Problem with persistence using IceStorm

jeangluis

I’m new working with ICE.
Recently I make this test.
I have two programs one to publish data (Publisher) and other to receive data (Subscript) both use IceStorm. IceStorm use Freeze data base for persistence propose, (Freeze.DbEnv.IceStorm.DbHome=db) In one PC (PC1) we have IceStorm with one Virtual IP. In other PC (PC2) we have the Publisher and Subscript applications.
In this state all work fine, Publish application send data and Subscript application receive OK. Now we turn OFF IceStorm, copy the configuration of IceStorm to other PC (PC3) and turn off PC1. PC3 have the same Virtual IP of PC1. When we turn On IceStorm, the Publish application DELAY S a big time (30 sec - 1 min) for reconnecting and send data again. How can I reduce this reconnection delay?
Exist some way for handle some timeout exception?
Thanks for your attention

matthew

What exactly is the publisher doing when it delays? What timeouts do you have configured? I suggest enabling network tracing (Ice.Trace.Network=2 or 3) in your publisher to determine the issue.

What precisely are you trying to do anyway? IceStorm has full support for fault tolerance built in, so if you are trying to do some roll-it-yourself fault tolerance using virtual IPs, I would recommend not doing that and instead use the built in solution.

jeangluis

Thanks for your answer!!!! I`ll try using timeout properties