Re: Improving connection scalability: GetSnapshotData()

On 04.09.2020 21:53, Andres Freund wrote:
>
>> May be it is because of more complex architecture of my server?
> Think we'll need profiles to know...

This is "perf top" of pgebch -c 100 -j 100 -M prepared -S

   12.16%  postgres                           [.] PinBuffer
   11.92%  postgres                           [.] LWLockAttemptLock
    6.46%  postgres                           [.] UnpinBuffer.constprop.11
    6.03%  postgres                           [.] LWLockRelease
    3.14%  postgres                           [.] BufferGetBlockNumber
    3.04%  postgres                           [.] ReadBuffer_common
    2.13%  [kernel]                           [k] _raw_spin_lock_irqsave
    2.11%  [kernel]                           [k] switch_mm_irqs_off
    1.95%  postgres                           [.] _bt_compare


Looks like most of the time is pent in buffers locks.
And which pgbench database scale factor you have used?