diff --git a/CHANGELOG.md b/CHANGELOG.md index 2bded5760a5..e79363625cc 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -1,5 +1,51 @@ # Fluid Release Notes +## v1.1.0 +### Highlights +* **CacheRuntime – a generic cache engine framework** – Two new CRDs, `CacheRuntime` and `CacheRuntimeClass`, let you onboard a cache system declaratively instead of writing a new Go engine. Describe the cluster topology (`MasterSlave` / `P2P-DHT` / `ClientOnly`), the per-component workload form and the operation commands in a `CacheRuntimeClass`, and the shared controller handles reconciliation, scheduling and lifecycle. See the [CacheRuntime Integration Guide](docs/en/dev/generic_cache_runtime_integration.md). +* **Curvine support via CacheRuntime** – Curvine is the first cache system integrated through the new framework, with a native image, cache-state reporting and dataset `files`/`UFSTotal` synchronization. +* **CacheRuntime data operations** – DataLoad, DataProcess and manual scaling work against CacheRuntime-backed Datasets, alongside tiered store configuration for worker and client components. +* **In-place update for CacheRuntime** – Backed by AdvancedStatefulSet, image-version and other spec changes roll out without recreating Pods. Field-level update semantics are documented in [CacheRuntime Spec Field Update Capabilities](docs/en/samples/cacheruntime/cacheruntime_spec_update.md). +* **DataProcess scheduling policies** – `Cron` and `OnEvent` policies allow recurring and event-driven data processing. +* **RDMA for JindoCache** – JindoCache engine can now use RDMA for higher cache read throughput. +* **ThinRuntime over CacheRuntime** – ThinRuntime can consume a CacheRuntime-backed Dataset, with validation that rejects invalid combinations. + +### Enhancements +* **Feature gates for FUSE host PID** – Host PID behaviour is now controlled by a feature gate rather than an implicit chart default. +* **Global runtime sync switch** – A cluster-level option can skip runtime syncing to reduce control-plane load. +* **JuiceFSRuntime worker `volumeClaimTemplates`** – Workers can claim persistent volumes directly. +* **Extended resource names on runtime Pods** – More resource names can be requested on master/worker/FUSE Pods. +* **JindoCache master self-healing** – The master recovers automatically after an unexpected crash. +* **CSI recovery loop** – Improved FUSE mount recovery, plus a fast path for resolving the mount Pod's node-selector label key. +* **Security hardening** – Least-privilege ClusterRole for the CacheRuntime controller, least-privilege GitHub Actions permissions across workflows, and enforced CPU/memory/storage limits in e2e manifests. +* **Dependency and base-image updates** – Helm 3.19.5 and Alpine 3.23.3. +* **Portable container images** – The hardcoded `Asia/Shanghai` timezone was removed from all images. +* **Developer experience** – Go modules replace the GOPATH requirement, Helm binaries were removed from the source tree, and a large share of the unit-test suite migrated from testify/gohook to Ginkgo/Gomega and gomonkey. Backward-compatibility e2e tests were added. +* **2026 roadmap** – Published in [ROADMAP.md](ROADMAP.md). + +### Bug Fixes +* Recover the GroupVersionKind of an owner whose TypeMeta is incomplete. +* Recreate the ThinRuntime of a reference Dataset stuck in `NotBound`. +* Prevent nodeSelector changes on the JuiceFS FUSE DaemonSet. +* Support multi-mount OSS secret projections in Jindo. +* Merge Dataset annotations across multi-round FUSE injection. +* Implement the missing `Validate` logic for `ReferenceDatasetEngine`. +* Implement `OnEventStatusHandler.GetOperationStatus` for DataLoad and DataMigrate. +* Watch Job resources in the DataLoad and DataMigrate controllers. +* Retry on conflict when updating `ufsTotal` status, and retry subpath/mount checks in the webhook. +* Preserve required-affinity branch semantics in the webhook. +* Fix Pod creation failure when mounting a Dataset with a long name. +* Prevent controller panic on a wrapped NotFound error in `OperationReconciler`. +* Handle the `updatedb.conf` path being a directory in the CSI plugin. +* Fix a data race in `ExecCommandInContainerWithTimeout`. +* Remove redundant UFS readiness checks in JindoCache and add an early runtime check. +* Correct cache client status reporting and propagate context through volume, ConfigMap, exec and dataflow-affinity paths. + +### Deprecations and Removals +* **GooseFSRuntime is deprecated** – The engine and its related resources have been removed. Migrate to another runtime before upgrading. + +--- + ## v1.0.8 ### Highlights * **Native Sidecar support** – Enable Kubernetes Native Sidecar injection for FUSE deployments to resolve lifecycle management, startup order, and resource isolation issues. diff --git a/README-zh_CN.md b/README-zh_CN.md index e02926b8316..17615b5d04a 100644 --- a/README-zh_CN.md +++ b/README-zh_CN.md @@ -16,9 +16,9 @@ |![更新](static/bell-outline-badge.svg) 最新进展：| |------------------| |**🎉 Fluid 正式晋升为 CNCF 孵化项目：** 2026年1月8日, CNCF 技术监督委员会（TOC）已投票通过，将 Fluid 纳入 CNCF 孵化阶段——这是项目成熟度和社区影响力的重要里程碑。| +|v1.1.0版发布：TBD，Fluid v1.1.0 发布！本版本引入 **CacheRuntime**，以声明式方式对接各类缓存系统。版本更新介绍详情参见 [CHANGELOG](CHANGELOG.md)。| |v1.0.8版发布：2025年10月31日，Fluid v1.0.8 发布！版本更新介绍详情参见 [CHANGELOG](CHANGELOG.md)。| |v1.0.7版发布：2025年9月21日，Fluid v1.0.7 发布！版本更新介绍详情参见 [CHANGELOG](CHANGELOG.md)。| -|v1.0.6版发布：2025年7月12日，Fluid v1.0.6 发布！版本更新介绍详情参见 [CHANGELOG](CHANGELOG.md)。| |进入CNCF：2021年4月27日, Fluid通过CNCF Technical Oversight Committee (TOC)投票决定被接受进入CNCF，成为[CNCF Sandbox Project](https://lists.cncf.io/g/cncf-toc/message/5822)。| ## 什么是Fluid diff --git a/README.md b/README.md index f63516fbc29..1eefe23a79c 100644 --- a/README.md +++ b/README.md @@ -33,9 +33,9 @@ English | [简体中文](./README-zh_CN.md) | ![notification](static/bell-outline-badge.svg) What is NEW! | | ------------------------------------------------------------ | |**🎉 January 8, 2026**: Fluid has been promoted to **CNCF Incubating** status! The CNCF Technical Oversight Committee (TOC) officially accepted Fluid as an incubating project, a major milestone in its maturity and community adoption.| +|v1.1.0 Release: TBD. Fluid v1.1.0 introduces **CacheRuntime**, a generic framework for onboarding cache systems declaratively. Please check the [CHANGELOG](CHANGELOG.md) for details. | |v1.0.8 Release: Oct. 31st, 2025. Fluid v1.0.8. Please check the [CHANGELOG](CHANGELOG.md) for details. | |v1.0.7 Release: Sep. 21st, 2025. Fluid v1.0.7. Please check the [CHANGELOG](CHANGELOG.md) for details. | -|v1.0.6 Release: Jul. 12th, 2025. Fluid v1.0.6. Please check the [CHANGELOG](CHANGELOG.md) for details. | | Apr. 27th, 2021. Fluid accepted by **CNCF**! Fluid project was [accepted as an official CNCF Sandbox Project](https://lists.cncf.io/g/cncf-toc/message/5822) by CNCF Technical Oversight Committee (TOC) with a majority vote after the review process. New beginning for Fluid! . |
diff --git a/docs/en/TOC.md b/docs/en/TOC.md index 2dc88c733aa..6d0f3b3de67 100644 --- a/docs/en/TOC.md +++ b/docs/en/TOC.md @@ -38,6 +38,7 @@ + Advanced - [Accelerate Data Access by MEM or SSD](samples/accelerate_data_by_mem_or_ssd.md) - [Alluxio Tieredstore Configuration](samples/tieredstore_config.md) + - [CacheRuntime TieredStore Configuration](userguide/cache_runtime_tieredstore.md) - [Alluxio S3 High-Concurrency Read Tuning](samples/alluxio_s3_high_concurrency.md) - [Pod Scheduling Optimization](operation/pod_schedule_optimization.md) - [Pod Scheduling Base on Runtime Tiered Locality](operation/tiered_locality_schedule.md) diff --git a/docs/zh/TOC.md b/docs/zh/TOC.md index d00a1825043..4910956ca5e 100644 --- a/docs/zh/TOC.md +++ b/docs/zh/TOC.md @@ -43,6 +43,7 @@ + 进阶使用 - [使用内存加速和SSD加速配置](samples/accelerate_data_by_mem_or_ssd.md) - [AlluxioRuntime分层存储配置](samples/tieredstore_config.md) + - [CacheRuntime分层存储配置](userguide/cache_runtime_tieredstore.md) - [Alluxio S3 高并发读调优](samples/alluxio_s3_high_concurrency.md) - [通过Webhook机制优化Pod调度](operation/pod_schedule_optimization.md) - [基于Runtime分层位置信息的应用Pod调度](operation/tiered_locality_schedule.md)