At this point I am not certain if time or number of connections is the problem, but I have a fully patched fedora 18 server running apache-tomcat 7.0.35 and firebird 2.5.2 (all 64 bit) which starts out just fine, but after some time (sometimes 30 minutes, sometimes 4 hours, sometimes 24+ hours), I get the message:
Unable to complete network request to host "localhost"
Once this occurs, it continues and I have to restart tomcat or reboot the machine.
Firebird uses port 3050 for network connections to the DB and I use tomcat's connection pool for transactions.
The "per_source" parameter has been set in both xinetd.conf (30) and the firebird xinetd file (50).
I don't understand if 30 or 50 is the real limit here, but I think I am exceeding it. The messages file say that the service is being stopped for 10 seconds sometimes and this appears to be the problem.
I am removing the 30 limit in xinetd.conf (back to 10) and increasing the firebird count to 200.
Am I doing the right thing?
Unable to complete network request to host "localhost"
Once this occurs, it continues and I have to restart tomcat or reboot the machine.
Firebird uses port 3050 for network connections to the DB and I use tomcat's connection pool for transactions.
The "per_source" parameter has been set in both xinetd.conf (30) and the firebird xinetd file (50).
I don't understand if 30 or 50 is the real limit here, but I think I am exceeding it. The messages file say that the service is being stopped for 10 seconds sometimes and this appears to be the problem.
I am removing the 30 limit in xinetd.conf (back to 10) and increasing the firebird count to 200.
Am I doing the right thing?
