Shutdown API for shutting down nodes in a safe manner

[rramachand21]

Is your feature request related to a problem? Please describe.
When a user wants to remove node out of the cluster, there is no easy way to move the ongoing tasks and index shards to other nodes. Also shards should not be migrated to the node being removed out of the cluster.

Describe the solution you'd like
Provide an API to safely shutdown nodes

[CEHENKLE]

Hello @rramachand21! Can you please fill this out with more information about the enhancement you want to make? We need more information to understand the problem and the enhancement.

Thank you!

[CEHENKLE]

Hello @rramachand21 Could you add more details to this issue? If you aren't able to provide more details by 10/22, I will close the issue - but feel free to reopen it later when you've got more information.

Thanks,
/C

[CEHENKLE]

@rramachand21 I've resolved this issue, but please feel free to re-open it and provide more information.

Thanks,
/C

[rramachand21]

@CEHENKLE how can I reopen this issue? Would love to add more details to this. Do I need to open a new issue?

[CEHENKLE]

I've reopened it, thanks.

[CEHENKLE]

@rramachand21 Any further thoughts on this? Thanks!

/C

[rramachand21]

@CEHENKLE updated the overview section

[GRomR1]

add related content in ES

• Feature request: graceful stop/restart elastic/elasticsearch#96147
• https://github.com/elastic/elasticsearch/blob/main/docs/reference/shutdown/apis/shutdown-api.asciidoc

this feature useful for me too. I would like using preStop hooks in Kubernetes to graceful stopping OS nodes. Like this script https://github.com/elastic/cloud-on-k8s/blob/main/pkg/controller/elasticsearch/nodespec/lifecycle_hook.go#L29

[ben-gineer]

Any updates on this? Since OpenSearch 2.12, I'm unable to cleanly shut down the service in a Dockerfile

[dblock]

@ben-gineer AFAIK nobody is working on this, please contribute!

[tmanninger]

Added PR for this issue.