Thursday, November 29, 2018

RPC traffic breaks when HSRP active fails over

I am setting up a new project for an industrial scada client and they have a distributed pair of servers essentially at different sites. The two servers communicate using RPC calls and talk directly via IP on some high numbered TCP port (40xxx). Anyway I am testing the network to ensure that when HSRP fails over to the secondary router that the communication doesnt fail with it, and so far all my tests indicate that this isn't the case. The 2 servers are synchronized and can withstand a delay of up to 200ms between them however HSRP by default is slower than that and even though I have tuned the timers to 200ms/600ms (hello/hold) I am still seeing a loss of synchronization and all the RPC traffic fail at the same time in the server logs. Is there anyway to preserve RPC over a change in HSRP state on a cisco router? I am using 3850's I believe the RPC traffic maintains the synchronization.



No comments:

Post a Comment