We have an application base on SQL, this application is provided by server throught a private WAN link with maximun capacity of 2Mbps current usages is 800Kbps. We are experimenting problem with application time out and we have solved it adjusting the MTU parameters in each station to 1460 bytes. The connection to SQL servers start in our WS6509 using ethernet which use as MTU to 1500 by default, but so far from this L3 switch is a cloud for me, because this equipments are other owner's administrators.

Here is a traceroute

C:\Documents and Settings\xxxx>tracert 192.168.239.194

Tracing route to 192.168.239.194 over a maximum of 30 hops

1 <1 ms <1 ms <1 ms 10.32.6.200

2 <1 ms <1 ms <1 ms 10.32.0.101

3 21 ms 23 ms 24 ms 192.168.252.1

4 13 ms 13 ms 13 ms 192.168.239.2

5 13 ms 13 ms 13 ms 192.168.239.194

The connectivity was working fine recentetly and we are frustated to have to solve in this way this problem because I know that is uncommon solutions because the degradation in efficiency for TCP connections.

My question for you is that if changing the MTU parameters in each station (which is not scalable) should be the solution or there is (are) other ones that I can follow.

Thanks in advanced