhinosf1> Opensearch Dashboards is so bad
[14:47:26] <MirahezeRelay>	 <paladox, replying to originalauthority> it's logstash and kibana but just under the Opensearch Dashboards name
[15:37:53] <MirahezeRelay>	 <originalauthority> Its not logstash
[15:37:58] <MirahezeRelay>	 <originalauthority> Logstash is a completely different thing
[15:39:28] <MirahezeRelay>	 <abaddriverlol> [1/2] did somebody edit a login template or how can I interpret this
[15:39:28] <MirahezeRelay>	 <abaddriverlol> [2/2] https://cdn.discordapp.com/attachments/1006789349498699827/1376223149200379954/image.png?ex=68348b2f&is=683339af&hm=390f133183b6b573c76d60845dc4b4d7a49444a4affa5af4875d7ce0fa0229f8&
[15:39:59] <MirahezeRelay>	 <originalauthority> Kafka job queue so good theres 910k jobs for one job lol
[15:40:02] <MirahezeRelay>	 <originalauthority> Sucks ass
[15:40:25] <MirahezeRelay>	 <pixldev, replying to originalauthority> what
[15:40:35] <MirahezeRelay>	 <originalauthority> What do you mean what
[15:44:16] <MirahezeRelay>	 <pixldev, replying to originalauthority> what hat
[15:44:20] <MirahezeRelay>	 <pixldev> what that
[15:57:17] <MirahezeRelay>	 <abaddriverlol, replying to originalauthority> isn't it one job type but multiple jobs
[15:57:25] <MirahezeRelay>	 <originalauthority> Yes
[15:57:29] <MirahezeRelay>	 <originalauthority> Thats what i meant hehe
[15:57:34] <MirahezeRelay>	 <abaddriverlol> ah
[16:07:27] <MirahezeRelay>	 <rhinosf1, replying to abaddriverlol> Someone made like 30 edits when editing their user page
[16:07:32] <MirahezeRelay>	 <rhinosf1> It'll go down fine eventually
[16:07:39] <MirahezeRelay>	 <abaddriverlol, replying to rhinosf1> i saw
[16:07:46] <MirahezeRelay>	 <abaddriverlol> but how does that result in 900k jobs
[16:08:22] <MirahezeRelay>	 <rhinosf1> Oh they ain't even stopped yet
[16:08:35] <MirahezeRelay>	 <rhinosf1> Cause GUP generates like 3 jobs on every wiki for every edit
[16:09:34] <MirahezeRelay>	 <abaddriverlol, replying to rhinosf1> they made 24 template edits too but none of those are transcluded on other user pages i think
[16:09:41] <MirahezeRelay>	 <abaddriverlol> doesn't GUP check whether the edits are actually relevant
[16:10:41] <MirahezeRelay>	 <rhinosf1> I think it'll only apply to the edits to their user page
[16:10:48] <MirahezeRelay>	 <rhinosf1> otherwise we'd have seen a lot more
[16:11:05] <MirahezeRelay>	 <rhinosf1> If he'd made 24 edits to a userbox used by a lot of people, I'd have blocked him
[16:11:15] <MirahezeRelay>	 <rhinosf1> Because then we'd have like a billion jobs
[16:12:20] <MirahezeRelay>	 <abaddriverlol> oh yeah it's only the user page, the edits perfectly match up with the job queue graph
[16:12:55] <MirahezeRelay>	 <rhinosf1> Once it calms itself down, it'll probably drop half the jobs in the space of like half an hour
[16:13:41] <MirahezeRelay>	 <abaddriverlol> ah
[16:13:42] <MirahezeRelay>	 <abaddriverlol> cool
[16:13:48] <MirahezeRelay>	 <rhinosf1> https://github.com/miraheze/puppet/blob/7c18a17189fedf66f7ffe5e486b395c8a6eb2e51/hieradata/hosts/changeprop201.yaml#L38
[16:14:01] <MirahezeRelay>	 <rhinosf1> @paladox does that mean it can only run 15 at once?
[16:14:21] <MirahezeRelay>	 <rhinosf1> If so that's really low
[16:14:24] <MirahezeRelay>	 <rhinosf1> They are cheap jobs
[16:14:31] <MirahezeRelay>	 <originalauthority> But it shouldn't
[16:15:14] <MirahezeRelay>	 <originalauthority> Because the job queue support de duplication, so the next job that enters the queue for the same page should evict the other jobs with the same root page
[16:16:12] <MirahezeRelay>	 <rhinosf1> @originalauthority there should be some deduplication too ye
[16:18:25] <MirahezeRelay>	 <paladox, replying to rhinosf1> Well concurrency. But on https://grafana.wikitide.net/d/CbmStnlGk/jobqueue-job?orgId=1&from=now-2d&to=now&timezone=browser&var-job=LocalGlobalUserPageCacheUpdateJob&refresh=5m it says like a 150 jobs ran.
[16:19:24] <MirahezeRelay>	 <rhinosf1, replying to paladox> Is it worth upping it?
[16:19:31] <MirahezeRelay>	 <rhinosf1> I'm fairly sure they are extremely cheap
[16:22:14] <MirahezeRelay>	 <abaddriverlol> yeah it looks like it basically just purges the pages <https://gerrit.wikimedia.org/r/plugins/gitiles/mediawiki/extensions/GlobalUserPage/+/b06719264da28a2f1ab1f9c1863a0bf6ff313c43/includes/LocalCacheUpdateJob.php#38>
[16:23:44] <MirahezeRelay>	 <paladox> I’m not sure. We’ve only got 4 mwtask servers and I don’t really want to overload it.
[16:32:24] <MirahezeRelay>	 <rhinosf1> @paladox fairly sure we could run a thousand on them jobs at once and it wouldn't overload much
[16:34:09] <MirahezeRelay>	 <rhinosf1> 150 jobs in 5 minutes for that is slow
[16:34:16] <MirahezeRelay>	 <rhinosf1> That's like 1 every 2 seconds
[17:54:15] <MirahezeRelay>	 <cosmicalpha, replying to paladox> I want to change this btw. Have 2 mwtask servers (just 1 as a backup) and then have like 4-6 additional jobrunner-only (not task) servers if necessary. But we don't need 4 mwtask servers even.
[17:56:16] <MirahezeRelay>	 <cosmicalpha> Also have 1 or 2 jobrunner servers just for videoscaler maybe.
[18:23:29] <MirahezeRelay>	 <rhinosf1, replying to cosmicalpha> We should probably rename mwtask to mwmaint or something
[18:23:42] <MirahezeRelay>	 <rhinosf1> and have jobunnerXXX
[18:23:53] <MirahezeRelay>	 <rhinosf1> with like 2 maybe prioritised for videoscaling
[18:24:27] <MirahezeRelay>	 <rhinosf1> not fully videoscaler but that prioritises it over other stuff when needed and can only run video scaling
[19:29:11] <MirahezeRelay>	 <agentisai> videoscaler is pending GPUs
[19:49:00] <MirahezeRelay>	 <notaracham> Once PayPal money lands, we should be well situated to add 'em, with any luck
[20:39:43] <MirahezeRelay>	 <pixldev, replying to notaracham> Oh! How much
[20:45:10] <MirahezeRelay>	 <notaracham> $9k or thereabouts was sitting thanks to a broken automated transfer.  Very well situated for working capital, though at our rate of expansion, we'll see...
[20:45:55] <MirahezeRelay>	 <pixldev, replying to notaracham> huh
[20:46:11] <MirahezeRelay>	 <notaracham> Oh, how much for the GPUs?
[20:46:42] <MirahezeRelay>	 <notaracham> @agentisai or CA would remember specifics, I think we were looking around 750 for what we wanted to do.
[21:01:59] <MirahezeRelay>	 <pixldev, replying to notaracham> only? I thought it’d be way more
[21:14:00] <MirahezeRelay>	 <notaracham> 750 might have been the black Friday sale price
[21:28:14] <MirahezeRelay>	 <pixldev, replying to notaracham> Ah
[21:43:56] <MirahezeRelay>	 <rhinosf1, replying to notaracham> Bit past Black Friday
[21:46:19] <MirahezeRelay>	 <notaracham> Just a little bit, yes.
[21:52:38] <MirahezeRelay>	 <pixldev, replying to notaracham> Get a memorial day weekend price
[23:14:16] <MirahezeRelay>	 <agentisai, replying to notaracham> sounds about right