From 4c21e5b7c45b281b1feed29cbbad3bc0c0aab01c Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?S=C3=A9bastien=20Han?= Date: Tue, 20 Apr 2021 09:15:35 +0200 Subject: [PATCH] docs: reset the pending release notes MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Since the v1.6 release is completed, reset the pending release notes for v1.7. Signed-off-by: Sébastien Han --- PendingReleaseNotes.md | 24 +----------------------- 1 file changed, 1 insertion(+), 23 deletions(-) diff --git a/PendingReleaseNotes.md b/PendingReleaseNotes.md index c2de057cb..838dfa6fb 100644 --- a/PendingReleaseNotes.md +++ b/PendingReleaseNotes.md @@ -1,6 +1,6 @@ # Major Themes -v1.6... +v1.7... ## K8s Version Support @@ -9,31 +9,9 @@ v1.6... ## Breaking Changes ### Ceph -* Support for adding OSDs via Drive Groups was removed. Please refer to the - [Ceph upgrade guide](Documentation/ceph-upgrade.md#migrate-the-drive-group-spec) for migration - instructions. - See https://github.com/rook/rook/issues/7275 for more information. ## Features ### Core ### Ceph - -* Ceph Pacific support and along with it: - * Multiple Ceph Filesystems - * Networking dualstack -* CephClient CRD has been converted to use the controller-runtime library -* Extending the support of vault KMS configuration for Ceph RGW -* Enable disruption budgets (PDBs) by default for Mon, RGW, MDS, and OSD daemons -* Add CephFilesystemMirror CRD to deploy cephfs-mirror daemon -* Multiple Ceph mgr daemons are supported for stretch clusters and other clusters where HA of the mgr is more critical -* OSDs: - * as of Nautilus 14.2.14 and Octopus 15.2.9 if the OSD scenario is simple (one OSD per disk) we won't use LVM to prepare the disk anymore - * for Pacific (16.2.x), Rook is able to update multiple OSD Deployments at the same time to speed - up updates and upgrades for larger Ceph clusters -* Disable CSI GRPC metrics by default -* Update CephCSI to v3.3.0 -* Monitors failover can be [disabled](Documentation/ceph-mon-health.md#failing-over-a-monitor), this is useful for monitor going under planned maintenance where automatic failover is not desired -* RGW: Rook has started using deployment's replicatset functionality instead of deploying multiple deployments for each rgw -* Support running Volume Replication Controller on the RBD provisioner pod. Depends on Cephcsi v3.3.0