Piece_expiration.db still huge after SL graceful exit

Synology+Docker, node ver. 1.114.6. Node size ~3-4TB. Many nodes have the same problem.
I graceful exited Saltlake a few months back, force forgeted the sat, succesfuly ran the filewalker 2 times already, but piece_expiration.db is still huge, 20GB in size.
Shouldn’t it be shriked? Is there a bug in the code that made the node to forget to delete the Saltlake pieces info from db?
What’s the solution?
If I delete the db, should I also delete the piece_expirations dir, with all those .dat files?

Edit: It seems that the Vaccum did the trick and shrinked the db file to 1.4MB. WOW!
I also shrinked the db for an older node that did’t had SL from almost 2 years, from 344MB to 4MB.
So, I guess, the Vaccum is necessary afterall.

4 Likes