We have a NAS application which can be accessed by both HTTP and HTTPS connections.

The issue we are facing is that the tcp server instance that initiates the HTTP access dies after a few hours of inactivity(the NAS application was kept idle for around 10 hours). However, the tcpserver that launches HTTPS connections remained running.So the connection failed through HTTP access but the connection succeeded through HTTPS access.

The converse has also been observed where the tcpserver for HTTPS is dead but tcpserver for HTTP is running.

The version of tcpserver that we are using here is ucspi-tcp-0.88.tar.gz.

I have the following queries:

a)Has anyone faced the same problem before?
b)Is it an issue with this package version of tcpserver ?
c)Will upgrading to newer versions of tcpserver bail me out of the issue?
d)Also, could anyone let me know what are the possible causes for the tcpserver to die ?

Is the tcp server actually dying or is http/https just refusing any more connections ?

It should listen interface, if it is alive. What does "netstat -ntulp" give you?

Actually, this is a problem reported by one of our clients. The bug report states that the tcp server actually dies.

We were not able to reproduce this issue. But we are trying to figure out if the issue has been faced by any one else. I would post the logs once I myself recreate it.

If anyone has faced this issue earlier or if anyone could suggest some alternatives, it would help me out a lot.

Also, I would like to know the various triggers that can kill the tcp server.