Koozali.org: home of the SME Server

Legacy Forums => General Discussion (Legacy) => Topic started by: Francis Van Peborgh on April 02, 2003, 05:24:49 PM

Title: CRON Message about FETCHMAIL
Post by: Francis Van Peborgh on April 02, 2003, 05:24:49 PM
Hi,

Since I upgraded to 5.6, I regularly get messages to ROOT with the following contents :

fetchmail: another foreground fetchmail is running at 10443.

The port the message reports is each time different, some days there are two or three such messages.

Can anyone help me ?

Thanks,

Francis
Title: Re: CRON Message about FETCHMAIL
Post by: David Clarke on April 02, 2003, 10:41:24 PM
This seems to be when a cron-triggered fetchmail overlaps the previous one.

It occurs often on my home 'test' machine, which is only a P75, and which often takes a minute or more to process each mail ( as it is running a duplicate of a production system that does virus and spam-scanning). The email settings try and fetch new mail every five minutes, but this only works when I have activated the dial-up link to the ISP, so the first time a connection is made the mail message from cron appears if all of the accumulated mail cannot be processed in five minutes.

Two solutions occur to me (neither is implemented, I just tolerate the messages):

1. Modify the /etc/startmail script to test for the existence of the fetchmail lockfile and exit silently rather than trying to do another fetchmail.

2. Manually change ( or edit the web-interface scripts to offer more options ) /etc/crontab to run fetchmail every 10 or 15 minutes.

David Clarke.
Trent Micro Systems, Nottingham, England.
Title: Re: CRON Message about FETCHMAIL
Post by: Scott Smith on April 02, 2003, 11:16:30 PM
Your option #2 is not necessary. Simply change the mail collection frequency via the web manager.

For my purposes, I modified the /etc/startmail template to redirect the output of /etc/fetchmail to /dev/null. I want mail collection every 5 minutes. If fetchmail happens to still be working on the prior invocation, that's fine. I just don't want to know about it every time it happens. Redirecting output to /dev/null works.

Why? You get the message because cron takes any output to stdout/stderr that a process generates and emails it to admin. When fetchmail warns that another session is running, that message is what causes the email. Redirect the output of fetchmail to /dev/null -- no more emails.