Skip to main content
Participant
June 2, 2007
Question

Jrun.exe consuming up 1GB of RAM

  • June 2, 2007
  • 3 replies
  • 600 views
I am in a shared hosting environment with approx. 2k of customer sites running on a web farm. ( windows 2003 standard , IIS 6.0, cold fusion MX 7)

I have noticed lately that the jrun.exe process which controls the cf application server has been consuming up to 1gb of ram on our servers. This is a gross abnormality and obviously indicates a possible leak somewhere in this process. I have been using the iis diagnostic tool but as it is the jrun process that is taking the ram, I cant get to the root of this issue.

How can I get what application is causing this issue or what is running within jrun real time. I have had much easier time identifying an issue within a worker process ( IIS application pool) but this is different. I have not dealt to much with cold fusion issues. Can anyone shed some light here.
    This topic has been closed for replies.

    3 replies

    BKBK
    Community Expert
    Community Expert
    June 5, 2007
    What would be a method I could identify the site which is making this call nside of the jrun process?
    Study the log files. Does anything jump out?

    BKBK
    Community Expert
    Community Expert
    June 3, 2007
    > Jrun.exe consuming up 1GB of RAM

    Possibly infinite loops. A common example is infinite redirection by a cflocation tag in Application.cfm or in Application.cfc.

    Participant
    June 5, 2007
    What would be a method I could identify the site which is making this call nside of the jrun process? This is hwere I am stuck, I cant isolate what application in a specific site is causing the problem.
    Participant
    June 2, 2007
    Also I suspect that we are way off on max simultaneous requests. I have inherited this cf farm and was built by others. They have set the max requests to 1000. From looking at another thread on this topic, this looks to be way off.

    The hardware is dell 1950 4gb RAM. The customers per IIS farm varies from 500 to 2k. This is an active passive design. Failover is handled via a load balancer.