I'd say since we fixed T215107: Global rename of The_Photographer → Wilfredor: supervision needed with rERENeb847453fece: Batch renames in image tables renames are being very slow.
This rename started at 15:09, 4 March 2019 (UTC) and as of the time I'm writting (15:49, 4 March 2019 (UTC)) the rename is still in progress.
The account does not have a particularly high edit count.
Could eb847453fece be the cause of this recent slowness?
Same issue I've reported in T217203#4988080. It seems this is happening from time to time. Some requests are processed slower than before, whereas some other are stuck at the beginning for a while. I've also noticed that the issue is not related to single requests at a time. For example, I could see up to 7 renames getting stuck until they all started together.
Maybe jobs are held in the queue for long? It would be weird, we have used use the same kind of job already for edits, the patch just moved images there.
It looks to me that it's not the renaming process itself but the speed in which the job is sent to the wiki to process. That is, renames are now "queued" for longer than before. EventBus issue?
Example: https://meta.wikimedia.org/wiki/Special:GlobalRenameProgress?username=WhiteDreamsIndaHouse started at 09:51, 7 March 2019 but it was not until 09:54, 7 March 2019 when the first rename happened.