Document 120-day default retention period for ML jobs in log anomalies and log categories

[Copilot]

Summary

This PR updates the documentation to reflect the new default retention period of 120 days for ML results in the predefined ML configurations for logs that power the log anomalies and log categories pages, introduced in version 9.2.

Context

Following the changes introduced in elastic/kibana#231080, the predefined ML configurations for logs now set a default retention period of 120 days for ML results, where previously there was no default retention period (results were retained indefinitely).

Changes

Added concise notes in active voice to three documentation files, with version-specific tags:

1. solutions/observability/logs/inspect-log-anomalies.md

   • Added a note in the "Enable log rate analysis and anomaly detection" section
   • Informs users that log anomaly ML jobs retain results for 120 days by default
   • Includes :applies_to: stack: ga 9.2 tag

2. solutions/observability/logs/categorize-log-entries.md

   • Added a note in the "Create log categories" section
   • Informs users that log categorization ML jobs retain results for 120 days by default
   • Includes :applies_to: stack: ga 9.2 tag

3. explore-analyze/machine-learning/anomaly-detection/anomaly-detection-scale.md

   • Updated the "Optimize the results retention" section
   • Clarified that while anomaly detection results are retained indefinitely by default, predefined ML configurations for logs are an exception with 120-day retention (from version 9.2)
   • Includes inline {applies_to}stack: ga 9.2`` tag

Each note:

• States the 120-day default retention period
• Explains how to customize the retention period via the results_retention_days setting
• Clearly indicates this feature is available from version 9.2 onwards

Impact

• Target release: 9.2
• Documentation sets affected: Elastic On-Prem and Cloud (all)
• User benefit: Users are now informed about the automatic cleanup of older ML results, helping them understand storage management and plan accordingly. Version-specific tags ensure users understand when this feature became available.

Related

• Kibana PR: Set the default retention period for the Logs anomaly detection configurations to 120 days kibana#231080
• Contact: @gbamparop

Fixes #2476

Original prompt

---

This section details on the original issue you should resolve

<issue_title>[Internal]: Setting a default retention period for the ML results for the jobs enabling log anomalies and log categories</issue_title>
<issue_description>### Description

The predefined ML modules for logs that power the log anomalies and log categories pages didn't set a default retention period for the ML results, this will be updated to a default retention of 120 days.

Resources

• Set the default retention period for the Logs anomaly detection configurations to 120 days kibana#231080

Which documentation set does this change impact?

Elastic On-Prem and Cloud (all)

Feature differences

There was no default retention period before and it will be set to 120 days.

What release is this request related to?

9.2

Serverless release

Date TBD

Collaboration model

The documentation team

Point of contact.

Main contact:

• @gbamparop
  </issue_description>

<agent_instructions>A simple note should suffice where relevant. Updates are probably focused in the following parts of the documentation: https://github.com/elastic/docs-content/tree/main/solutions/observability/logs and/or https://github.com/elastic/docs-content/tree/main/explore-analyze/machine-learning</agent_instructions>

Comments on the Issue (you are @copilot in this section)

Fixes #2476

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

[github-actions]

🔍 Preview links for changed docs

• explore-analyze/machine-learning/anomaly-detection/anomaly-detection-scale.md
• solutions/observability/logs/categorize-log-entries.md
• solutions/observability/logs/inspect-log-anomalies.md

[florent-leborgne]

@copilot shorten these notes and use active voice.
We should also edit the following file that states that "results are retained indefinitely by default.": explore-analyze/machine-learning/anomaly-detection/anomaly-detection-scale.md

[Copilot]

@copilot shorten these notes and use active voice.
We should also edit the following file that states that "results are retained indefinitely by default.": explore-analyze/machine-learning/anomaly-de...

Done. Shortened the notes to use active voice and updated anomaly-detection-scale.md to clarify that log modules are an exception to the indefinite retention default. (502f69e → e8c7d43)

[florent-leborgne]

@gbamparop experimenting with copilot here. Do these changes look enough to you?

[florent-leborgne]

@copilot we need it clear that this default retention is only from version 9.2. Use applies_to tags for this

• inline: {applies_to}stack: ga 9.2
• In notes, you can specify it this way:
  :::{note}
  :applies_to: stack: ga 9.2

[Copilot]

@copilot we need it clear that this default retention is only from version 9.2. Use applies_to tags for this

• inline: {applies_to}stack: ga 9.2
• In notes, you can specify it this way:
  ...

Added applies_to tags for stack: ga 9.2 to all three files. (6f8a1b2)