[BUG] Status code 500 for hybrid queries with preference=_local flag

[ksingh17i]

What is the bug?

We are getting status 500 on open search dashboard when running hybrid queries with preference=_local and request_cache=true flags. The initial investigation shows the error occurs if the index has 1 shard.

How can one reproduce the bug?

1. Create one index with only one shard
2. Populate the data
3. Run hybrid query with the search pipeline and with preference=_local and request_cache=true flags

What is the expected behavior?

The query should return the matching result with status code 200

What is your host/environment?

AWS Managed OpenSearch 2.11

Do you have any screenshots?

Please find the screenshot attached

Do you have any additional context?

[ksingh17i]

adding @navneet1v in loop, Navneet do we have any ETA for this issue?

[vamshin]

@ksingh17i We are evaluating the effort. Will update the ETA soon.

[ksingh17i]

Hi @vamshin, could you please help prioritizing this and let us know by when can we expect a fix for this.

[vamshin]

@ksingh17i we could not get to prioritize this issue. As a work around, you can create more than 1 shard and use request_cache=true. This should work as expected. Also will you be able to help contribute the fix?

[navneet1v]

@martin-gaievski lets close this issue as the feature is merged.