On Fri, Oct 11, 2019 at 09:20:59AM -0400, Stephen John Smoogen wrote:
OK I think we have a problem with the above value. the db servers
seems to be running at a load average of over 140 consistently and we
The load has actually always been nearly that high when the db dump is
happening.
have been getting timeouts and problems with nagios. I think we
should
look at dropping that to 32 and see if that causes the load average to
lower. While the ssd's are much faster, they do not seem to have the
cache and it looks like IO is backing up.
The timeouts however are new.
I'm not sure what is going on. I'll do more investigation today...
The issues seem to have started after I added memory to the vm.
I suppose if we wanted we could go all the way back to the orig postgres
config and drop the vm back to 32gb memory, but it seems kind of crazy
that more memory would cause more slowness. ;(
kevin