[jboss-dev-forums] [Design of JBossCache] - Re: FLUSH and state transfer during high invocation load

[manik.surtani@jboss.com]

"vblagojevic at jboss.com" wrote : This is not a JGroups bug but a valid property related to FIFO message delivery. This unfortunately includes STOP_FLUSH_OK messages and thus FLUSH cannot complete gracefully at A until CP returns.

Isn't this a generic JGroups issue and not something that is specific to JBoss Cache?

Regardless of the application, if someone's using the RpcDispatcher in synchronous mode, any RPC call has the potential of getting delayed up the stack and FLUSH preventing that call from returning.

I'm happy with fixing it in JBC - especially for 2.0.0 since any fix in FLUSH would involve another JGroups release - but just wondering if this is something other apps using FLUSH + RpcDispatcher would experience as well.


View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4068468#4068468

Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4068468