You would only get this message if you have more than one fetchmail instance running. In normal operation, this would indicate that fetchmail is taking longer than the defined cycle time to complete, therefore another fetchmail is firing off and creating the error condition. There are several options for handling this.
You can increase the cycle time. For example, if you have the system checking every 5 minutes, change it to every 15 and see if the problem goes away.
You can modify the /etc/startmail template to direct all output from /etc/fetchmail to /dev/null. You're getting the message because cron sends everything output to the console by a background task to the root user. Redirecting the output of the task by adding >/dev/null 2>&1 to the command line will solve the problem of the messages appearing -- although the collisions may still occur.
Modify the /etc/startmail script (via a custom template) to first check to see if fetchmail is already running. Either have it terminate and wait for the next cycle, or sleep 60 and try again. I'd have it terminate, but it's your choice.
Scott
6 Replies
1975 Views