Using Wildfly Infinispan in a stand-alone topology

simplex-software Jan 19, 2017 5:33 AM

Hello,

We have a production architecture consisting in a number of stand-alone Wildfly servers behind a load-balancer but without cluster. The HTTP requests are distributed through the load-balancer but there is no session replication, etc.Is it worth using infinispan in this case ? I mean, would it be any way to have the distributed cache such that any server has access to the information stored by any other, without having to configure our servers in a cluster ? Probably not but just asking ...

Many thanks in advance,

Nicolas

1. Re: Using Wildfly Infinispan in a stand-alone topology

pferraro Jan 24, 2017 10:50 PM (in response to simplex-software)

I think you have some inflated notion of what a "cluster" is. A distributed cache, where any server has access to the information stored by another, *is* a cluster. To enable distributed web sessions, all you need to do is:
1. ensure your web application define <distributable/> in their web.xml
2. start your servers using the standalone-ha.xml instead of standalone.xml (or standalone-full-ha.xml instead of standalone-full.xml)
In most cases, the default configuration works just fine.
Actions