Tobias Brox <tobias@xxxxxxxxxxxxx> writes: >> Blocked on someone else's lock, maybe? > hardly, the autovacuum is the only one having such a low transaction id, > and also the only one hanging around when waiting a bit and rechecking > the pg_locks table. Hmph. Is the autovac process actually doing anything (strace would be revealing)? If not, can you attach to the autovac process with gdb and get a stack trace to see where it's blocked? regards, tom lane