Daniel Widdis

> 1. Are there other AD specific performance metrics we could be considering? For example cache hits, training time etc. Perhaps someone from AD plugin could help compile the list?...

> First, you wrote "The [recommended instance size](https://docs.aws.amazon.com/opensearch-service/latest/developerguide/managedomains-dedicatedmasternodes.html) for a 30-instance cluster similar to the baseline is a c5.2xlarge.search with 16 GiB memory and the same vCPUs as the baseline....

A quick clarification on the new cluster and memory allocation requirements. The blog post we're replicating for the first phase of this comparison identified the memory requirements for the models....

After learning many lessons about initialization order, subnets, stale builds, and persnickety operating system incompatibilities, I've got a small cluster running enough to begin replicating the HCAD test in [this...

Full-scale performance test. Cluster from original HCAD "1 million in 1 minute" was 36 R5.2xlarge data nodes. Same size/cost cluster for this test: 30 r5.2xlarge data nodes. 1 r5.12xlarge extension...

Another run: 1 million results. Start: 20:25:16.238 Finish: 20:26:06.965 50.727 seconds.

> @dbwiddis are these data-points for running extensions on a remote node or out of process? These were quick snapshot points of one metric (run time) for a remote node...

We don't have a create components method, and instead are directly implementing the required components. Some are associated with other extension points. The following are part of `createComponents` and are...

The following plugins may not be relevant for extensions. They are documented here for completeness: - ClusterPlugin - [ ] createAllocationDeciders - [ ] getShardsAllocators - [ ] getExistingShardsAllocators -...

> why are [#315 (comment)](https://github.com/opensearch-project/opensearch-sdk-java/issues/315#issuecomment-1377742853) not relevant for extensions? @minalsha I used the qualifier "may not be relevant". I'd appreciate another set of eyes on this (@saratvemulapalli). - ClusterPlugin directly...