I ran into this error a few days ago for the first time. I think this narrows it down to the last code deployment (last Thursday?)

@Legoktm Done aswell.

Using Google Translate it becomes more clear on why this is requested. The wiki has been plagued with excessive vandalism from long-term abusers. An example of this can be found on https://uk.wikivoyage.org/w/index.php?title=%D0%92%D1%96%D0%BA%D1%96%D0%BC%D0%B0%D0%BD%D0%B4%D1%80%D0%B8:%D0%9A%D0%BD%D0%B0%D0%B9%D0%BF%D0%B0&action=history. I do think this "RfC" should atleast pass a running time of a week. Apart from that it looks like a reasonable request to me

See T258336. Database host crashed yesterday, which in turn creates replication lag. See below for snippit from that task

In T258336#6317095, @Stashbot wrote:

Mentioned in SAL (#wikimedia-operations) [2020-07-18T20:52:50Z] <marostegui> Due to db1082 crash there will be replication lag on s5 on labsdb hosts - T258336

Part of this issue sounds like T252236, where I describe that newer browser versions severely break CentralAuth, resulting in no global login...

Not only Chrome will have issues with, but Firefox too. As a matter of fact. I run pre-release versions of both browsers (I know, not recommended, but doesn't give much problems usually) and on both the login process is severly broken. Let's say I login on my homewiki (nlwiki). I will get a cookie for *.wikipedia,org as its the same domain. Same goes for wikis that are on *.wikimedia.org, because as far I understand a cookie will be set for that through login.wikimedia.org. Any other domains (Wikiquote, Wikinews, Wikidata, etc...) will not log you in and will require you to login normally, which can be a real pain if you are someone who works on a multitude of different projects (SWMT) and use 2FA. I can however note, that current release versions of these browsers and those based on it still work okayish (still requires me to force-reload page, when in the past that wasn't even neccesary).

Do you mean you don't want that message to show up? You have two ways of getting rid of it, the first one is blanking and the second is deleting the page.

I don't think my message was understand quite correct, not that it makes a difference as the linked page explains it well. This request from a bureaucrat on a Wikipedia. There was no local discussion on that wiki, hence no consensus. I thought in that case it would be handled differently, but that's not the case (looking at the page). That's on me for not knowing (I haven't been hanging out here until recently, just to help where-ever I can).

@Aklapper It was more of a question on how stuff like this is usually handled. I thought that team could answer that question, but maybe there is someone more appropriate that could answer it.

Yea, but generally you need consenses from the local wikis that are going to affected by such a change. There are no local discussions as far I see. Don't know what the exact procedure would be if it were to be requested from like a Wikipedia instead of a Wikibooks for example.

I am not someone from ops, but after looking at last month of edits I couldn't find any page that discusses this change. Furthermore it seems you are the only really active user on that wiki. Changes like this need to reflect consensus, and one person is not consensus.

Tool-wise the database code for this project is also shared (after looking at the individual index.php files) with other tools under the tools.meta project. See github here above. It could be very well possible the oudated meta_p causes for issues, as it probably assumes its still there.

@Krenair @Ladsgroup sqwiki on Labs is affected too, it seems, see: https://sq.wikipedia.beta.wmflabs.org/wiki/Speciale:NdryshimeS%C3%ABFundmi

Invalid. Asked about this on IRC, and personal accounts are not used for database access anymore (you should use tool accounts for this for what I have been told). Documention on Wikitech (https://wikitech.wikimedia.org/wiki/Help:Tool_Labs/Database) should be updated to reflect this.

Thank you very much for that! Just one question let: If I would be added as maintainer to another project (not that it will happen really soon) this problem could theoretically come back, right? Or do I have to see this as: your are fine for the forseeable future?