Repository navigation
Release cortex 0.4.0 - #1177
Merged
Merged
Release cortex 0.4.0#1177
Conversation
This PR contains the following updates: | Package | Update | Change | |---|---|---| | [kube-prometheus-stack](https://redirect.github.com/prometheus-operator/kube-prometheus) ([source](https://redirect.github.com/prometheus-community/helm-charts)) | patch | `88.5.2` → `88.5.4` | --- ### Release Notes <details> <summary>prometheus-community/helm-charts (kube-prometheus-stack)</summary> ### [`v88.5.4`](https://redirect.github.com/prometheus-community/helm-charts/releases/tag/kube-prometheus-stack-88.5.4) [Compare Source](https://redirect.github.com/prometheus-community/helm-charts/compare/kube-prometheus-stack-88.5.3...kube-prometheus-stack-88.5.4) kube-prometheus-stack collects Kubernetes manifests, Grafana dashboards, and Prometheus rules combined with documentation and scripts to provide easy to operate end-to-end Kubernetes cluster monitoring with Prometheus using the Prometheus Operator. #### What's Changed - \[kube-prometheus-stack] Update Helm release grafana to v12.11.2 by [@​renovate](https://redirect.github.com/renovate)\[bot] in [#​7213](https://redirect.github.com/prometheus-community/helm-charts/pull/7213) **Full Changelog**: <prometheus-community/helm-charts@prometheus-ipmi-exporter-0.8.2...kube-prometheus-stack-88.5.4> ### [`v88.5.3`](https://redirect.github.com/prometheus-community/helm-charts/releases/tag/kube-prometheus-stack-88.5.3) kube-prometheus-stack collects Kubernetes manifests, Grafana dashboards, and Prometheus rules combined with documentation and scripts to provide easy to operate end-to-end Kubernetes cluster monitoring with Prometheus using the Prometheus Operator. #### What's Changed - \[kube-prometheus-stack] Update ghcr.io/jkroepke/kube-webhook-certgen Docker tag to v1.8.7 by [@​renovate](https://redirect.github.com/renovate)\[bot] in [#​7208](https://redirect.github.com/prometheus-community/helm-charts/pull/7208) **Full Changelog**: <prometheus-community/helm-charts@prometheus-operator-admission-webhook-0.43.2...kube-prometheus-stack-88.5.3> </details> --- ### Configuration 📅 **Schedule**: (in timezone Europe/Berlin) - Branch creation - "after 6pm every weekday,every weekend,before 8am every weekday" - Automerge - At any time (no schedule defined) 🚦 **Automerge**: Enabled. ♻ **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. --- - [ ] <!-- rebase-check -->If you want to rebase/retry this PR, check this box --- This PR was generated by [Mend Renovate](https://mend.io/renovate/). View the [repository job log](https://developer.mend.io/github/cobaltcore-dev/cortex). <!--renovate-debug:eyJjcmVhdGVkSW5WZXIiOiI0NC4zOS4wIiwidXBkYXRlZEluVmVyIjoiNDQuMzkuMCIsInRhcmdldEJyYW5jaCI6Im1haW4iLCJsYWJlbHMiOltdfQ==--> Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
The OpenStack Keystone service catalog contains internal, admin, and public endpoints. While for internal and admin endpoints of the Cortex Placement Shim we can use an internal URL via cluster- internal DNS, for the public endpoint that'll be accessed by clients like the CLI, we need an Ingress resource. Note that Ingress is chosen over HTTPRoute (part of the Gateway API) for now because we didn't migrate there yet. Signed-off-by: Philipp Matthes <p.matthes@sap.com>
This PR contains the following updates: | Package | Type | Update | Change | |---|---|---|---| | debian | | digest | `3a39a05` → `d7e1218` | | debian | final | digest | `3a39a05` → `d7e1218` | --- ### Configuration 📅 **Schedule**: (in timezone Europe/Berlin) - Branch creation - At any time (no schedule defined) - Automerge - At any time (no schedule defined) 🚦 **Automerge**: Disabled by config. Please merge this manually once you are satisfied. ♻ **Rebasing**: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox. 🔕 **Ignore**: Close this PR and you won't be reminded about these updates again. --- - [ ] <!-- rebase-check -->If you want to rebase/retry this PR, check this box --- This PR was generated by [Mend Renovate](https://mend.io/renovate/). View the [repository job log](https://developer.mend.io/github/cobaltcore-dev/cortex). <!--renovate-debug:eyJjcmVhdGVkSW5WZXIiOiI0NC4zOS4wIiwidXBkYXRlZEluVmVyIjoiNDQuMzkuMCIsInRhcmdldEJyYW5jaCI6Im1haW4iLCJsYWJlbHMiOltdfQ==--> Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
The daily CVE scan detected fixable vulnerabilities in the published `cortex-postgres` image. A test rebuild confirms that rebuilding reduces the CVE count (via `apt-get upgrade` picking up security patches). Merging this PR triggers the image rebuild and publish pipeline. This PR was created automatically by the `rebuild-postgres` workflow. Co-authored-by: umswmayj <140147670+umswmayj@users.noreply.github.com>
Adds a transparent, in-process write-through cache that masks informer
lag: after a write the controller can immediately Get/List the object
even before the informer has caught up.
So what this cache does is:
- On Create/Update/Patch stores an entry in the cache (if the gvk is
configured); on Delete sets a tombstone.
- Reads merge the overlay with the informer result: a tombstone yields
NotFound, a live entry wins over (or supplements) the informer, and
List/MatchingFields include pending entries.
- Entries are evicted once the real object shows up in the informer at a
resourceVersion ≥ the cached one, or after a TTL as a safety net.
This is integrated into the multicluster client:
- multicluster.Client wraps each home/remote cluster.Cluster via
pendingcache.WrapCluster, so every cluster gets its own overlay +
informer and eviction stays cluster-local. Routing is unchanged -
reads/writes/IndexField just flow through the wrapped cluster's
GetClient()/GetFieldIndexer().
- The overlay's lifecycle (eviction handlers + TTL cleanup) is a
manager.Runnable (NeedLeaderElection = false) added alongside the raw
cluster, so the informer is started exactly once.
The gvks that are supposed to be cached and other config options are
listed below:
```yml
cache:
enabled: false # opt-in; off ⇒ raw clusters, zero overhead
gvks: [cortex.cloud/v1alpha1/Reservation] # only these GVKs are overlaid; others pass through
ttl: 2m # max entry lifetime (default 2m)
```
```mermaid
flowchart TB
Ctrl["Controller"]
MCL["multicluster.Client\n(routing / fan-out)"]
subgraph home["overlayCluster (home)"]
OH["Overlay (cache)"]
CH["cluster + informer"]
OH --> CH
end
subgraph rem["overlayCluster (remote N)"]
OR["Overlay (cache)"]
CR["cluster + informer"]
OR --> CR
end
Ctrl --> MCL
MCL --> OH
MCL --> OR
CH -. "informer events → evict" .-> OH
CR -. "informer events → evict" .-> OR
```
---------
Signed-off-by: Markus Wieland <markus.wieland@sap.com>
Co-authored-by: mblos <156897072+mblos@users.noreply.github.com>
Adds a high-watermark size metric for the pending-cache overlay. Because
overlay entries are evicted within milliseconds, a plain gauge would
almost always read zero and give no signal. So `cache.Monitor` tracks
the peak entry count per (gvk, host) between scrapes and resets it on
each Collect, alongside the current size at scrape.
- `cortex_cache_overlay_entries_max{gvk,host}` - high-watermark since
last scrape (reset on scrape)
- `cortex_cache_overlay_entries{gvk,host} ` - current entry count
With that we can track that if the eviction is working as intended or if
there is a memory leak
---------
Signed-off-by: Markus Wieland <markus.wieland@sap.com>
This PR contains the following updates: | Package | Type | Update | Change | |---|---|---|---| | [github.com/sapcc/go-bits](https://redirect.github.com/sapcc/go-bits) | require | digest | `75bdd20` → `2ac89aa` | --- ### Configuration 📅 **Schedule**: (in timezone Europe/Berlin) - Branch creation - "after 6pm every weekday,every weekend,before 8am every weekday" - Automerge - At any time (no schedule defined) 🚦 **Automerge**: Enabled. ♻ **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. --- - [ ] <!-- rebase-check -->If you want to rebase/retry this PR, check this box --- This PR was generated by [Mend Renovate](https://mend.io/renovate/). View the [repository job log](https://developer.mend.io/github/cobaltcore-dev/cortex). <!--renovate-debug:eyJjcmVhdGVkSW5WZXIiOiI0NC40Ni4wIiwidXBkYXRlZEluVmVyIjoiNDQuNDYuMCIsInRhcmdldEJyYW5jaCI6Im1haW4iLCJsYWJlbHMiOltdfQ==--> Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
Adds a `physical_host_size` label to the VMware host capacity metrics so hosts can be filtered by their physical memory size category (e.g. 4TiB, 8TiB). A VMware building block groups multiple physical hosts into a single placement resource provider, so the per-host size is recovered from the pooled memory inventory using `available_ram = total - reserved` `amount_hosts = available_ram / max_unit` `physical_host_size = max_unit + reserved / amount_hosts`. The size is computed in the host-details extractor, rounded to whole GiB, and reported as.. .. TiB when it reaches 1024 GiB .. or GiB otherwise (falling back to unknown when inventory is missing). Signed-off-by: Markus Wieland <markus.wieland@sap.com>
Rebrands the `host-details` knowledge to `vmware-host-details` and scopes it to VMware hosts only. KVM hosts are served by a dedicated CRD and ironic hosts are not needed, so both are now excluded directly in the extractor SQL rather than filtered in the KPIs. --------- Signed-off-by: Markus Wieland <markus.wieland@sap.com>
#1170) `go mod tidy` bumped go.mod to `1.27` as a side effect of a go-bits update, but the Dockerfile and Renovate config were still locked to 1.26, breaking Docker builds. I don't see any reason why we should limit the go version at all, so I removed these constraints. Signed-off-by: Markus Wieland <markus.wieland@sap.com>
The new HostOversubscriptionController adds automated detection and eviction of oversubscribed host slots, but operators have no documentation to understand its modes, config, metrics, or alerts. This adds a section to the CR reservations doc covering operational modes (disabled, detection-only, active remediation), grace period behavior, eviction target selection strategy, and the two new Prometheus metrics and alerts. Assisted-by: Claude Code:claude-sonnet-4-20250514 [Bash] [Read] Signed-off-by: cortex-ai-agents[bot] <279748396+cortex-ai-agents[bot]@users.noreply.github.com> Co-authored-by: cortex-ai-agents[bot] <279748396+cortex-ai-agents[bot]@users.noreply.github.com>
This PR contains the following updates: | Package | Type | Update | Change | |---|---|---|---| | [github.com/sapcc/go-bits](https://redirect.github.com/sapcc/go-bits) | require | digest | `2ac89aa` → `7669cbd` | --- ### Configuration 📅 **Schedule**: (in timezone Europe/Berlin) - Branch creation - "after 6pm every weekday,every weekend,before 8am every weekday" - Automerge - At any time (no schedule defined) 🚦 **Automerge**: Enabled. ♻ **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. --- - [ ] <!-- rebase-check -->If you want to rebase/retry this PR, check this box --- This PR was generated by [Mend Renovate](https://mend.io/renovate/). View the [repository job log](https://developer.mend.io/github/cobaltcore-dev/cortex). <!--renovate-debug:eyJjcmVhdGVkSW5WZXIiOiI0NC40Ni4wIiwidXBkYXRlZEluVmVyIjoiNDQuNDYuMCIsInRhcmdldEJyYW5jaCI6Im1haW4iLCJsYWJlbHMiOltdfQ==--> Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
#1163) Cortex does not yet support Nova's forced-destination behavior. This causes problems for flows like onboarding tests that force a specific host, because: - Cortex ignores the forced host, and - Cortex then filters the host out (e.g. because it is in onboarding), so the VM can never land on the forced host. This PR replicates Nova's forced-destination behavior (see [sapcc/nova](https://github.com/sapcc/nova/blob/05f384a938e3d6a8740a8f404d79d767d6ebdbd7/nova/scheduler/host_manager.py#L610-L620)). When `force_hosts` or `force_nodes` is set in the request spec and no `_nova_check_type` is present in scheduler_hints, the filter/weigher pipeline is skipped entirely and only the forced destinations are returned. I added a kill switch to the config `forcedDestinationEnabled`. That way, if something breaks we can easily disable it again without doing an emergency deployment. --------- Signed-off-by: Markus Wieland <markus.wieland@sap.com> Signed-off-by: Markus Wieland <wie.markus@web.de> Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
This PR contains the following updates: | Package | Change | [Age](https://docs.renovatebot.com/merge-confidence/) | [Adoption](https://docs.renovatebot.com/merge-confidence/) | [Passing](https://docs.renovatebot.com/merge-confidence/) | [Confidence](https://docs.renovatebot.com/merge-confidence/) | Type | Update | |---|---|---|---|---|---|---|---| | [github.com/golangci/golangci-lint/v2](https://redirect.github.com/golangci/golangci-lint) | `v2.13.1` → `v2.13.2` |  |  |  |  | | patch | | [github.com/ironcore-dev/ironcore](https://redirect.github.com/ironcore-dev/ironcore) | `v0.5.0` → `v0.6.1` |  |  |  |  | require | minor | | [k8s.io/api](https://redirect.github.com/kubernetes/api) | `v0.36.4` → `v0.37.0` |  |  |  |  | require | minor | | [k8s.io/apimachinery](https://redirect.github.com/kubernetes/apimachinery) | `v0.36.4` → `v0.37.0` |  |  |  |  | require | minor | | [k8s.io/client-go](https://redirect.github.com/kubernetes/client-go) | `v0.36.4` → `v0.37.0` |  |  |  |  | require | minor | | [kube-prometheus-stack](https://redirect.github.com/prometheus-operator/kube-prometheus) ([source](https://redirect.github.com/prometheus-community/helm-charts)) | `88.5.4` → `88.6.0` |  |  |  |  | | minor | --- ### Release Notes <details> <summary>golangci/golangci-lint (github.com/golangci/golangci-lint/v2)</summary> ### [`v2.13.2`](https://redirect.github.com/golangci/golangci-lint/blob/HEAD/CHANGELOG.md#v2132) [Compare Source](https://redirect.github.com/golangci/golangci-lint/compare/v2.13.1...v2.13.2) *Released on 2026-08-28* 1. Bug fixes - Decrease cache entropy 2. Linters bug fixes - `iface`: from 1.5.0 to 1.5.1 - `staticcheck`: from 0.8.0 to 0.8.1 - `unparam`: from [`3f964bc`](https://redirect.github.com/golangci/golangci-lint/commit/3f964bcb5673) to [`2fa3d84`](https://redirect.github.com/golangci/golangci-lint/commit/2fa3d841b0c8) - `canonicalheader`: from v1.1.2 to a temporary fork </details> <details> <summary>ironcore-dev/ironcore (github.com/ironcore-dev/ironcore)</summary> ### [`v0.6.1`](https://redirect.github.com/ironcore-dev/ironcore/releases/tag/v0.6.1) [Compare Source](https://redirect.github.com/ironcore-dev/ironcore/compare/v0.6.0...v0.6.1) #### Changes #### 🐛 Bug Fixes - Add mutating admission policy RBAC to apiserver role [@​lukasfrank](https://redirect.github.com/lukasfrank) ([#​1550](https://redirect.github.com/ironcore-dev/ironcore/issues/1550)) - Fix inverted CSR recognizer check in certificate approval controller [@​SamiraBriongos](https://redirect.github.com/SamiraBriongos) ([#​1547](https://redirect.github.com/ironcore-dev/ironcore/issues/1547)) ### [`v0.6.0`](https://redirect.github.com/ironcore-dev/ironcore/releases/tag/v0.6.0) [Compare Source](https://redirect.github.com/ironcore-dev/ironcore/compare/v0.5.0...v0.6.0) #### Changes #### 🚀 Features - Add `pool-lifecycle-controller` bootstrap identity [@​lukasfrank](https://redirect.github.com/lukasfrank) ([#​1544](https://redirect.github.com/ironcore-dev/ironcore/issues/1544)) - Enhance scheduling to respect `MachinePool` `Ready` condition [@​lukasfrank](https://redirect.github.com/lukasfrank) ([#​1536](https://redirect.github.com/ironcore-dev/ironcore/issues/1536)) - Upgrade `k8s.io/*` deps to v0.36.x [@​afritzler](https://redirect.github.com/afritzler) ([#​1504](https://redirect.github.com/ironcore-dev/ironcore/issues/1504)) #### 🐛 Bug Fixes - Use volume provider ID instead of volume name for volume snapshot [@​gonzolino](https://redirect.github.com/gonzolino) ([#​1543](https://redirect.github.com/ironcore-dev/ironcore/issues/1543)) - Use 'go list' directly to avoid an external 'jq' dependency in codegen script [@​ushabelgur](https://redirect.github.com/ushabelgur) ([#​1540](https://redirect.github.com/ironcore-dev/ironcore/issues/1540)) - Stop rewriting machine conditions on every reconcile [@​lukasfrank](https://redirect.github.com/lukasfrank) ([#​1532](https://redirect.github.com/ironcore-dev/ironcore/issues/1532)) - Fix allocatable calculation in poollets [@​ushabelgur](https://redirect.github.com/ushabelgur) ([#​1529](https://redirect.github.com/ironcore-dev/ironcore/issues/1529)) #### 🧰 Maintenance - Bump golang from 1.26 to 1.27 in the docker-dependencies group across 1 directory @​[dependabot\[bot\]](https://redirect.github.com/apps/dependabot) ([#​1546](https://redirect.github.com/ironcore-dev/ironcore/issues/1546)) - Bump the go-dependencies group with 2 updates @​[dependabot\[bot\]](https://redirect.github.com/apps/dependabot) ([#​1545](https://redirect.github.com/ironcore-dev/ironcore/issues/1545)) - Bump the go-dependencies group with 3 updates @​[dependabot\[bot\]](https://redirect.github.com/apps/dependabot) ([#​1539](https://redirect.github.com/ironcore-dev/ironcore/issues/1539)) - Bump google.golang.org/grpc from 1.82.1 to 1.83.0 in the go-dependencies group across 1 directory @​[dependabot\[bot\]](https://redirect.github.com/apps/dependabot) ([#​1535](https://redirect.github.com/ironcore-dev/ironcore/issues/1535)) - Bump actions/stale from 10 to 11 in the gh-actions-dependencies group @​[dependabot\[bot\]](https://redirect.github.com/apps/dependabot) ([#​1534](https://redirect.github.com/ironcore-dev/ironcore/issues/1534)) </details> <details> <summary>kubernetes/api (k8s.io/api)</summary> ### [`v0.37.0`](https://redirect.github.com/kubernetes/api/compare/v0.36.4...v0.37.0) [Compare Source](https://redirect.github.com/kubernetes/api/compare/v0.36.4...v0.37.0) </details> <details> <summary>kubernetes/apimachinery (k8s.io/apimachinery)</summary> ### [`v0.37.0`](https://redirect.github.com/kubernetes/apimachinery/compare/v0.36.4...v0.37.0) [Compare Source](https://redirect.github.com/kubernetes/apimachinery/compare/v0.36.4...v0.37.0) </details> <details> <summary>kubernetes/client-go (k8s.io/client-go)</summary> ### [`v0.37.0`](https://redirect.github.com/kubernetes/client-go/compare/v0.37.0-rc.1...v0.37.0) [Compare Source](https://redirect.github.com/kubernetes/client-go/compare/v0.36.4...v0.37.0) </details> <details> <summary>prometheus-community/helm-charts (kube-prometheus-stack)</summary> ### [`v88.6.0`](https://redirect.github.com/prometheus-community/helm-charts/releases/tag/kube-prometheus-stack-88.6.0) [Compare Source](https://redirect.github.com/prometheus-community/helm-charts/compare/kube-prometheus-stack-88.5.4...kube-prometheus-stack-88.6.0) kube-prometheus-stack collects Kubernetes manifests, Grafana dashboards, and Prometheus rules combined with documentation and scripts to provide easy to operate end-to-end Kubernetes cluster monitoring with Prometheus using the Prometheus Operator. ##### What's Changed - \[kube-prometheus-stack] Add HTTPRoute timeout support by [@​U-S-jun](https://redirect.github.com/U-S-jun) in [#​7221](https://redirect.github.com/prometheus-community/helm-charts/pull/7221) **Full Changelog**: <prometheus-community/helm-charts@prometheus-redis-exporter-6.30.0...kube-prometheus-stack-88.6.0> </details> --- ### Configuration 📅 **Schedule**: (in timezone Europe/Berlin) - Branch creation - "after 6pm every weekday,every weekend,before 8am every weekday" - Automerge - At any time (no schedule defined) 🚦 **Automerge**: Enabled. ♻ **Rebasing**: Whenever PR is behind base branch, or you tick the rebase/retry checkbox. 👻 **Immortal**: This PR will be recreated if closed unmerged. Get [config help](https://redirect.github.com/renovatebot/renovate/discussions) if that's undesired. --- - [ ] <!-- rebase-check -->If you want to rebase/retry this PR, check this box --- This PR was generated by [Mend Renovate](https://mend.io/renovate/). View the [repository job log](https://developer.mend.io/github/cobaltcore-dev/cortex). <!--renovate-debug:eyJjcmVhdGVkSW5WZXIiOiI0NC40Ni4wIiwidXBkYXRlZEluVmVyIjoiNDQuNDYuMCIsInRhcmdldEJyYW5jaCI6Im1haW4iLCJsYWJlbHMiOltdfQ==--> Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> Co-authored-by: Markus Wieland <markus.wieland@sap.com>
The forced-destination shortcut in the Nova external scheduler was placed after the canRunScheduler validation, which rejects requests with missing or incomplete weights. Nova does not always send weights for forced-destination requests (force_hosts/force_nodes), so they would be incorrectly rejected with HTTP 400 before the forced-destination bypass could trigger. Moving the forced-destination check earlier ensures forced requests are handled before weight validation runs. A new test case verifies that forced-destination requests with no weights succeed instead of being rejected. Assisted-by: Claude Code:claude-sonnet-4-20250514 [Bash] [Read] Signed-off-by: cortex-ai-agents[bot] <279748396+cortex-ai-agents[bot]@users.noreply.github.com> Co-authored-by: cortex-ai-agents[bot] <279748396+cortex-ai-agents[bot]@users.noreply.github.com>
The pending cache is a major subsystem (~2000 lines across 7 files in pkg/cache/) that solves informer lag for configurable GVKs by overlaying writes on the client.Client interface, integrated via the multicluster ClusterWrapper pattern. It has per-object keyed mutexes, tombstone logic for deletes, informer-driven eviction, TTL fallback, field index awareness, and its own Prometheus monitoring with a PrometheusRule alert. None of this was documented anywhere, leaving readers with no orientation for what the subsystem does, why it exists, or how it integrates with the multicluster client and monitoring stack. This adds a focused docs page covering the overlay's purpose, lifecycle, integration, configuration, and monitoring. Assisted-by: Claude Code:claude-sonnet-4-20250514 [Bash] [Read] --------- Signed-off-by: cortex-ai-agents[bot] <279748396+cortex-ai-agents[bot]@users.noreply.github.com> Signed-off-by: Markus Wieland <wie.markus@web.de> Co-authored-by: cortex-ai-agents[bot] <279748396+cortex-ai-agents[bot]@users.noreply.github.com> Co-authored-by: Markus Wieland <markus.wieland@sap.com>
…#1174) The forced destination feature replicates Nova's native behavior of bypassing the scheduling pipeline when force_hosts or force_nodes is set. The architecture doc describes the scheduling pipeline but does not mention this bypass, leaving operators unaware of the behavior or how to disable it via the forcedDestinationEnabled config flag. This adds a short subsection under External Scheduler Delegation covering when the bypass triggers, the _nova_check_type exception, the config kill-switch, and a code pointer. Assisted-by: Claude Code:claude-sonnet-4-20250514 [Bash] [Read] --------- Signed-off-by: cortex-ai-agents[bot] <279748396+cortex-ai-agents[bot]@users.noreply.github.com> Signed-off-by: Markus Wieland <wie.markus@web.de> Co-authored-by: cortex-ai-agents[bot] <279748396+cortex-ai-agents[bot]@users.noreply.github.com> Co-authored-by: Markus Wieland <markus.wieland@sap.com>
Collaborator
Author
|
@claude pls run the /release 1777 command |
SoWieMarkus
marked this pull request as ready for review
August 31, 2026 07:26
SoWieMarkus
requested review from
PhilippMatthes,
auhlig,
juliusclausnitzer,
mblos and
umswmayj
as code owners
August 31, 2026 07:26
Contributor
|
Claude finished @SoWieMarkus's task in 1m 36s —— View job
|
Collaborator
Author
|
@claude pls run the /release 1177 command |
Contributor
|
Claude finished @SoWieMarkus's task in 17m 12s —— View job Release Orchestrator — PR #1177 Complete ✅
Release #1177 Post-Open Summary
|
Merged
## Summary - Release prep for #1177: changelog entry and helm chart version bumps - Merge this before merging #1177 ## Changes - Updated `CHANGELOG.md` with release notes for cortex 0.4.0 - Bumped chart versions across 11 Helm Chart.yaml files (bundles, library, and dev charts) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: cortex-ai-agents[bot] <279748396+cortex-ai-agents[bot]@users.noreply.github.com> Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
Contributor
Test Coverage ReportTest Coverage 📊: 72.4% |
bump app version [skip ci] ``` bumped cortex: sha-23f975db -> sha-91df27a0 bumped cortex-shim: sha-23f975db -> sha-17e0a4fe bumped cortex-postgres: sha-23f975db -> sha-cf70e178 ``` Signed-off-by: SoWieMarkus <44964229+SoWieMarkus@users.noreply.github.com> Co-authored-by: SoWieMarkus <44964229+SoWieMarkus@users.noreply.github.com>
PhilippMatthes
approved these changes
Aug 31, 2026
juliusclausnitzer
approved these changes
Aug 31, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Release cortex 0.4.0
cortex v0.4.0 (sha-23f975db)
New features:
client.Clientthat eliminates informer lag for configurable GVKs; after a successful write the overlay keeps the object in memory so subsequent reads return the fresh version immediately instead of the stale informer snapshot; includes per-object keyed mutex, tombstone tracking, field selector support, and fallback TTL expiry (#1042)cortex_cache_overlay_entries(current count) andcortex_cache_overlay_entries_max(high-watermark) Prometheus gauge metrics per GVK/host (#1160)force_hosts/force_nodesbehavior; when a scheduling request specifies forced destinations and no_nova_check_typehint is set, Cortex skips the entire filter/weigher pipeline and returns only the forced hosts directly; configurable viaforcedDestinationEnabled(#1163)canRunScheduler— critical because Nova may not send weights for forced requests (#1176)physical_host_sizelabel derived from memory inventory (e.g."4TiB","512GiB") to VMware host capacity metrics (#1165)Breaking changes:
host-detailstovmware-host-details— existing Knowledge CRDs with the old name will be orphaned and must be deleted manually; extractor renamed fromsap_host_details_extractortovmware_host_details_extractor; SQL rewritten to filter at the query level, eliminating Go-side hypervisor type/family filtering (#1166)Non-breaking changes:
k8s.io/api,k8s.io/apimachinery,k8s.io/client-goto v0.37.0 (#1169)github.com/sapcc/go-bits(#1168, #1171)kube-prometheus-stackto v88.6.0 (#1152)cortex-shim v0.1.15 (sha-23f975db)
New features:
ingress.yamltemplate with configurable className, annotations, rules, and TLS support (#1158)Includes updated dependencies (Go 1.27, k8s.io v0.37.0).
cortex-postgres v0.6.15 (sha-23f975db)
debian:trixie-slimbase image digest (#1159)cortex-nova v0.0.89
CortexNovaCacheOverlayNotDraining(5m window, warning severity)cache.enabled(true),cache.gvks,forcedDestinationEnabled(true)host-details→vmware-host-detailsIncludes updated charts cortex v0.4.0 and cortex-postgres v0.6.15.
cortex-cinder v0.0.89
Includes updated charts cortex v0.4.0 and cortex-postgres v0.6.15.
cortex-manila v0.0.89
Includes updated charts cortex v0.4.0 and cortex-postgres v0.6.15.
cortex-crds v0.0.89
Includes updated chart cortex v0.4.0.
cortex-ironcore v0.0.89
Includes updated chart cortex v0.4.0.
cortex-pods v0.0.89
Includes updated chart cortex v0.4.0.
cortex-placement-shim v0.1.15
Includes updated chart cortex-shim v0.1.15.
Dependencies