ASP 0113, 0147 and IIS state

Hi folks,

I'm experiancing some slow downs and timeouts on my web servers.
The majority of the time we get periods of slow returning pages with periods
of excessive wait times
for requests which throw ASP 0113 and ASP 0147 errors. The thing is the
servers do not even
appear to be breathing hard. The only thing I am seeing is ASP requests
queuing.
Does not appear to be load induced.

No excessive CPU utilization (indicitive of looping)
No excessive memory utilization.

I do run the windbg user mode dumps on the servers but have not
recieved any user dumps based on exception errors.

We are also not able to specifically match the timing to any lock or
deadlocks
against the DB server.

Our monitor sees that the server is returning ASP pages in the 20 second
time frame a(a very basic test page)
and takes the load balanced server out-of-service. After a few seconds the
connections are dropped and it puts
the server back in. No IIS restart, No server reboot, No restarting of the
custom com+ applications is required.
Adding more servers has not alleviated the problem....

Can I use IIS state in as automated method to capture the wait state? We
have 12 servers that
may or may not fail so catching this on a particular server is not so easy.
the good part is the fact
that this seems to be happening durning a particular time frame.

Thanks,

Bev T.

Beverly ,

If your server is crashing or hanging you might want to
download our ReJuvin8 tool. It's free for the first
15-days. It can keep your site running despite problems.

It's designed to detect problems then restart or reboot.
It can restart ANY programs, services, or processes in
precise order. It doesn't fix bugs but it does give you some
breathing room while you are finding them. ReJuvin8 also
gives you email and telephone alert messages and graphic
status reports. It runs as a Windows Service.

It can also test, alert, and report on URLs running anywhere
on any server on your net connection. So if you use multiple
servers you can use them to watch each other, too.

ReJuvin8 has both an automatic and a custom mode. In the
auto mode you make all the settings by pointing and
clicking. This will handle most applications.

The custom mode uses heartbeat synch files to get absolute
detection that key processes are working. Tools and
instructions are provided to help you do this. We also offer
customization as a contract service.

Charles


Get ReJuvin8 at: <a style='text-decoration: underline;' href="http://www.rejuvin8.net?ac=ch1" target="_blank">http://www.rejuvin8.net?ac=ch1</a>

Also, see it and reviews at
<a style='text-decoration: underline;' href="http://www.hotscripts.com/cgi-bin/search.cgi?bool=AND&query=rejuvin8&catid=all" target="_blank">http://www.hotscripts.com/cgi-bin/search.cgi?bool=AND&query=rejuvin8&catid=all</a>




"Beverly Treawell" <prgmrblu DeleteThis @hotmail.com> wrote in message
news:%23cLsA1I$DHA.2432@TK2MSFTNGP11.phx.gbl...
 > Hi folks,
 >
 > I'm experiancing some slow downs and timeouts on my web servers.
 > The majority of the time we get periods of slow returning pages with
periods
 > of excessive wait times
 > for requests which throw ASP 0113 and ASP 0147 errors. The thing is the
 > servers do not even
 > appear to be breathing hard. The only thing I am seeing is ASP requests
 > queuing.
 > Does not appear to be load induced.
 >
 > No excessive CPU utilization (indicitive of looping)
 > No excessive memory utilization.
 >
 > I do run the windbg user mode dumps on the servers but have not
 > recieved any user dumps based on exception errors.
 >
 > We are also not able to specifically match the timing to any lock or
 > deadlocks
 > against the DB server.
 >
 > Our monitor sees that the server is returning ASP pages in the 20 second
 > time frame a(a very basic test page)
 > and takes the load balanced server out-of-service. After a few seconds
the
 > connections are dropped and it puts
 > the server back in. No IIS restart, No server reboot, No restarting of
the
 > custom com+ applications is required.
 > Adding more servers has not alleviated the problem....
 >
 > Can I use IIS state in as automated method to capture the wait state? We
 > have 12 servers that
 > may or may not fail so catching this on a particular server is not so
easy.
 > the good part is the fact
 > that this seems to be happening durning a particular time frame.
 >
 > Thanks,
 >
 > Bev T.
 >
 >
 ><!-- ~MESSAGE_AFTER~ -->