Thanks for this, I read too quickly! I've attached the 2 perf reports. From the 2nd one, I can see lots of time waiting for TOAST table locks on the geometry column, but I definitely don't fully understand the implications or why LockManager would be struggling here. Thanks for the continued help! ---Paul Paul Friedman CTO 677 Harrison St | San Francisco, CA 94107 M: (650) 270-7676 E-mail: paul.friedman@xxxxxxxxxxxxxxxxxxx -----Original Message----- From: Andres Freund <andres@xxxxxxxxxxx> Sent: Monday, April 12, 2021 4:04 PM To: Paul Friedman <paul.friedman@xxxxxxxxxxxxxxxxxxx> Cc: pgsql-performance@xxxxxxxxxxxxxxxxxxxx Subject: Re: LWLocks by LockManager slowing large DB Hi, On 2021-04-12 15:56:08 -0700, Paul Friedman wrote: > Also, I didn't understand your comment about a 'futex profile', could > you point me in the right direction here? My earlier mail included a section about running a perf profile showing the callers of the futex() system call, which in turn is what lwlocks end up using on linux when the lock is contended. Check the second half of: https://www.postgresql.org/message-id/20210412215738.xytq33wlciljyva5%40al ap3.anarazel.de Greetings, Andres Freund
Attachment:
perf1.out
Description: Binary data
Attachment:
perf2.out
Description: Binary data
