Macromedia application server is unreachable

Report · Jun 09, 2008

Hi,
I have a W2K3 server running CFMX7 with report server and verity. Starting on Friday, the server would display an error of:

Server Error

Either the Macromedia application server is unreachable or it does not have a mapping to process this request.

anytime that you tried to hit a CF page. After a couple of minutes, your pages were accessible again. It happens after a restart of the server or service and also intermintently. It also happens at times that aren't particularly high use.

I ended up reloading CF on Sunday, report server, search functions, Updater 1 and 2 and the DST patch and then copied over the old \LIB folder. That helped out alot, but it still periodically goes down for a couple of minutes. Is there anything I can do or check? I've never seen this before. I don't see that any CF process is consuming that many resources (nothing out of the norm).
Thank you,
Nathaniel

Report · Jun 09, 2008

Check for errors in the cfusionmx7/logs/application.log and cfusionmx7/logs/exception.log around the time the server goes down. Also check if there are any hs_err_pid*.log files in runtime/bin directory. Do the timestamps match outages? If so, review them and see if they point to something obvious. Recently I had a customer who had these and it pointed to his database client software was crashing.

I would also download CFMX702 from http://www.adobe.com/support/coldfusion/downloads.html page bottom. It is 702 so you would not need to load the updaters. Insert your license and you are all set.

Report · Jun 11, 2008

I can't speak to the intermittent problem, but sometimes when the server is starting up a race condition seems to occur. IIS will start, ColdFusion will start, but the connector fails to start correctly.

If you reboot your server and you get that error, simply restart the World Wide Web Publishing Service and life should be golden.

I've had this error on all flavors of CF6, 7 and even 8. Recreating the connector will fix the problem too, but that is excessive and still won't fix it long term.

Report · Jun 16, 2008

Thanks for all of the help. After a lot of looking and trending, it turned out that the SQL backup running on the backend server was getting hung up. Once we tweaked the maintence plan, we stopped having outages.