What exactly are you doing to check your website every two minutes? If I'm
not mistaken, CF Executive should be monitoring unresponsive threads and
restart the CF Application server as needed. There's a settings within the
CF Administrator to adjust the count of number of unresponsive threads at
which you have the service restarted.
I'm running CF 4.01 Pro as well, and I also cycle the server nightly to
recapure leaked memory (just the application server - takes about five
seconds). For the most part, we have no problems with stability. Every
once in a great while, I'll notice the server becomes unresponsive, but it
always comes back, proabably after CF Exec has restarted it.
Also, every once in a blue moon I'll try to cycle IIS to overcome a file
locking problem, and the only time I'm able to do this is during times of
low server load. When the web server is busy, it usually stops IIS after
about 30 seconds, but is left in a state that won't allow IIS to restart.
Are you running SQL Server on the same machine as IIS and CF?
Jim
----- Original Message -----
From: "Bruce Ford" <[EMAIL PROTECTED]>
To: "CF-Server" <[EMAIL PROTECTED]>
Sent: Monday, November 13, 2000 12:41 AM
Subject: Stability problem with CF 4.0.1 Prof.
> I have an ongoing problem with the stability of our Cold Fusion Server:
>
> Version details:
> CF Server: 4.01 Professional.
> OS: NT 4 SP 5
> Web Server: IIS 4
> Database: SQL Server 7 SP 2 via ODBC
> Hardware: Dual PIII 500 MHz / 512 MB RAM
>
> At present the web site is checked every 2 minutes and if a problem is
> detected CF server is restarted. If that does not fix things CF and IIS
> are restarted 2 minutes later. Eventually, after 2 more attempts at
> correcting via restarts of CF and IIS (sometimes "net stop" does not work
> and I need to kill the inetinfo process), I escalate to a restart of the
> server. That usually fixes it!
>
> The problem seems unrelated to hardware, as it occurs on both the main and
> standby servers, I had to switch to the standby server when corruption
> caused WINSRV.DLL to disappear and the main server blue-screened on
restart.
>
> Cold Fusion server is regularly restarted at 03:40am every day.
>
> The failure pattern correlates very loosely with usage but there seems
> little reason for the observed peaks in the failure pattern for
00:00-00:59
> and 02:00-02:59 hours local time. Interestingly in October I have no
> failures on Saturdays or between 03:00 and 06:59.
>
> I am concerned that it is associated with the CF code as restarting CF
> server seems to fix these errors on occasion but I am running out of
ideas.
>
> I have so far fixed constructs where CFSET was not setting a variable eg.
> <CFSET StructInsert(struct,key,var)> and ensured all references to
> application and session scoped variables are within <CFLOCK> ... </CFLOCK>
> using NAME=#application.applicationName# and NAME=#session.sessionID# as
> appropriate. Although these have helped stability to some extent the
> failures continue.
>
> The last item I am correcting is the use of nested, ungrouped <CFOUTPUT>
> tags due to the use of CFINCLUDEs and CFMODULEs. I have no idea if these
> are significant or not but as the CF documentation says "You cannot nest
> CFOUTPUT tags." I am hopeful.
>
> I have put in process an upgrade to CF 4.5.1 but that may be some time in
> coming and if the problem is within the CF code I am unsure if it will
help.
>
> Is there anything else I should be investigating?
>
> Bruce Ford [EMAIL PROTECTED]
> Online Teaching Systems Coordinator
> Teaching and Learning Support Services Ph: +61 7 3864 9773
> Queensland University of Technology
------------------------------------------------------------------------------
To unsubscribe, send a message to [EMAIL PROTECTED] with
'unsubscribe' in the body or visit the list page at www.houseoffusion.com
