Update dependency prometheus/node_exporter to v1.11.1

[renovate]

This PR contains the following updates:

Package	Update	Change
prometheus/node_exporter	minor	1.10.2 → 1.11.1

---

Release Notes

prometheus/node_exporter (prometheus/node_exporter)

v1.11.1: 1.11.1 / 2026-04-07

Compare Source

• [BUGFIX] Fix kernel_hung for no data #​3613

v1.11.0: 1.11.0 / 2026-04-04

Compare Source

• [CHANGE] Adds distroless container image. #​3592
• [FEATURE] infiniband: Add hw_counters metrics #​2827
• [FEATURE] kernel_hung: Add collector for kernel hung tasks #​3470
• [FEATURE] mixin: Add "All" option to cluster variable in dashboards #​3509
• [FEATURE] netlink: Add node_network_altnames_info metric #​3541
• [FEATURE] thermal: Add support for CPU temperature metrics on Apple Silicon #​3547
• [FEATURE] bcachefs: Add new collector for bcachefs #​3523
• [ENHANCEMENT] mixin: Make devices in network graphs configurable #​3525
• [ENHANCEMENT] nvme: Add namespace parameters #​3360
• [BUGFIX] netdev: Respect deviceFilter for node_network_address_info metric #​3340
• [BUGFIX] mountstats: Fix multiple transports statistics labeling #​3358
• [BUGFIX] diskstats: Fix panic when using device-include flag only #​3603

---

Configuration

📅 Schedule: (in timezone America/New_York)

• Branch creation
  • "after 2am and before 8am on monday"
• Automerge
  • At any time (no schedule defined)

🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.

♻ Rebasing: Whenever PR is behind base branch, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about this update again.

---

• If you want to rebase/retry this PR, check this box

---

This PR was generated by Mend Renovate. View the repository job log.

[github-actions]

prometheus/node_exporter (ansible) 1.10.2 -> 1.11.1

Risk: 🟢 Safe

The Deep Dive

Update Scope

Bumps node_exporter_version in ansible/group_vars/all/node_exporter.yaml from 1.10.2 → 1.11.1 (minor + patch). The role is geerlingguy.node_exporter, which downloads and unpacks the official upstream binary tarball — no container image change is involved. Affected hosts are the kubernetes group (k1–k5) and the pi group (pantrypi, garagepi), per group_vars/kubernetes.yaml and group_vars/pi.yaml setting node_exporter_enabled: true. Runtime flags (--collector.textfile.directory=..., --log.level=info) are unchanged.

Performance & Stability

• BUGFIX diskstats: Fix panic when using device-include flag only (#3603) — not triggered here because the deployed node_exporter_options does not pass --collector.diskstats.device-include.
• BUGFIX mountstats: Fix multiple transports statistics labeling (#3358) — only relevant if NFS clients use multi-transport (nconnect>1) mounts; user has no mountstats-specific config.
• BUGFIX netdev: Respect deviceFilter for node_network_address_info metric (#3340) — not relevant; no --collector.netdev.device-(in|ex)clude flag is set.

Features & UX

Each item below was verified at v1.11.1 by reading the registering source file. All collector flags would be added to node_exporter_options in ansible/group_vars/all/node_exporter.yaml if you wanted to opt out.

• kernel_hung: collector for kernel hung tasks (#3470) — registered defaultEnabled in collector/kernel_hung_linux.go (registerCollector("kernel_hung", defaultEnabled, ...)). Reads /proc/sys/kernel/hung_task_detect_count, only present on Linux ≥ 6.13. Already active by default; disable with --no-collector.kernel_hung. Returns ErrNoData quietly when the sysctl is absent (PR #3613).
• bcachefs: new collector (#3523) — registered defaultEnabled in collector/bcachefs_linux.go (registerCollector(subsystem, defaultEnabled, NewBcachefsCollector)). Reads /sys/fs/bcachefs; returns ErrNoData when absent. Already active by default; disable with --no-collector.bcachefs. User has no bcachefs filesystems.
• infiniband: hw_counters metrics (#2827) — emitted by the existing infiniband collector, registered defaultEnabled in collector/infiniband_linux.go (registerCollector("infiniband", defaultEnabled, NewInfiniBandCollector)). Already active by default; disable with --no-collector.infiniband. Only emits for InfiniBand HCAs, none in the homelab.
• netlink: node_network_altnames_info metric (#3541) — added to the netclass collector's RTNL/netlink path in collector/netclass_rtnl_linux.go. The netlink path is gated by --collector.netclass.netlink (default false, see collector/netclass_linux.go). Not enabled in the deployed node_exporter_options, so this metric will NOT appear without first adding --collector.netclass.netlink.
• thermal: Apple Silicon CPU temperature (#3547) — Darwin-only build target; the Linux binary deployed by geerlingguy.node_exporter is unaffected.
• mixin: cluster All option (#3509) and configurable network devices (#3525) — Grafana dashboard JSON only; this repo does not consume the upstream mixin.
• ENHANCEMENT nvme: namespace parameters (#3360) — emitted by the existing nvme collector, registered defaultEnabled in collector/nvme_linux.go (registerCollector("nvme", defaultEnabled, NewNVMeCollector)). Already active by default; disable with --no-collector.nvme. Adds NVMe namespace labels on hosts with NVMe drives; purely additive.

Key Fixes

• Fix kernel_hung for no data (#3613) — the headline 1.11.1 fix. In 1.11.0, the new kernel_hung collector logs ERROR collector failed name=kernel_hung err="open /proc/sys/kernel/hung_task_detect_count: no such file or directory" on every scrape on kernels that lack the /proc/sys/kernel/hung_task_detect_count sysctl, which was added in Linux 6.13 (see issue 3612). The kubernetes nodes are installed via the os-install role using preseed.trixie.cfg (Debian 13 trixie, kernel 6.12), and the Pi image is pinned to 2026-04-13-raspios-trixie-arm64 in scripts/rpi-image-customize (also kernel 6.12) — both below 6.13, so 1.11.0 would have logged this error continuously across the whole fleet. 1.11.1 returns ErrNoData instead, so jumping straight from 1.10.2 → 1.11.1 sidesteps the noisy-log window entirely.

Newer Versions

1.11.1 is the latest release (next entry is master / unreleased in CHANGELOG.md, confirmed via gh release list --repo prometheus/node_exporter). No newer fixes pending.

Hazards & Risks

None identified. Upstream introduced no breaking changes between 1.10.2 and 1.11.1, the only [CHANGE] entry (distroless container image, #3592) does not affect this deployment because the role pulls a binary tarball from GitHub releases, and the user's runtime flags don't intersect any of the bug-fix code paths. New default-enabled collectors degrade gracefully when their data sources are absent.

Sources

• node_exporter v1.11.1 release notes
• node_exporter v1.11.0 release notes
• Compare v1.10.2...v1.11.1
• CHANGELOG.md @ v1.11.1
• Issue #3612: kernel_hung collector fails loudly
• PR #3613: Fix kernel_hung for no data
• geerlingguy.node_exporter role tasks

---

🟢 Verdict: Safe

Straightforward minor upgrade with no CVEs, no breaking changes, and no flags affected by the bug fixes. Jumping directly to 1.11.1 also skips the 1.11.0 kernel_hung error-log regression. Merge and let Semaphore deploy to the kubernetes and pi host groups; expect a few new (mostly empty) metric series in Prometheus.