I was writing my thesis chapter about ejb clustering and, when describing
the replicated jndi tree I had this question:

I have a client. He wants to access a "well-known" EJB. The jndi tree is all
over the cluster, so I can use whatever node to do the jndi initial request.
But.. the client does not explicitely get a node assigned to do the
requests, does it? Does it multicast the initial request? Or gets a jndi
server list by multicasting a "request server list" packet, and then chooses
one?

I think this behaviour is determined by the jndi.properties file, in
particular the
org.jnp.interfaces.NamingContextFactory method/object is the responsible for
it... right?
But... if multicast is used, that means the client must always be on the
same LAN segment where the cluster nodes are in... unless we explicitly
provide a servername where to it should connect.
If there is the need to explicilty provide a servername to do jndi queries,
we'll not have HA or LB in that client interactions with the cluster... or
can we?

Thank you

-- 
Joao Clemente - jpcl at rnl ist utl pt



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
JBoss-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-user

Reply via email to