The user Arsog1985 wants to be renamed to "Sigma'am" per this request. However, they have over 51,000 edits, and a sysadmin is needed to supervise their rename. This task is, therefore, a request for sysadmin attention.
The reason for the request can be per [[meta:Special:GlobalRenameQueue/request/37897|request]] and [[phab:180903]].
Thanks.
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | None | T169440 Pending global renames in need of sysadmin supervision (tracking) | |||
| Resolved | Linedwell | T180903 Rename user "Arsog1985" to "Sigma'am" on Central Auth : supervision needed |
Typically, I've the feeling the operations tag could mean:
To overview renames tend to be the responsibility of the system administrators familiar with the CentralAuth code, which isn't generally considered as a scope of Operations, as it's a Mediawiki product.
Experience showed the script behaves well from a performance point of view, or cause issues at DBA level (so the tag is probably not a bad idea, and your reaction indicates you could be interested by it).
The worst possibility is a potential update bomb (the username is stored in too much MediaWiki tables, touching index columns).
In the past, the main concern was if the code was able to gracefully resume, and so it was interesting for the CentralAuth developers. A little less now.
I created the template based on previous tasks for this kind of requests. All had the SRE tag. If it is not needed anymore, then yes; I will update the prefilled link.
I don't want to hijack this thread, but I do believe the SRE tag should still be present.
As Dereckon said, the process rarely affects any database, might generate a few seconds of lag, which is not relevant.
Any Opsen can supervise this change (remember there is always an on-duty Opsen), and if something more serious happens, they can directly ping any DBA to take a look.
Relying only on specific DBAs (there are only 2 of us) can be a serious bottleneck.
I don't want to hijack this thread, but I do believe the SRE tag should still be present.
As Dereckon said, the process rarely affects any database, might generate a few seconds of lag, which is not relevant.
Any Opsen can supervise this change (remember there is always an on-duty Opsen), and if something more serious happens, they can directly ping any DBA to take a look.
Relying only on specific DBAs (there are only 2 of us) can be a serious bottleneck.
Contradicting Marostegui, I do not think either SRE or DBA should be tags here. The fact that @Marostegui offered to help in past instances doesn't mean this is an ops or dba job- in case of a problem, an op cannot help, as it would be a code problem; no server has gone down because of a rename. It happens frequently that some renames get stuck, and I have no idea how to react to that.