CODEWITHSUNDEEP
tomcatjbossluceneamazon-ec2hibernate-search
Shadowman
Shadowman

Reputation: 12039

Hibernate Search on EC2?

We're in the process of finalizing one of our e-commerce applications that will be deployed onto Amazon EC2. The app is built using JBoss Seam, and makes heavy use of Hibernate/JPA and Hibernate Search. We plan to run with multiple app servers in a cluster, and are currently targeting Tomcat 7, although we're also investigating migrating the codebase over to JBoss AS 7.1.1.

In thinking about our cluster setup, one of the big questions we have is around how the Hibernate Search indexing will work. How do we reliably assure that a search request going to app server #1 will return the same results as app server #2? Is there a way to keep the indexes in sync on 2 different app servers? I've glanced at the documentation for JMS master/slave setup, but we don't really have a master/slave architecture. We're looking at more master/master, as either app server could be the one updating an index.

Has anyone come across any useful guides, tips, or documentation that explain how to best run Hibernate Search in an environment such as EC2? We're eager to hear any recommendations you can provide; performance, indexing, high-availability, clustering, etc. Anything you can recommend is GREATLY appreciated!

Upvotes: 2

Views: 918

Answers (1)

Guilherme Torres Castro
Guilherme Torres Castro

Reputation: 15350

I think it's not possible using a "master/master architecture". But using master/slave you can keep the indexes in sync and should worked for you.

Take a look here

Tomcat does not support JMS, however you can use another JMS provider (like OpenJMS) or migrate to tomEE or JBoss.

About, amazon i found this:

Upvotes: 2

Related Questions