I have a .NET Windows app (framework 1.1) that makes calls to a SQL Server on
the same subnet as our systems, but in a different domain, using SQL
security. I open a database connection once when the application starts, and
never explicitly close it.
A database call is on a 5 second timer to poll for status stored in the
database. The call works fine most of the time, but periodically the
execution will result in a call to an apparently random IP that ZoneAlarm
attributes to the application calling on the DNS port. None of the IP
addresses Zone Alarm displays are in our domain and the domain is not
forwarding DNS requests to them either. So why would my application decide
after several minutes of inactivity to try to contact a DNS server, and where
would it be getting the IP addresses of servers it shouldn't know about?
If I do not tell ZoneAlarm to let the traffic through then the SQL call will
time out, resulting in an exception thrown by the timer tick.
This behavior just began in the past couple days and there have been no
major code changes beyond an update to Janus Systems controls I use.
Does this behavior ring any bells with anyone?
