Versions : 2.11
Describe the issue: We have several Opensearch clusters (>20) with master and data nodes. Now we want to enable ml nodes for vector search capability. Rather we add dedicated ml nodes to each cluster, Can we create 1 dedicated opensearch cluster with ml nodes alone and use that from all existing opensearch clusters for ml work? Basically, we want to reduce the hardware cost as its hard to size ml nodes for each opensearch cluster.
Configuration: