>>>> AR == Alex Rousskov <rousskov@measurement-factory.com> writes:
AR> There are no known memory leaks in 2.5.4. Requests that are waiting
AR> for connection slot to become available (see open_conn_lmt) do consume
AR> memory and will increase memory footprint if the queue is growing. If
AR> the queue is growing, Polygraph may run out of memory (sooner or
AR> later, depending on the queue growth speed and available memory).
AR>
AR> Finally, as I tried to tell chin60 several times, if cache response
AR> time is too high for a given workload, it probably means that the
AR> request rate is too high for that cache configuration, and that
AR> polyclt running PolyMix workload may eventually run our of memory
AR> under these conditions. This behavior is intentional to prevent
AR> best-effort "PolyMix" tests (which PolyMix becomes if the queues are
AR> not drained) from being considered successful.
I haven't noticed that polymix workloads have rate set up for clients.
What is the maximum request rate you can get out of one polyclt? Web page says that polyclt running on FreeBSD can generate up to 700h/s. In my case, I can't get more then 150-180h/s out of polyclt running modification of simple.pg on Solaris. Is such a huge difference expected?
Thank you,
-- Artem VeremeyFortune cookie here: Oh, I get it!! "The BEACH goes on", huh, SONNY??
This archive was generated by hypermail 2b29 : Tue Jul 10 2001 - 12:00:18 MDT