0

We are currently evaluating infinispan server as a cache server for our application running in JBOSS.

We are planning to use hotrod endpoints to access infinispan cache from our application nodes.

We have around 300+ test environments, what would be right strategy for infinispan server installation.?

I do not want to create 300+ instances of infinispan servers, which would lead to management mess.

Is there a way we can run one infinispan instance exposing 30 hotrod endpoints to server 30 test environments ? Thereby, we would need to setup only 10 instances of infinispan server where each infinispan instance supporting 30 environments/application server.

Any assistance on the approach would be highly appreciated.

Thanks Santhosh S

rockycres
  • 71
  • 8

1 Answers1

0

Since this is a test environment, I assume you are not really interested in hard segregation of data for each one. Why not just have a single Hot Rod endpoint, and have 1 (or more) cache(s) for each of your environments ?

Tristan Tarrant
  • 1,299
  • 6
  • 6
  • We want to use infinispan as a second level cache provider for hibernate. If we create multiple cache containers and use single Hot Rod endpoint, how does hibernate cache entities would be segregated ? From looking at the documentation, I couldn'd figure a way to separate hibernate cache between multiple test environments if I use singe HotRod end point. How does hibernate cache named inside Infinispan for lookup ? – rockycres Mar 26 '20 at 04:40
  • Unfortunately the Infinispan 2LC provider for Hibernate only supports embedded cache managers – Tristan Tarrant Mar 30 '20 at 08:12
  • Infinispan segregation of data by using multiple cache container is not possible in a single infinispan instance for serving multiple environment is not possible if the application uses Infinispan as 2LC provider for hibernate.. can you confirm if this is correct ? So what is way forward ? Looks like running one infinispan instance per test env is the only way. – rockycres Apr 01 '20 at 00:55