RE: Testing Squid with Polygraph

From: Ahuja Sadhna (NRC/Boston) (
Date: Wed Oct 03 2001 - 13:04:51 MDT

Hi Duane,
        Thanks for your prompt reply. Here are answers to the questions
you asked:

> How many disk drives in your Squid? What kind?

Just one drive - an IBM ATA drive, 9GB.

> Did you try a proxy-only (all replies uncachable) workload?

Yes, there are no errors in this case.

> Did you recompile Squid after increasing kern.maxfiles? Does
> Squid see the new filedescriptor limit?

I hadn't done this earlier, but now I have. Squid sees the new limit as
its SQUID_MAXFD value. With this change, after running the tests for
some time, I see the following message continuously on Squid:
comm_open: socket_failure: (55) No buffer space available
But, the Polygraph tests continue to run normally (or at least they look
normal to me).

> Did Squid run out of network mbufs? What does 'netstat -m' say?
> Did you increase NMBCLUSTERS on the Squid box?

I didn't increase NMBCLUSTERS. Here is the output of netstat -m, while
Squid was complaining about no buffer space:

1158/1968/4096 mbufs in use (current/peak/max)
        115 mbufs allocated to data
        1043 mbufs allocated to packet headers
98/930/1024 mbuf clusters in use (current/peak/max)
2352 bytes alloacted to network (76% of mb_map in use)
0 requests for memory denied
0 requests for memory delayed
0 calls to protocol drain routines

> "Host is down" implies that polyclt cannot determine Squid's
> ethernet address. What does 'arp -n' say? What kind of NICs and
> switch are you using?

I am sorry I was getting the "Host is down" message because of a stupid
configuration error on my part. I have resolved it now.

Will the results that I am logging, while Squid is still complaining,
make any sense? Should I reduce the request rate even lower than 50
req/sec? What else could be wrong?

Thanks for your help,

> On Tue, 2 Oct 2001, Ahuja Sadhna (NRC/Boston) wrote:
> > Hi,
> > I am trying to test my Squid cache with the polymix-3 workload.
> > (Note: Sorry, I am still using older versions - Polygraph 2.5.4 and
> > Squid 2.3-STABLE4). I am running polysrv on 1 machine and polyclt on
> > another. The request rate being used is 50 req/sec, which I think is
> > quite low. But, after running for a few minutes, I start seeing
> > "Connection Refused" and "Host is down" errors on the
> polyclt side and
> > then no more requests are generated. Squid does not seem to complain
> > about anything. (It did complain about 'no buffer space
> available' when
> > I first tried to run the test with higher request rates).
> > I also doubled up the values of kern.maxfiles and
> > kern.maxfilesperproc on the FreeBSD box running Squid, but
> same results.
> > I have used the same machines with Polygraph to test a cache
> > from another vendor with request rate = 400 req/sec, so I
> don't think
> > the Polygraph machines are a problem here. So, is there something
> > special that I need to do to test Squid?

This archive was generated by hypermail 2b29 : Mon Feb 06 2006 - 12:00:20 MST