I'm looking at the documentation for outscaling search. In particular, the diagram here is interesting.

It's quite possible my understanding is wrong, but it seems to me that you could simply configure the Query Engine settings on your various CM servers all to point to the URL of your "master" search host.

Have I understood the architecture correctly? What added value is there in having slave servers? There's a hint in the documentation for configuring slaves that it might have to do with replication. Do we then have fully automatic replication of the index?

2 Answers 2


The primary reason for any out-scaled search setup (master-slave) is to distribute the search load across servers to provide a timely responsive to any search query being executed in Solr. Outscaled search will be very useful in large volume implementations

The index is continuously updated by the master and it is replicated on the slave servers automatically. Since all the servers have the index, the queries are executed against them and hence the sharing of search load makes the system more responsive.

There is an automatic replication which happens from the master servers to the slave servers whenever an item is created/changed/deleted in Tridion. On the slave servers, you need to provide the

indexer engine host url property = master server url:8983

With Tridion 2013 though, there is an issue with indexing (in master slave configuration) which is documented here with the solution as well Tridion KB


Yes, your understanding of architecture is correct.

Solr engine takes care of replication by itself. The replication always happens from master to slave. Reference: Index replication

The replication process takes the indexed set from master and replicates on all the slaves.

If slave also starts indexing then during replication, it will overwrite the items indexed by other servers. That is why we need slaves.

On a side note, there is race condition that happens between tridion core and search indexer due to which items created on slave servers are not picked up by master server. To avoid this issue polling interval needs to be set for search indexer service on master server.

  • Modified answer to provide more details on replication and how it works.
    – Hiren Kaku
    Jun 1, 2015 at 17:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.