My company connects to a major company's (I'll call them Acme Inc.) online
catalog - that's how we place orders with them. Most of my users are on
Windows NT 4 Workstation 4 with SP6a, or XP with SP1, all using IE6 with SP1.
A few weekends ago Acme made changes to their internet server, with a
major upgrade to Apache, and adding TomCat (replacing an Oracle connector) to
connect to an Oracle DB. They also moved to a Linux machine, from Solaris
(UNIX).
Following that weekend, our users with IE6 SP1 starting having major
problems when connected to Acme. Connecting to Acme was not a problem. While
adding items (usually after 2 or 3) to the Acme shopping cart, IE6 would lock
up. If the user got to the point of submitting the order, IE would lock up
when the "checkout" button was clicked. The cursor was not locked, and would
turn into a hand when hovering over a hyperlink. There were no error messages
of any kind, and the "back" button worked, but at this point the user was
hosed in terms of placing an order.
We have a small percentage of users with XP SP2/IE6 SP2 (which we're
rolling out slowly), and none of them have this problem, so we are looking at
SP1 as the issue, since our SP2 users don't have this issue.
Acme claims that no other customers are having this problem. One Acme
techie said that adding items to their cart executes Javascript and causes
the page to reload, so Javascript issues should be looked into. Acme is doing
some testing on their end, including building an XP SP1 machine, to see if
they can replicate the problem.
Acme also tried to monitor their web server while we connected (during
which time our SP1 browser locked up), but they didn't see the offending
transaction - apparently our browser locked up without sending them any data.
Does anyone have any suggestions on how to determine what's happening on
the browser when it locks up? (By the way, upgrading the older configurations
to SP2
is unfortunately not an option). I looked at the IE6 SP1 settings, and I
don't see anything specific to Javascipt.
One last note - we use online catalogs for other companies, and we have
had no problems with them, just Acme, so we know it's something they did. Any
suggestions would be greatly appreciated!

"Padrino" <Padrino@discussions.microsoft.com> wrote in message
news:F6C718A6-9745-487C-9072-74F040158870@microsoft.com
....
> Does anyone have any suggestions on how to determine what's happening on
> the browser when it locks up?

Have you tried running a script debugger to see what happens before it hangs?
An inferior but still useful substitute would be to activate prompting for scripts,
ActiveX, etc. just to see what the pattern there is. Running FileMon and
RegMon to trace file system and registry accesses leading up to the hang
and then comparing the equivalent traces of what happens when a transaction
works would be useful too. Do you even know if the symptom is a hang
or a loop? Activating maximal column statistics in Task Manager could help there
and again, depending on how loose the loop is, the other tools could help
clarify what it was doing then.

Finally, are the IE6sp1 systems fully updated with the latest cumulative
update for IE or is some lower level of maintenance present?


Good luck

Robert Aldwinckle
---

Thanks for your informative reply. I'll do some research on script debugging,
as I never have dealt with it. Meanwhile, if you have a suggestion for a
debugger, it would be appreciated. In any case, Happy New Year!

"Robert Aldwinckle" wrote:

> "Padrino" <Padrino@discussions.microsoft.com> wrote in message
> news:F6C718A6-9745-487C-9072-74F040158870@microsoft.com
> ....
> > Does anyone have any suggestions on how to determine what's happening on
> > the browser when it locks up?
>
> Have you tried running a script debugger to see what happens before it hangs?
> An inferior but still useful substitute would be to activate prompting for scripts,
> ActiveX, etc. just to see what the pattern there is. Running FileMon and
> RegMon to trace file system and registry accesses leading up to the hang
> and then comparing the equivalent traces of what happens when a transaction
> works would be useful too. Do you even know if the symptom is a hang
> or a loop? Activating maximal column statistics in Task Manager could help there
> and again, depending on how loose the loop is, the other tools could help
> clarify what it was doing then.
>
> Finally, are the IE6sp1 systems fully updated with the latest cumulative
> update for IE or is some lower level of maintenance present?
>
>
> Good luck
>
> Robert Aldwinckle
> ---
>
>
>