Nick Knize

> You may need to run it 2-3 times to reproduce the behavior. Is there not a reproducible `-Dtests.seed`?

Why are PRs getting merged before passing gradle check?

Is this a meta issue? Is it done? Can this be closed? Or is there a backport needed somewhere?

> With segrep, every replica can search accurate point-in-time snapshots of the index Noting here there's a [Point in Time API design issue](https://github.com/opensearch-project/OpenSearch/issues/1147) and [corresponding PR](https://github.com/opensearch-project/OpenSearch/pull/2745) to incrementally improve the...

> use Lucene's `SearcherLifetimeManager` class to efficiently hold open any past point-in-time readers that are still leased by running scan/scroll requests This sounds like a better mechanism to use within...

> Do Primaries handle searching too? For now, yes. This was discussed for "priority queries" (e.g., alerting use cases) where the primary would have documents available for search before replicas....

/cc @vigyasharma to join the discussion!

I'm actually quite confused how this is a different issue than https://github.com/opensearch-project/job-scheduler/issues/147 where there was strong opposition to moving it to the core. Tangentially, it looks like there is an...

> From what I understand, the main opposition to moving Job Scheduler in previous discussions was due to the mandated installation required by making Job Scheduler a module. I see....

I'll dig in a little further in a follow-up but wanted to throw one quick thought: > Is FST the right data structure for checking existence/absence. BloomFilter format can be...