kernel-mshv: CONFIG_PSI was dropped in the dom0 v2411.19.1 config update
Maintainers usually reply within 1 day
Nobody has claimed this yet.
Assessment
- Difficulty
- 2/5
- Estimated time
- 1-3 hours
- Newbie friendliness
- 72/100
- Issue type
- Bug
- Clarity
- Clearly specified
- Activity status
- Active
- Tech stack
- linux
- Domain
- operating-systems
Research direction
Start by reading SPECS/kernel-mshv/config and the source defconfig mentioned in the issue, then compare the PSI setting with the history around #9720, #13648, and #17187. Restore CONFIG_PSI=y in the relevant configuration if the regression is confirmed, and verify the built kernel config and /proc/pressure/{cpu,io,memory} as described in the reproduction steps.
Written by the indexing model from the issue text.
Description
Environment
- Azure Linux version and/or image name: 3.0, AKS Azure Linux 3 Kata (Pod Sandboxing) node image 202610.08.0
- Kernel version: 6.6.137.mshv2-2.azl3
- Architecture: x86_64
- Affected package(s) and version(s), if applicable: kernel-mshv 6.6.137.mshv2-2.azl3
- Custom image or package modifications, if applicable: none
Deployment / platform
- Azure VM
- AKS
- Azure Linux container image
- Other (please describe)
Describe the bug
kernel-mshv is built without CONFIG_PSI, so Kata hosts have no pressure stall information (/proc/pressure/* and the cgroup v2 *.pressure files are absent).
PSI was enabled for this kernel in #9720 ("kernel-mshv: enable CONFIG_PSI for cgroup-memory-telemetry"). It was then removed by #13648 ("dom0 packages: Update to dom0 release v2411.19.1"): the parent of that commit has CONFIG_PSI=y in SPECS/kernel-mshv/config, and the commit itself has # CONFIG_PSI is not set. The PR doesn't mention PSI, so this looks like an unintended regression from the config update rather than a deliberate change. #17187 later restored several other options but not PSI.
Steps to reproduce
- Boot a node running
kernel-mshv6.6.137.mshv2-2.azl3 (for example, an AKS Kata node pool). - Run
grep CONFIG_PSI /boot/config-$(uname -r)andls /proc/pressure /sys/fs/cgroup/cpu.pressure.
Expected behavior
CONFIG_PSI=y, as after #9720 and as in the standard Azure Linux 3 kernel, so /proc/pressure/{cpu,io,memory} exist.
Actual behavior
# CONFIG_PSI is not set
ls: cannot access '/proc/pressure': No such file or directory
ls: cannot access '/sys/fs/cgroup/cpu.pressure': No such file or directory
No psi= parameter is on the kernel command line; because the option is compiled out, it can't be enabled at boot.
Additional context
Impact on AKS Kata hosts: Prometheus node-exporter's pressure collector returns no data (node_scrape_collector_success{collector="pressure"} 0, no node_pressure_* series), and node health checks that rely on CPU/memory pressure can't use PSI. Other kernel-mshv users who rely on PSI (including the cgroup memory telemetry that motivated #9720) are likely affected too.
Could CONFIG_PSI=y be restored, along with its source defconfig, so a future dom0 sync doesn't drop it again? If PSI was intentionally disabled for dom0, could you share the reason so we can document the expected behavior?
- Dominant language
- Python
- Stars
- 5.4k
- Forks
- 708
- Avg merge
- 2d 10h
- Merged PRs (30d)
- 412
Getting set up
- No Dockerfile or Docker Compose file
- No pull request template
- Read the contributing guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from microsoft/azurelinux
-
[4.0] Offline installer does not handle HTTP URLs in inst.ks kernel parameterPossibly taken @sonarbhavarth claimed this 102 days ago. Open4.0 enhancement
Difficulty 2/5 1-3 hours Newbie friendliness 76/100
microsoft/azurelinux#17915 · 1 comment ·
Maintainers usually reply within 1 day
-
[4.0] Live installer initrd missing NFS dracut module preventing NFS based PXE deploymentPossibly taken @IzanVil claimed this 51 days ago. Open4.0 enhancement
Difficulty 2/5 1-3 hours Newbie friendliness 70/100
microsoft/azurelinux#17913 · 1 comment ·
Maintainers usually reply within 1 day
-
documentation
Difficulty 5/5 Over a week Newbie friendliness 35/100
microsoft/azurelinux#19419 ·
Maintainers usually reply within 1 day
-
[4.0] rust may need libssh2 disabled for FIPS modePossibly taken @tobiasb-ms claimed this 4 days ago. Openbug
microsoft/azurelinux#19277 · 2 comments · 1 assignee ·
Maintainers usually reply within 1 day
-
[4.0] aarch64 repo publishes noarch sub-packages from ExclusiveArch: x86_64 SRPMs (uninstallable: syslinux-nonlinux, rocm-rpm-macros-modules, python3-tensile-devel)Possibly taken @UditDewan claimed this 4 days ago. Open4.0
Difficulty 4/5 3-5 days Newbie friendliness 35/100
microsoft/azurelinux#18985 · 2 comments ·
Maintainers usually reply within 1 day
All issues in microsoft/azurelinux
Similar issues
-
Difficulty 2/5 1-3 hours Newbie friendliness 85/100
Maintainers usually reply within 1 day
-
infrastructure qa
Difficulty 2/5 1-3 hours Newbie friendliness 84/100
khuisman/mcp-gee-sweet#1068 ·
Maintainers usually reply within 1 day
-
area:ecosystem documentation
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
Chelis-Lang/chelis#3866 ·
Maintainers usually reply within 1 day
-
`lvm_compact` has no self-hosted twin, and `docs/158` §5 says every language-face component mustOpenbug sev-3 vm
Difficulty 2/5 1-3 hours Newbie friendliness 65/100
-
issue: Picking a prompt from the / menu in a channel inserts a command chip instead of the promptOpenbug confirmed issue
Difficulty 2/5 Half a day Newbie friendliness 78/100
open-webui/open-webui#32712 · 2 comments ·
Maintainers usually reply within 1 day