Hi All,
I am a newbie to Tuscany. I have requirement like in my EAR application,
as part of application workflow, it is required to invoke some services
(which is also part of same EAR currently). We want to expose these services
as SOA based services, so that these services can be invoked by any external
client and also through an EAR application.
I have few concerns related to clustering and load balancing behavior
for
services hosted on Tuscany Runtime.
Consider J2EE application scenario, say, I have set up 2 node Weblogic
cluster,
and both the nodes are up and running. So the client request may reach to
any
of the nodes and it will be served by one of the node. In case, if one of
the
node goes down, then also client request will be handled transparently by
the
other running node.
I was looking at the documentations on the tuscany site but could not
get
information about clustering and load-balancing behavior.
So my questions are -
1) Can we host tuscany in any Appserver?
site documentation says that it can be done but please point me to the
document
that explains how to achieve this.
2) Assuming it is possible to host tuscany in Appserver say weblogic, then
can
I host same services (same SCA application) in 2 different nodes of a
cluster, so that
same service is available on both the tuscany nodes hosted on respective
appserver
nodes. And even if one appserver node crashes, tuscany services are
available to
serve the request by external clients or by EAR application in other node?
3) If such a clustering behavior is supported, then for a two node cluster,
is load-balancing done using round-robbin or some other algorithm?
4) How to dynamically deploy SCA applications to a already running Tuscany
node/domain?
Your response would be highly appreciated.
Thanks,
Sandesh
