Skip to content

Fix the disk watermark example in ISM error resolutions - #13346

Merged
kolchfa-aws merged 2 commits into
opensearch-project:mainfrom
fjvbn2003:ism-disk-watermark-example
Oct 19, 2026
Merged

kolchfa-aws merged 2 commits into
opensearch-project:mainfrom
fjvbn2003:ism-disk-watermark-example

Conversation

@fjvbn2003

Copy link
Copy Markdown
Contributor

Description

The ISM error resolutions page's section "The index is blocked" tells readers to raise the low disk watermark with this request:

PUT _cluster/settings
{ "transient": { "cluster": { "routing": { "allocation": { "disk": { "watermark": { "low": "25.0gb" } } } } } } }

The request fails, and the advice behind it would not work either:

  1. 25.0gb is rejected: Failed to parse bytes value [25.0gb]. Fractional bytes values have been deprecated since Legacy 6.2.
  2. 25gb is also rejected, because high and flood_stage are still percentages (the defaults): unable to consistently parse [...low=25gb], [...high=90%], and [...flood_stage=95%] as percentage or bytes (DiskThresholdSettings.doValidate).
  3. The read-only-allow-delete block is applied when a node exceeds the flood-stage watermark, and DiskThresholdMonitor releases it automatically once the node is below the high watermark. Raising low does neither. When the node is already over the flood stage, raising low alone is rejected anyway (low disk watermark [90%] more than high disk watermark [10%] in the test below).

This PR replaces the example with one that raises low, high, and flood_stage together using the same unit. It also adds one sentence about the automatic release and resetting the watermarks to null afterward.

How I tested it

I used a single-node OpenSearch 3.8.0 container (security disabled, cluster.routing.allocation.disk.watermark.enable_for_single_data_node=true). To simulate a full disk, I set all three watermarks to 10% on a node at about 50% disk usage:

Step Result
Set watermarks to 10%/10%/10% After about 27 s, wm-test gets index.blocks.read_only_allow_delete: true. Indexing a document returns 429 cluster_block_exception ... disk usage exceeded flood-stage watermark.
Current doc request ("low": "25.0gb") 400, fractional bytes
Same with "low": "25gb" 400, inconsistent percentage/bytes
"low": "90%" alone 400, low disk watermark [90%] more than high disk watermark [10%]
Proposed request (low/high/flood_stage as percentages) 200. The block is removed within about 9 s, and indexing succeeds again.
Reset to null 200

I also ran the exact JSON block from the updated page (nested form, 90%/95%/97%) against the same node. It returned acknowledged: true.

Issues Resolved

None. This is a doc fix and I did not open a separate issue.

Version

all

Frontend features

N/A

Checklist

  • By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license and subject to the Developers Certificate of Origin.
    For more information on following Developer Certificate of Origin and signing off your commits, please check here.

This change was drafted with AI assistance (Claude Code) and reviewed by me.

🤖 Generated with Claude Code

The "The index is blocked" resolution suggested raising only the low
disk watermark to "25.0gb". OpenSearch rejects that request twice over:
fractional byte values are not accepted, and a byte value for `low`
cannot be combined with the default percentage values of `high` and
`flood_stage`. Raising `low` would also not remove the
read-only-allow-delete block, which follows the flood-stage watermark.

Raise all three watermarks with the same unit, and note that the block
is released automatically once disk usage falls below the high
watermark.

Signed-off-by: YoungjuKim <[email protected]>
@github-actions

Copy link
Copy Markdown

Thank you for submitting your PR. The PR states are In progress (or Draft) -> Tech review -> Doc review -> Merged.

Before you submit your PR for doc review, make sure the content is technically accurate. If you need help finding a tech reviewer, tag a maintainer.

When you're ready for doc review, tag the assignee of this PR. The doc reviewer may push edits to the PR directly or leave comments and editorial suggestions for you to address (let us know in a comment if you have a preference).

@github-actions github-actions Bot added the Tech review PR: Tech review in progress label Oct 16, 2026
Signed-off-by: Fanit Kolchina <[email protected]>

@kolchfa-aws kolchfa-aws left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you, @fjvbn2003! Updated to show both byte and percentage versions. LGTM

@kolchfa-aws
kolchfa-aws merged commit 24d0c62 into opensearch-project:main Oct 19, 2026
4 checks passed
kolchfa-aws added a commit that referenced this pull request Oct 19, 2026
)

* Fix the disk watermark example in ISM error resolutions

The "The index is blocked" resolution suggested raising only the low
disk watermark to "25.0gb". OpenSearch rejects that request twice over:
fractional byte values are not accepted, and a byte value for `low`
cannot be combined with the default percentage values of `high` and
`flood_stage`. Raising `low` would also not remove the
read-only-allow-delete block, which follows the flood-stage watermark.

Raise all three watermarks with the same unit, and note that the block
is released automatically once disk usage falls below the high
watermark.



* Update to show both versions



---------




(cherry picked from commit 24d0c62)

Signed-off-by: YoungjuKim <[email protected]>
Signed-off-by: Fanit Kolchina <[email protected]>
Signed-off-by: opensearch-ci-bot <[email protected]>
Co-authored-by: Youngju KIM <[email protected]>
Co-authored-by: Fanit Kolchina <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

backport 3.9 Tech review PR: Tech review in progress

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants