diff --git a/ROADMAP.md b/ROADMAP.md index 66e6bc2aa..18cbe40ce 100644 --- a/ROADMAP.md +++ b/ROADMAP.md @@ -1,31 +1,41 @@ # Roadmap -This document defines a high level roadmap for Rook development. The dates below are subject to change but should give a general idea of what we are planning. We use the [milestone](https://github.com/rook/rook/milestones) feature in Github so look there for the most up-to-date and issue plan. - -## Rook 0.7 - -- Ceph Block Storage (Cluster and Pool CRDs) declared beta -- Mon reliability (restarts, failing over too fast, ip changes, etc.) -- Durability of state (local storage support, config is regenerat-able) -- Run everywhere (rook-agent uses ceph-fuse, nbd-rbd / tcmu runner) -- Adding / removing nodes and disk drives (lifecycle issues, failures, etc.) -- Run with Least Privileged and possibly without privileged containers -- Shutdown / restart issues -- Improved data placement and pool configuration (CRUSH maps) -- Placement group balancer support (ceph-mgr module) -- Use new Prometheus ceph-mgr module -- Performance Testing for Block storage -- Make API and CLI Optional (plan to remove) -- Dynamic Volume Provisioning for CephFS +This document defines a high level roadmap for Rook development. +The dates below are subject to change but should give a general idea of what we are planning. +We use the [milestone](https://github.com/rook/rook/milestones) feature in Github so look there for the most up-to-date and issue plan. ## Rook 0.8 -- Operator runs HA -- Ceph Object storage declared Beta -- Logging (levels, granularity, etc.) -- Support ingress controller for RGW (SSL termination) -- Support for multi-region replication for RGW -- Object storage User CRD -- Automated Upgrade -- CRD validation, cleanup, progress, status (requires engagement with API sig) -- Volume Snapshotting (consider aligning with SIG-storage) +* Multiple storage backends + * Full design + * Refactor code base and repositories to enable + * Consider support for Minio, potentially early support for other backends time permitting +* Switch CRDs to API Aggregation +* Run on arbitrary PVs +* Remove Rook API and CLI +* Migrate CI and release pipelines to a solution hosted by the CNCF +* Run with Least Privileged and possibly without privileged containers +* Shutdown / restart issues +* Support Kubernetes 1.7+ only +* Ceph features and improvements + * Disk management (1 OSD per pod, adding/removing disks) + * Placement group balancer support (ceph-mgr module) + * Mon reliability (restarts, failing over too fast, ip changes, etc.) + * Mimic support + +## Rook 0.9 + +* Automated upgrades +* Durability of state (local storage support, config can be regenerated) +* Custom resource validation, progress, status +* Design for Volume Snapshotting and policies (consider aligning with SIG-storage) +* Ceph features and improvements + * Improved data placement and pool configuration (CRUSH maps) + * Dynamic Volume Provisioning for CephFS + * ceph-volume provisioning for OSDs + * Ceph CSI plug-in that runs everywhere (rook-agent uses ceph-fuse, nbd-rbd / tcmu runner) +* Cluster and Pool types are declared Beta + +## Rook 1.0 + +* Declare numerous types as v1 \ No newline at end of file