On 7/12/2013 12:51 PM, Sagar Jadhav wrote:
The reason for going that route is because our application is clustered and if the indexing information is on the filesystem, I am not sure whether that would be replicated. At the same time since its a product it needs to be packaged with the product and also from a proprietary reason we are not allowed to use the filesystem.
Solr can do replication from a master server to slaves. If you implement as SolrCloud, then you would have a clustered solution with no master/slave designations. SolrCloud requires a three server minimum for a robust deployment. The third server can be a wimpy thing that only runs zookeeper.
Putting your index in a DB is just a bad idea. It would be hard to find help with it, and performance would not be good.
Thanks, Shawn
