Highlights
This is a minor release 0.36.0 containing changes to GPU plugin's monitoring functionality and some smaller changes/fixes.
Generic
- Update k8s components to 0.36.0 version (#2275)
- Golang 1.26 (#2246)
- Go improvements (#2248)
- e2e: move to ginkgo labels (#2258)
- OpenShift documentation created (#2265)
GPU
- Support new 2.x xpumd's device info API (#2278)
- Move from KMD specific monitoring resources to a singular resource
gpu.intel.com/monitoring(#2244)
Operator
- Support for e2e operator tests for OCP and vanilla k8s
Bugs fixed
- QAT: Fix provisioning config not applying for an existing QAT CR (#2270)
- GPU: readme renders incorrectly (#2239)
- GPU: Include /dev/mei* devices in the monitoring resource (#2237)
What's Changed (excluding dependabot)
- Post 0.35 updates to main by @tkatila in #2231
- gpu: include /dev/mei* devices for _monitoring resource by @tkatila in #2242
- modernize Go and fix prealloc findings by @mythi in #2248
- GPU: move to single/non-kmd tied monitoring resource by @tkatila in #2244
- Pin setup-envtest to release-0.23 branch by @Copilot in #2257
- cdi: add npu and gpu examples for docker use by @tkatila in #2253
- e2e: use labels to filter tests by @tkatila in #2258
- Remove stale nodes/proxy trivyignore, merge .trivyignore into .trivyignore.yaml by @Copilot in #2260
- Fix table rendering in GPU plugin README by adding missing blank line by @Nesar976 in #2263
- publish: add missing 0.35 release branch for triggers by @tkatila in #2264
- e2e: rewrite operator tests with OCP support by @tkatila in #2259
- build: use Go 1.26 by @Copilot in #2246
- k8s 1.36.0 by @mythi in #2275
- operator: qat: fix provisioning config change propagation and asym;sym validation by @Copilot in #2271
- GPU plugin: add xpumd 2.x as alternative health data source by @Copilot in #2279
- Move Openshift documentation to this repository by @Copilot in #2265
- build: drop -N -l flags by @tkatila in #2288
New Contributors
Full Changelog: v0.35.0...v0.36.0