I have a Bordermanager 3.7 server running on NW5.1, SP6, BM37SP2, with

Surfcontrol. Netware volumes are all standard volumes.

We had a cron job to unload proxy -cc about once a week.

This had been working ok, until we rebuilt and reinstalled the BM
server
on new hardware this past Summer.

Now when the cron job runs if often hangs at the proxy console.

After the server is restarted, often we find we are unable to access
our
own website because the sys:\etc\hosts file has become corrupt. Once
the
hosts file is restored web site access resumes.

What can we do to troubleshoot this problem to improve the reliability
of
the proxy re-cycling process. Thank you.

John B.