Remove cluster.routing.allocation.disk.watermark.enable_for_single_data_node setting #114207
Conversation
arteam added >breaking WIP :Distributed Coordination/Allocation 
| We haven't yet decided that we want to do this in v9, let's hold off here until that's settled. |
…ata_node` setting Prior to 7.8, whenever a cluster had only a single data node, the watermarks would not be respected. This was incompatible with how storage based autoscaling on ESS/ECH works and in order to change this in 7.8+ in a backwards compatible way, we had to introduce the `cluster.routing.allocation.disk.watermark.enable_for_single_data_node node` setting. The setting was deprecated in elastic#73733 (7.14), and was made to accept only `true` in elastic#73737 (8.0).
arteam force-pushed the remove-watermark-enable-for-single-data-node branch from 50f6cae to ea54648 Compare 
@DaveCTurner Yes, I'm aware that the breaking change hasn't been approved yet, that's why I created this PR as a draft and added a |
…le-for-single-data-node
henningandersen left a comment
henningandersen left a comment There was a problem hiding this comment.
Looks good, though I think we should add a breaking change changelog entry to this PR. I thought that was half-automatic, but perhaps not from draft?
elasticsearchmachine added the needs:triage 
| Hi @arteam, I've created a changelog YAML for you. Note that since this PR is labelled |
| @elasticmachine update branch |
arteam added Team:Distributed Coordination | Pinging @elastic/es-distributed-coordination (Team:Distributed Coordination) |
| Hi @henningandersen! I've made the PR ready to review and added the breaking changelog entry. |
…le-for-single-data-node
henningandersen left a comment
henningandersen left a comment There was a problem hiding this comment.
LGTM (with one breaking change documentation change).
Co-authored-by: Henning Andersen <33268011+henningandersen@users.noreply.github.com>
Co-authored-by: Henning Andersen <33268011+henningandersen@users.noreply.github.com>
| @elasticmachine update branch |
…le-for-single-data-node
arteam force-pushed the remove-watermark-enable-for-single-data-node branch from afed3ad to 7dab95c Compare arteam added the auto-merge-without-approval 
…ata_node` setting (#114207) Prior to 7.8, whenever a cluster had only a single data node, the watermarks would not be respected. This was incompatible with how storage based autoscaling works and in order to change this in 7.8+ in a backwards compatible way, we had to introduce the `cluster.routing.allocation.disk.watermark.enable_for_single_data_node node` setting. The setting was deprecated in #73733 (7.14), and was made to accept only `true` in #73737 (8.0).
…ata_node` setting (elastic#114207) Prior to 7.8, whenever a cluster had only a single data node, the watermarks would not be respected. This was incompatible with how storage based autoscaling works and in order to change this in 7.8+ in a backwards compatible way, we had to introduce the `cluster.routing.allocation.disk.watermark.enable_for_single_data_node node` setting. The setting was deprecated in elastic#73733 (7.14), and was made to accept only `true` in elastic#73737 (8.0).
| @arteam is this PR relevant to the serverless changelog? [FYI this question is based on 9.0 breaking changes] |
| @leemthompo No, this change doesn't have any effect on ES Serverless |
6 participants
Prior to 7.8, whenever a cluster had only a single data node, the watermarks would not be respected. This was incompatible with how storage based autoscaling works and in order to change this in 7.8+ in a backwards compatible way, we had to introduce the
cluster.routing.allocation.disk.watermark.enable_for_single_data_node nodesetting.The setting was deprecated in #73733 (7.14), and was made to accept only
truein #73737 (8.0).