We have a load balanced setup using two physical servers and one server behind a firewall used for editing. The EPiServer version is 4.62, running on .NET 1.1.
The strange behaviour that we have seen sometimes is that the external servers begin to request so much data from the oracle db that it chokes the 100Mbit ethernet connecting the servers to the db. There's one query in particular that amounts to almost all the traffic, and mostly from one of the servers. We've discovered that the behaviour is triggered from the editor machine. There's also a difference in the clock settings on the two external servers, the difference is 5 and 7 minutes compared to the (correct) time on the editor machine. There's cache listening set up between these servers.
Does anyone know if the time differences of the servers can be responsible for this behaviour?
We have a load balanced setup using two physical servers and one server behind a firewall used for editing. The EPiServer version is 4.62, running on .NET 1.1.
The strange behaviour that we have seen sometimes is that the external servers begin to request so much data from the oracle db that it chokes the 100Mbit ethernet connecting the servers to the db. There's one query in particular that amounts to almost all the traffic, and mostly from one of the servers. We've discovered that the behaviour is triggered from the editor machine. There's also a difference in the clock settings on the two external servers, the difference is 5 and 7 minutes compared to the (correct) time on the editor machine. There's cache listening set up between these servers.
Does anyone know if the time differences of the servers can be responsible for this behaviour?