This is a very, very common complaint I hear from developers, testers, and production support.

I find there are a couple reasons that this can happen:

1. The flow has faulted sometime after the wait activity, but never dehydrated again, so it appears to be “stuck” on the wait activity. Put in some more waits to force dehydration soon after your “problematic” wait to ensure the wait really isn’t your issue.

2. In 10.1.3, the default wait in JDev is set to 1 day; please, be sure to set your wait times properly.

3. The flow really is just waiting too long. Oracle advertises the wait time as a “minimum”. You may want to increase the number of threads you have running for your scheduler.

• Stop your application server
• Take a backup copy of the following file:
  \bpel\domains\[domain-name]\config\resources-quartz.properties
• Change the resources-quartz.properties file:
  com.oracle.bpel.expirationAgent.threadCount is defaulted to 10. You can increase this to 50 or 100, or more depending on your load.
• Restart the application server