[jorju]

Thanks for your answer,

Are there any logs available that may explicit the fact that when the stopped server returns to live it copies the clustered HTTP session ?

Anyway, are there any logs (maybe additional to those in server.log) that monitorize how the cluster handles a fail back (http session copy, jndi, etc.)?

[derry]

Hello!

I put code in HEAD that should get the session to stick on the node it was reactivated on (server2 in your scenario) even when server1 comes back to life... This should solve your problems...

CU
Thomas

[jakefear]

Is there any hope of this patch being applied to the 3.0.x series? I need this feature as well, and I am hoping I will not need to wait for a stable 3.2 or later version.

Thanks,
Jake

[derry]

Hello!

The code is in Branch_3_0 and Branch_3_2 now. Feel free to test...

CU
Thomas