[release-4.17] OCPBUGS-43664: Add vendor and architecture specific tuning options #1191
Conversation
openshift-ci-robot
added
the
jira/valid-reference
|
@MarSik: This pull request references CNF-14090 which is a valid jira issue. Warning: The referenced jira issue has an invalid target version for the target branch this PR targets: expected the story to target either version "4.17." or "openshift-4.17.", but it targets "openshift-4.18" instead. In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: MarSik The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
openshift-ci
bot
added
the
approved
|
/jira cherry-pick CNF-14090 |
|
@MarSik: Ignoring requests to cherry-pick non-bug issues: CNF-14090 In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
|
/jira cherry-pick OCPBUGS-43660 |
|
@MarSik: Jira Issue OCPBUGS-43660 has been cloned as Jira Issue OCPBUGS-43664. Will retitle bug to link to clone. In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
openshift-ci
bot
changed the title
|
@MarSik: This pull request references Jira Issue OCPBUGS-43664, which is invalid:
Comment The bug has been updated to refer to the pull request using the external bug tracker. In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
openshift-ci-robot
added
the
jira/invalid-bug
|
/jira cherry-pick OCPBUGS-43665 |
|
@MarSik: Jira Issue OCPBUGS-43665 has been cloned as Jira Issue OCPBUGS-43666. Will retitle bug to link to clone. In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
openshift-ci
bot
changed the title
|
@MarSik: This pull request references Jira Issue OCPBUGS-43666, which is invalid:
Comment The bug has been updated to refer to the pull request using the external bug tracker. In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
|
/jira refresh |
|
@MarSik: This pull request references Jira Issue OCPBUGS-43666, which is invalid:
Comment In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
MarSik
changed the title
MarSik
changed the title
|
@MarSik: This pull request references Jira Issue OCPBUGS-43664, which is invalid:
Comment In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
MarSik
changed the title
|
/jira refresh |
|
@MarSik: This pull request references Jira Issue OCPBUGS-43664, which is invalid:
Comment The bug has been updated to refer to the pull request using the external bug tracker. In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
|
/jira refresh |
|
@MarSik: This pull request references Jira Issue OCPBUGS-43664, which is invalid:
Comment In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
|
|
||
|
|
||
| cmdline_realtime_intel=tsc=reliable nmi_watchdog=0 mce=off | ||
|
|
There was a problem hiding this comment.
Nit: why so many empty lines? Similarly elsewhere. Was this in the original PR too?
There was a problem hiding this comment.
The templating generates it like this.
There was a problem hiding this comment.
And there is no way to fix it? It is pretty ugly.
There was a problem hiding this comment.
It is the same for the old profiles and this is the generated content. Yes, there is a way to fix it by using {{- xxx -}} properly in all the templates. But not in this PR as it is a wider issue.
|
@MarSik - are you sure you want to cherry-pick this now? I think it would make sense to wait for at least some degree of QE to happen for this on OCP 4.18 before the cherry-pick - given that it changes kernel parameters and such, there is potential for significant breakage if something is wrong. |
|
@bartwensley QE approval of 4.18 is needed to merge this indeed. But I want to be ready. |
|
@bartwensley Btw, the idea is that this does not change any kernel arguments for Intel. Just for the other new platforms. And it is easier to verify with the PR posted (clusterbot can build a PR, yesterday it gave me an AMD VM though..) |
|
/retest |
|
4.17 still needs backports to fix CI failures, will need to be rebased once both of these PRs are merged |
|
/retitle [release-4.17] OCPBUGS-43664: Add vendor and architecture specific tuning options |
openshift-ci
bot
changed the title
…hift#1083) * CNF-14090: Add vendor and architecture specific tuning options - Performance tuning support for 3 platforms (amd/x86,arm/aarch64,intel/x86) is added in this change - When a valid platform is detected the additional platform specific tuning will be imported alongside the default tuning - This makes use of a new helper function added to tuned to detect the system name and architecture - Update unit tests to account for the various changes - Add new unit tests to cover the platform specific tuning * CNF-14090: Re-sync e2e test yaml for tuning changes * CNF-14090: Use variable composition for idle_poll - idle=poll is only supported on x86 - Update tests to account for changes - Add explaination comments to empty values in openshift-node-performance * CNF-14090: Re-sync e2e test yaml for tuning changes * CNF-14090: Fix active/passive pstates * CNF-14090: Re-sync e2e test yaml for tuning changes
* OCPBUGS-43665: Drop amd_iommu=on from amd tuning - "=on" is not a valid value for amd_iommu - amd_iommu is enabled by default unless you specify "amd_iommu=off", unlike intel - See kernel docs for more information (https://docs.kernel.org/admin-guide/kernel-parameters.html) * OCPBUGS-43665: Update render-sync for performance profile change
* Fix kernel arguments ordering on Intel An upgrade from previous version causes one extra reboot due to differently ordered kernel arguments. This is a side effect of platform specific tuned profile split we merged in openshift#1083 This fix updates the Intel specific tuned profile to follow the same ordering that was used in the past. It does so by exploiting a specific tuned behavior of the bootloader plugin. It orders the kernel argument cmdline_suffix keys based on the order of first appearance. Any additional appearance just changes the value, but not the ordering. The change is only needed for Intel, because we have never supported other platforms before and so upgrade is not an issue. * Sync rendered manifests
MarSik
force-pushed
the
backport-14090
branch
from
4bad3b1
to
7632491
Compare
|
/retest-required |
|
/label backport-risk-assessed |
openshift-ci
bot
added
the
backport-risk-assessed
|
/retest-required |
|
@MarSik: all tests passed! Full PR test history. Your PR dashboard. Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. I understand the commands that are listed here. |
|
@MarSik: This pull request references Jira Issue OCPBUGS-43664, which is invalid:
Comment The bug has been updated to refer to the pull request using the external bug tracker. In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
CNF-14090: Re-sync e2e test yaml for tuning changes
CNF-14090: Use variable composition for idle_poll
CNF-14090: Re-sync e2e test yaml for tuning changes
CNF-14090: Fix active/passive pstates
CNF-14090: Re-sync e2e test yaml for tuning changes
OCPBUGS-43665: Drop amd_iommu=on from amd tuning
OCPBUGS-43666: Fix kernel arguments ordering on Intel
An upgrade from previous version causes one extra reboot
due to differently ordered kernel arguments. This is
a side effect of platform specific tuned profile split
we merged in #1083
This fix updates the Intel specific tuned profile to
follow the same ordering that was used in the past.
It does so by exploiting a specific tuned behavior
of the bootloader plugin. It orders the kernel argument
cmdline_suffix keys based on the order of first appearance.
Any additional appearance just changes the value, but
not the ordering.
The change is only needed for Intel, because we have
never supported other platforms before and so upgrade
is not an issue.