Shipped · Observability

What cortexproject/cortex shipped

The public repository of Cortexmetrics · github.com/cortexproject/cortex

Written by FoxPlug from the repository's own releases; not affiliated with Cortexmetrics. Paste your own repository in the box below and FoxPlug writes the same weekly update for your product. This page is an automatic summary of the releases, pull requests and commits in github.com/cortexproject/cortex; Cortexmetrics did not write it and does not use or endorse FoxPlug. Every line links to the change it describes.

Get a weekly update like this for your product, free

Or Use it as a GitHub Action

See the update FoxPlug writes for cortex, free

Week of November 2, 2026

What shipped

Why it matters

This week brings a new storage option for anyone running blocks, ruler or alertmanager, and a ruler change that stops alert notifications from being silently dropped when the Alertmanager does not accept them. It also keeps 1.21 configs that use the old query range and HA failover timeout settings working, and adds a per-zone ownership metric for ring users.

Changelog entry

Example posts FoxPlug drafted from these changes. Not written or posted by the project.

Post for X

This week in Cortex: Alibaba Cloud OSS is now a storage backend for blocks, ruler and alertmanager; the ruler retries notifications Alertmanager did not accept; old 1.21 query range and HA failover timeout configs keep working; and compaction delay is configurable per block range

Post for LinkedIn

This week's changes in Cortex: Alibaba Cloud OSS is now available as a storage backend for blocks, ruler and alertmanager, selected with backend: oss. The ruler now retries notifications the Alertmanager failed to accept instead of dropping them. Configs that set the query range or HA failover timeout settings under their old 1.21 names keep working instead of breaking on upgrade. The compaction group graceful period can now be configured per block range with compactor.block-ranges-compaction-delay, and a per-zone ring member ownership metric was added. The Go base image moved to 1.27.3 for security fixes, and the docs gained a guide on setting up UTF-8 mode.

Week of October 19, 2026

What shipped

Why it matters

This week's changes fix sharding in the store-gateway's Parquet bucket store, prevent unnecessary retries of ruler staleness markers, and stop miscounting them as discarded samples. Dependency upgrades to promql-engine and Thanos bring lazier allocation in the query path, and the Alertmanager update applies upstream inhibit rule fixes.

Changelog entry

Example posts FoxPlug drafted from these changes. Not written or posted by the project.

Post for X

This week: store-gateway Parquet store now honors query sharding; ruler no longer retries unaccepted staleness markers; ingester stops counting them as discarded samples; promql-engine and Thanos upgraded; Alertmanager bumped to v0.34.1 for inhibit rule fixes.

Post for LinkedIn

This week in Cortex: the store-gateway's Parquet bucket store now honors query sharding in Series requests. The ruler stops retrying staleness markers that can never be accepted, and the ingester no longer counts those markers as discarded tenant samples. We also upgraded promql-engine and Thanos to pick up query-path changes that allocate and release iterators lazily, and updated Alertmanager to v0.34.1 to apply upstream inhibit rule bug fixes. Benchmark configuration was updated to include ParquetQueryConcurrency.

Week of October 5, 2026

What shipped

Why it matters

The compactor no longer reports the final cleanup of a tenant marked for deletion as failed on object stores such as GCS, Azure, Swift and OCI that return an error when deleting a missing object. The published Go modules SBOM is now generated from a clean checkout of the release tag, so it no longer includes untracked or gitignored files from the working directory. The fourth release candidate for v1.22.0 packages these fixes for testing.

Changelog entry

Example posts FoxPlug drafted from these changes. Not written or posted by the project.

Post for X

V1.22.0-rc.3 is out. It fixes compactor tenant cleanup being reported as failed on GCS, Azure, Swift and OCI, and generates the Go modules SBOM from a clean checkout of the release tag so untracked and gitignored files stay out of it.

Post for LinkedIn

V1.22.0-rc.3 is available, the fourth release candidate for v1.22.0. It contains two fixes on top of v1.22.0-rc.2: the compactor no longer reports the final cleanup of a tenant marked for deletion as failed on object stores that return an error when deleting a missing object, such as GCS, Azure, Swift and OCI, and the Go modules SBOM is now generated from a clean checkout of the release tag so untracked and gitignored files no longer end up in the published SBOM. Both fixes were backported to release-1.22.

Week of September 28, 2026

What shipped

Why it matters

This week includes a significant new feature for ruler federation and critical fixes for compactor cleanup and test reliability. The tenant cleanup bug fix addresses production issues across multiple cloud storage backends, while improved test initialization prevents flaky test failures that delay releases.

Changelog entry

Example posts FoxPlug drafted from these changes. Not written or posted by the project.

Post for X

This week: ruler tenant federation ships, compactor cleanup bugs fixed on GCS/Azure/Swift/OCI, and integration test reliability improvements to prevent race conditions during MinIO initialization.

Post for LinkedIn

Cortex progressed on multiple fronts this week. Ruler tenant federation is now available, enabling rule groups to evaluate against multiple tenants while writing results to the owning tenant. Critical bugs in compactor cleanup on GCS, Azure, Swift and OCI were fixed where final cleanup was incorrectly reported as failed. Test infrastructure improvements address race conditions in MinIO initialization that were causing flaky test runs.

Week of September 21, 2026

What shipped

Why it matters

V1.22.0-rc.2 addresses a critical bug in the new ingestion delay histogram metric and resolves infrastructure issues blocking integration tests. The Alertmanager upgrade and MinIO sourcing fix improve stability and compatibility for users running Cortex with these components.

Changelog entry

Example posts FoxPlug drafted from these changes. Not written or posted by the project.

Post for X

V1.22.0-rc.2 is out with a fix for the ingestion_delay_seconds histogram and updated MinIO sourcing for integration tests.

Post for LinkedIn

V1.22.0-rc.2 released: fixes cortex_ingester_ingestion_delay_seconds histogram observations, upgrades Alertmanager to v0.34.0, and resolves MinIO image sourcing to unblock integration tests. Native histogram verification and Parquet converter error message improvements also included.

Week of September 14, 2026

What shipped

Why it matters

This week addresses operational stability with fixes to ingester compaction, tenant federation, and cache handling that prevent silent failures and startup issues. API improvements add pagination for large rule sets, and memory optimizations reduce heap pressure on high-cardinality queries.

Changelog entry

Example posts FoxPlug drafted from these changes. Not written or posted by the project.

Post for X

Shipping fixes to ingesters, ruler pagination, promql-engine upgrade, and memory optimizations for query processing.

Post for LinkedIn

This week's Cortex updates focus on operational stability and performance: ListRules API now supports pagination for large rulesets; upgraded promql-engine with improved limit tracking; ingesters handle block shipping gracefully; removed duplicate series expiry; fixed tenant federation and regex resolver cache issues; memory optimizations for high-cardinality queries.

Weeks with too little public activity are left out rather than filled in. Last updated 2026-11-10.

Is this your repo? Ask us to remove this page.