Terry,
On Fri, Jul 13, 2001 at 04:17:41PM +0000, Terry Carlin wrote:
> I have not been able to check your response (if any) to my last email
> about the info I sent you indicating what I thought was a leak where the
> Win2k machine runs out of handles.
>
> Was there any further info about this or anything you need for me to do
> to help troubleshoot this problem?
I did reply by email. Unfortunately, it appears that you did not
receive it. See attached.
Please confirm when you receive this email. I very much want to work
with you to solve the resource leak problem(s).
Thanks,
Jason
--
Jason Tishler
Director, Software Engineering Phone: 732.264.8770 x235
Dot Hill Systems Corp. Fax: 732.264.8798
82 Bethany Road, Suite 7 Email: Jason.Tishler@dothill.com
Hazlet, NJ 07730 USA WWW: http://www.dothill.com
Terry,
[Sorry for the sluggish response time -- those pesky holidays and
vacations are always getting in the way...]
On Thu, Jul 05, 2001 at 09:11:09PM +0000, Terry Carlin wrote:
> On 7/4/01, 3:31:37 PM, Jason.Tishler@dothill.com (Jason Tishler) wrote
> regarding Re: PostgreSQL on Cygwin:
>
> > On Tue, Jul 03, 2001 at 09:08:09PM +0000, Terry Carlin wrote:
> > > I posted this message on the postgresql.ports.cygwin list and have not
> > > heard anything from anybody. Tom Lane suggested that I contact you
> > > directly with this problem.
>
> > I monitor the pgsql-cygwin@postgresql.org mailing list very closely
> > and I did not see any of your posts. To what mailing list does
> > "postgresql.ports.cygwin" correspond?
>
> I don't subscribe to the mailing list, I posted to
> comp.databases.postgresql.ports.cygwin news group.
My guess is that comp.databases.postgresql.ports.cygwin is tied to
pgsql-cygwin@postgresql.org and vice versa, but it appears that messages
get lost between the two occasionally.
Given your current role, I would recommend subscribing to pgsql-ports
and pgsql-cygwin -- the traffic is fairly light.
> > My WAG is that you have found a Cygwin resource leak -- my best guess is
> > open handles. Since the activity of a benchmark application is by its
> > very nature intensive and "long lived" as compared to someone typing
> > away at bash, it would be much more likely for you to stumble over a
> > resource leak than the typical Cygwin user.
>
> This is why I was running the benchmarks against PostgreSQL on CYGWIN.
>
> > Try adding more columns (e.g, Handle Count) in Task Manager to determine
> > whether or not my hypothesis is correct. Another option is to use
> > Performance Monitor.
>
> I added the Handle Count to the Task Manager and found some interesting
> stuff. These are the memory and handle statistics from this run.
>
> This machine has 256 mb ram
> PostgreSQL Memory size Process 1 6,632k, # of handles 756,889
> PostgreSQL Memory size Process 2 6,636k, # of handles 474,435
> PostgreSQL Memory size Process 3 6,640k, # of handles 247,353
> PostgreSQL Memory size Process 4 6,632k, # of handles 56,515
> Does look like we ran out of handles
I would like to reproduce your resource leak problem without using
Benchmark Factory. Can you suggest a way to "emulate" your testing by
driving psql or using JDBC? If I can reproduce the problem, then I may
be able to plug the leak. If so, then I will submit a patch to Cygwin
for consideration. Otherwise, I can post a hopefully minimal test case
that demonstrates the problem to the core Cygwin developers.
Thanks,
Jason
--
Jason Tishler
Director, Software Engineering Phone: 732.264.8770 x235
Dot Hill Systems Corp. Fax: 732.264.8798
82 Bethany Road, Suite 7 Email: Jason.Tishler@dothill.com
Hazlet, NJ 07730 USA WWW: http://www.dothill.com
