Skip to content

Update ECE cgroup v2 support note in software prerequisites - #8589

Merged
erikjenks merged 6 commits into
mainfrom
ece-cgroup-v2-support
Oct 26, 2026
Merged

erikjenks merged 6 commits into
mainfrom
ece-cgroup-v2-support

Conversation

@erikjenks

@erikjenks erikjenks commented Sep 29, 2026 •

Copy link
Copy Markdown
Contributor

Summary

Removes the blanket statement that ECE is not supported on cgroup v2. ECE 3.8.x, 4.2.x, and later support cgroup v2. Replaces the old note with a customer-facing summary covering the support boundary, OS-level cgroup version breakdown, and known limitations.

Changes

  • Removes: "ECE is not supported on Linux distributions that use cgroups version 2"
  • Rewrites the first bullet to clarify that cgroup version is determined by the OS, and lists which OS versions use v1 vs v2
  • Adds three additional bullets:
    • Stack < 7.15.0 OOM risk on cgroup v2 hosts
    • ES 8.0.0–8.4.3 startup failure (existing bullet, kept)
    • cpuacct.* → cpu.* metric field rename and docker.memory.rss.pct removal for custom dashboards/alerts

OS cgroup version breakdown (from ECE integration tests)

OS cgroup version
Ubuntu 22.04 v2
Ubuntu 24.04 v2
RHEL 9.x v2
RHEL 10.x v2
Rocky Linux 9.x v2
SLES 15 SP7 v2
Ubuntu 20.04 v1 only
RHEL 8.x v1 only
Rocky Linux 8.x v1 only

File changed

deploy-manage/deploy/cloud-enterprise/ece-software-prereq.md

🤖 Generated with Claude Code

Remove the blanket 'not supported on cgroup v2' statement. ECE 3.8.x,
4.2.x, and later support cgroup v2. Add customer-facing notes covering:
- Version support boundary (4.0 and 4.1 support cgroup v1 only)
- Stack < 7.15.0 OOM risk on cgroup v2 hosts
- Elasticsearch 8.0.0–8.4.3 startup failure (existing bullet, kept)
- cpuacct.* → cpu.* field rename and docker.memory.rss.pct removal

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
@erikjenks
erikjenks requested a review from a team as a code owner September 29, 2026 14:24
@github-actions

Copy link
Copy Markdown
Contributor

Elastic Docs AI PR menu

Check the box to run an AI review for this pull request.

  • Review docs changes (docs-review). Status: not started.

Powered by GitHub Agentic Workflows and docs-actions. For more information, reach out to the docs team.

@github-actions

github-actions Bot commented Sep 29, 2026 •

Copy link
Copy Markdown
Contributor

🔍 Preview links for changed docs

⏳ Building and deploying preview... View progress

This comment will be updated with preview links when the build is complete.

@github-actions

Copy link
Copy Markdown
Contributor

✅ Elastic Docs Style Checker (Vale)

No issues found on modified lines!


The Vale linter checks documentation changes against the Elastic Docs style guide. To use Vale locally or report issues, refer to Elastic style guide for Vale.

Ubuntu 20.04, RHEL 8.x, and Rocky Linux 8.x use cgroup v1.
Ubuntu 22.04+, RHEL 9.x+, Rocky Linux 9.x+, and SLES 15 use cgroup v2.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

@shainaraskas shainaraskas left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

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

a couple of optimization comments. from what I am finding in the docs you linked and the internet, we always want to say cgroups

@@ -29,8 +29,10 @@ We recommend using kernel 4.15.x or later on Ubuntu.
To check your kernel version, run `uname -r`.

::::{note}

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

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

this note has reached escape velocity imo

Suggested change
::::{note}
### cgroups support

* {{stack}} versions earlier than 7.15.0, including {{es}}, {{kib}}, APM Server, Integrations Server, and Beats, don't detect cgroup v2 memory limits. On cgroup v2 hosts, these components can use more memory than allocated and get stopped by the Linux out-of-memory (OOM) killer. Upgrade deployments to 7.15.0 or later before moving hosts to cgroup v2.
* {{es}} versions 8.0.0–8.4.3 fail to start on hosts using cgroup v2 with the bundled JDK. Refer to [our KB article](https://ela.st/es-may-fail-on-cgroup-v2) for more details.
* On cgroup v2 hosts, [platform monitoring](/deploy-manage/monitor/orchestrators/ece-platform-monitoring.md) reports process CPU metrics in `system.process.cgroup.cpu.*` instead of `system.process.cgroup.cpuacct.*`, and `docker.memory.rss.pct` is unavailable. Update any custom dashboards or alerts that query these fields, or they return no data.
::::

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

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

to match with making this a section

Suggested change
::::


::::{note}
* {{ece}} is not supported on Linux distributions that use [cgroups](https://man7.org/linux/man-pages/man7/cgroups.7.html) version 2.
* {{ece}} 3.8.x, 4.2.x, and later support Linux distributions that use [cgroups](https://man7.org/linux/man-pages/man7/cgroups.7.html) version 2. The cgroup version is determined by the OS, not by ECE. Ubuntu 20.04, RHEL 8.x, and Rocky Linux 8.x use cgroup v1. Ubuntu 22.04 and later, RHEL 9.x and later, Rocky Linux 9.x and later, and SLES 15 use cgroup v2. Earlier ECE versions, including 4.0 and 4.1, support only cgroup v1.

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

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

we technically only document 4.x in this system (because we have not injected 3.8 details holistically into this docset). ideal spot for 3.8 details is here

Suggested change
* {{ece}} 3.8.x, 4.2.x, and later support Linux distributions that use [cgroups](https://man7.org/linux/man-pages/man7/cgroups.7.html) version 2. The cgroup version is determined by the OS, not by ECE. Ubuntu 20.04, RHEL 8.x, and Rocky Linux 8.x use cgroup v1. Ubuntu 22.04 and later, RHEL 9.x and later, Rocky Linux 9.x and later, and SLES 15 use cgroup v2. Earlier ECE versions, including 4.0 and 4.1, support only cgroup v1.
Linux distributions that use [cgroups](https://man7.org/linux/man-pages/man7/cgroups.7.html) version 2 are only supported in {{ece}} 4.2 and later. {{ece}} versions 4.0-4.1 only support cgroups version 1.
The cgroups version is determined by the OS, not by ECE.
For example:
* Ubuntu 20.04, RHEL 8.x, and Rocky Linux 8.x use cgroups v1.
* Ubuntu 22.04 and later, RHEL 9.x and later, Rocky Linux 9.x and later, and SLES 15 use cgroups v2.

if you really want 3.8 detail

Suggested change
* {{ece}} 3.8.x, 4.2.x, and later support Linux distributions that use [cgroups](https://man7.org/linux/man-pages/man7/cgroups.7.html) version 2. The cgroup version is determined by the OS, not by ECE. Ubuntu 20.04, RHEL 8.x, and Rocky Linux 8.x use cgroup v1. Ubuntu 22.04 and later, RHEL 9.x and later, Rocky Linux 9.x and later, and SLES 15 use cgroup v2. Earlier ECE versions, including 4.0 and 4.1, support only cgroup v1.
Linux distributions that use [cgroups](https://man7.org/linux/man-pages/man7/cgroups.7.html) version 2 are only supported in {{ece}} 4.2 and later. {{ece}} versions 4.0-4.1 only support cgroups version 1.
:::{tip}
ECE 3.8 also supports cgroups v2.
:::
The cgroups version is determined by the OS, not by ECE.
For example:
* Ubuntu 20.04, RHEL 8.x, and Rocky Linux 8.x use cgroups v1.
* Ubuntu 22.04 and later, RHEL 9.x and later, Rocky Linux 9.x and later, and SLES 15 use cgroups v2.

Comment on lines +33 to +35
* {{stack}} versions earlier than 7.15.0, including {{es}}, {{kib}}, APM Server, Integrations Server, and Beats, don't detect cgroup v2 memory limits. On cgroup v2 hosts, these components can use more memory than allocated and get stopped by the Linux out-of-memory (OOM) killer. Upgrade deployments to 7.15.0 or later before moving hosts to cgroup v2.
* {{es}} versions 8.0.0–8.4.3 fail to start on hosts using cgroup v2 with the bundled JDK. Refer to [our KB article](https://ela.st/es-may-fail-on-cgroup-v2) for more details.
* On cgroup v2 hosts, [platform monitoring](/deploy-manage/monitor/orchestrators/ece-platform-monitoring.md) reports process CPU metrics in `system.process.cgroup.cpu.*` instead of `system.process.cgroup.cpuacct.*`, and `docker.memory.rss.pct` is unavailable. Update any custom dashboards or alerts that query these fields, or they return no data.

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

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

Suggested change
* {{stack}} versions earlier than 7.15.0, including {{es}}, {{kib}}, APM Server, Integrations Server, and Beats, don't detect cgroup v2 memory limits. On cgroup v2 hosts, these components can use more memory than allocated and get stopped by the Linux out-of-memory (OOM) killer. Upgrade deployments to 7.15.0 or later before moving hosts to cgroup v2.
* {{es}} versions 8.0.0–8.4.3 fail to start on hosts using cgroup v2 with the bundled JDK. Refer to [our KB article](https://ela.st/es-may-fail-on-cgroup-v2) for more details.
* On cgroup v2 hosts, [platform monitoring](/deploy-manage/monitor/orchestrators/ece-platform-monitoring.md) reports process CPU metrics in `system.process.cgroup.cpu.*` instead of `system.process.cgroup.cpuacct.*`, and `docker.memory.rss.pct` is unavailable. Update any custom dashboards or alerts that query these fields, or they return no data.
The following limitations apply to hosts using cgroups v2:
* {{stack}} versions earlier than 7.15, including {{es}}, {{kib}}, APM Server, Integrations Server, and Beats, don't detect cgroups v2 memory limits. On cgroups v2 hosts, these components can use more memory than allocated and get stopped by the Linux out-of-memory (OOM) killer. Upgrade deployments to {{stack}} version 7.15 or later before moving hosts to cgroups v2.
* {{es}} versions 8.0–8.4.3 fail to start on hosts using cgroups v2 with the bundled JDK. Refer to [our KB article](https://ela.st/es-may-fail-on-cgroup-v2) for more details.
* On cgroups v2 hosts, [platform monitoring](/deploy-manage/monitor/orchestrators/ece-platform-monitoring.md) reports process CPU metrics in `system.process.cgroup.cpu.*` instead of `system.process.cgroup.cpuacct.*`, and `docker.memory.rss.pct` is unavailable. Update any custom dashboards or alerts that query these fields.

erikjenks and others added 2 commits October 5, 2026 09:36
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>

The following limitations apply to hosts using cgroups v2:

* {{stack}} versions earlier than 7.15, including {{es}}, {{kib}}, APM Server, Integrations Server, and Beats, don't detect cgroups v2 memory limits. On cgroups v2 hosts, these components can use more memory than allocated and get stopped by the Linux out-of-memory (OOM) killer. Upgrade deployments to {{stack}} version 7.15 or later before moving hosts to cgroups v2.

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

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

Nit: Since ECE 4.x only allows versions 8+ this situation should not be possible.

@erikjenks
erikjenks enabled auto-merge October 26, 2026 13:05
@erikjenks
erikjenks added this pull request to the merge queue Oct 26, 2026
Merged via the queue into main with commit 3ec70bc Oct 26, 2026
12 checks passed
@erikjenks
erikjenks deleted the ece-cgroup-v2-support branch October 26, 2026 13:11
@gigerdo

gigerdo commented Oct 26, 2026

Copy link
Copy Markdown
Member

FYI @kunisen

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants