I was reading the 5th chapter of HFJS and I came across the section which explains Session Migration. Here according to the what is written/depicted
-->when a request is redirected to a different container (a container where the required session for generating a response for the incoming request does not reside,let's say in this case C2 in VM2) , it is the container(C2) who determines that the session resides on another VM (in this case let's say VM1).
-->Now I was wondering how would the the container(C2) in VM2 know about the current location of the required session? Shouldn't the knowledge of this exist at a slightly higher level like at Load Balancing Server as it acts as the intermediary between the different VMs and the client? Or are some of my basic assumptions wrong. IS the container C2 keept abreast of all activities that take place in all the other containers in a distributed environment?
I checked some discussions pertaining to this topic on JavaRanch and one answer stated that sessions exist on all the containers in a distibuted environment and are kept in sync via incessant synchronization efforts, now this is contradictory to what's given in the book which states(atleast I felt so) that session are migrated to a differnt VM on a need basis.
Some help for this confused soul please
Also I don't post that regularly so if I need to take this discussion to another forum ,please do prod me