Copy link to clipboard
Copied
Hi All,
For the past few weeks, the ColdFusion Server will randomly just hang... where a page request will just keep spinning... restarting the server has cleared it up... but we're having to do that more frequently.
Can you give me some insights into this problem or how to debug it on CF11 Server?
Version: ColdFusion 11 Update 7
IIS: Windows Server 2012R2 8.5.9600.16384
Database: SQL SERVER 2008 R2
I suspect the problem may be related to the number of workers/processes running at a given time? How do I increase this from the default installation?
Thanks,
John
Copy link to clipboard
Copied
Hi John,
Any errors or warnings in log files? These logs can sometimes provide a clue:
ColdFusion11\config\wsconfig\1\isapi_redirect.log
ColdFusion11\cfusion\logs\coldfusion-out.log or coldfusion-error.log
Workers and processes that sounds like tomcat tuning. Perhaps this blog reference will help:
ColdFusion 11 IIS Connector Tuning — Adobe ColdFusion Blog
HTH, Carl.
Copy link to clipboard
Copied
We're having a similar problem (at least to the "hang" and "spinning" level) with CF10.12. Capturing the packets between the webserver and our Oracle database shows a continual stream of packets asking Oracle for the last byte of a clob's substring, Oracle returning it, and CF re-requesting the same byte, etc., for days on end. The packets leading up to this point show that however many bytes CF is asking for in its (presumed) restart request, Oracle is sending them all back but CF is only counting half the returned bytes. For example, if CF re-asks for the last 32 bytes, Oracle returns them all but then CF turns around and asks for the last 16 on down to 1 where CF's effective integer divide by 2 results in it thinking it received zero bytes and so re-asks for that last one. This, then, burns up the network with an incredible multiple-days-long exchange of tiny "give me that last byte!", "Here you go -- again" exchanges.
Is this a known problem and has it been fixed beyond CF10.12?