>I don't really see anything suspicious in the profile. This looks more
>like a kernel scheduler issue than a postgres bottleneck one. It seems
>that somehow using nonblocking IO (started in 9.5) causes scheduling
>issues when pgbouncer is also local.
>
>Could you do perf stat -ddd -a sleep 10 or something during both runs? I
>suspect that the context switch ratios will be quite different.
Perf show that in 9.5 case context switches occurs about 2 times less.
Perf output is attached.
Regards,
Dmitriy Sarafannikov
