Re: postgresql latency & bgwriter not doing its job

Hello Aidan,

> If all you want is to avoid the write storms when fsyncs start happening on
> slow storage, can you not just adjust the kernel vm.dirty* tunables to
> start making the kernel write out dirty buffers much sooner instead of
> letting them accumulate until fsyncs force them out all at once?

I tried that by setting:
vm.dirty_expire_centisecs = 100
vm.dirty_writeback_centisecs = 100

So it should start writing returned buffers at most 2s after they are
returned, if I understood the doc correctly, instead of at most 35s.

The result is that with a 5000s 25tps pretty small load (the system can do
300tps with the default configuration), I lost 2091 (1.7%) of
transactions, that is they were beyond the 200ms schedule limit. Another
change is that overall the lost transactions are more spread than without
this setting, although there still is stretches of unresponsiveness.

So although the situation is significantly better, it is still far from
good with the much reduced value I tried.

--
Fabien.