[02:16:26] So pictures won't display at all for a whole 3-4 weeks? I thought it would be intermittent downages, but damn, that's brutal. [02:23:29] some images may have issues but most images will be ok. that said, all images are subject to intermittent issues during the migration. [02:24:28] i should also note that an image "issue" might also relate to a browser cache problem. so before assuming there is an image issue, please purge your cache. [02:27:46] please look to the future. in a couple weeks, images will be hosted on new servers that will be much more stable and responsive. as annoying as the present intermittency is, it's not like the images are lost. not sure I'd classify that as brutal. [02:49:16] Ah, okay... I have a bunch of images that just won't seem to load no matter what I do (clear cache, etc.) since yesterday, but I do see that some other images are working, so hopefully it's just things going in and out. [02:58:11] yeah, i'd not be happy right now, in your shoes, but if it's not better in a few weeks, then it's time to maybe file a task. but i'd recommend waiting until there's a message stating image migration is complete. [03:06:00] If it seems unique to your wiki and consistent, I'd recommend making the task sooner (few days instead of few weeks) [03:18:49] I wouldn't fret too much [03:18:58] The Miraheze logo disappeared for a solid 12+ hours and finally revived [03:30:53] For me [03:31:00] It appears it’s now a blank picture [03:32:09] its a conspiracy [03:32:14] WikiTide takeover! [03:32:58] Probably for the better [03:33:06] We call this: Miraheze 2 [13:03:20] Womp fucking Womp there goes db131 [13:04:07] morning mate [13:04:18] Morning legroom [13:12:12] db151 attempting to be repaired [13:47:12] db151 is back online btw [17:30:49] @pixldev I am geniunely confused what on earth useful parsoid's pre-emptive cache warming is doing [18:45:25] @pixldev since the task server picked up the read write change, we've gone down 20k jobs in 10 minutes [18:45:41] so that means all jobs caught up in like 1.5 hours [18:46:06] huh [18:46:21] so it was the write part that broke it [18:46:29] it was being read only [18:46:33] it's now clearing [18:46:37] i have no idea why [18:46:46] but new servers are doing well at clearing backlog [18:47:14] 117,558 jobs was the peak record [18:47:21] for parsoidprecachewarm [18:48:07] not sure what total peak for all jobs will be yet [18:48:55] Maybe cause the writes were disabed it couldnt save and complete jobs [18:50:20] Ye but it shouldn't be generating at that rate [18:50:27] That's a very high rate for cache warning [18:50:42] Logs wouldn’t give any clues? [18:50:48] That means 4-5 requests every second hit a cache miss [18:51:30] Well, that’s, uh whats the word [18:51:31] oh yes [18:51:33] Absurd [18:51:37] Exactly [20:38:10] @Site Reliability Engineers are we good to restart wiki creations? [20:38:36] I believe so yes, just let us know if you encounter any issues. [20:39:17] https://www.jewishnews.co.uk/fury-as-labour-mp-claims-holocaust-memorial-day-should-recognise-gaza-genocide/ [20:39:25] oh bugger wrong channel [20:39:34] wrong server even [20:42:39] Welcome to the next-gen [20:50:16] Yeah, I discourage posting political stuff in this server [20:54:59] Blake, light guidance: you're not a server moderator, please don't moderate. [20:57:53] <.ameisenigel> Is there any chance to adapt https://meta.wikimedia.org/wiki/User:Krinkle/Tools/Global_SUL for Miraheze. WikiDiscover is similar to SiteMatrix but the API results look a bit different. [20:59:23] Probably but it's a bad idea [21:00:00] [1/2] fyi i think i18n is broken here [21:00:00] [2/2] https://cdn.discordapp.com/attachments/1006789349498699827/1200908087700230225/Screenshot_2024-01-27_at_12.59.41_PM.png?ex=65c7e3cf&is=65b56ecf&hm=8f6c323b5a7327fe8bef159e2a93eb3c44c5c657a36d76a33264a1f5a8abee64& [21:00:05] simple fix if someone can tell me where to do [21:02:15] Check if the message exists in en.json on the relevant extension [21:02:22] Although that's a badly named message [21:02:32] Yep [21:02:35] Maybe search for that message and see where it is wrote [21:02:39] I'll check [21:02:57] <.ameisenigel, replying to rhinosf1> Why would it be bad? Currently there is no easy way to get a list of all wikis or to search through all wikis. [21:03:02] It might be managewiki-defaultpermissions-norights [21:03:18] Because all that does is create you an account on every wiki [21:03:22] en.json or extension.json [21:03:24] We have a lot more than wikimedia [21:03:32] en.json [21:03:41] CentralAuth doesn't really scale well [21:03:52] It will make it really slow for you to login [21:04:08] doesn't exist [21:04:33] Work out if it should exist or that's a typo [21:04:45] Probably should, I can just add a line and PR it [21:05:23] Given it looks like the wrong format for a message name from an extension. It shouldn't [21:05:47] It should be manage wiki permissions I think [21:06:01] maybe not [21:07:11] It's Special:ManageWikiDefaultPermissions [21:07:18] yeah [21:07:21] oh so maybe not [21:07:25] how should I remove? [21:08:54] Oh, yes. Creating an account on every wiki would be a very good way to make your account unusable [21:12:59] I'm fixing this [21:13:39] got it, thanks [21:16:08] @bwm0 https://github.com/miraheze/ManageWiki/pull/443 [21:16:32] ty