diff --git a/Documentation/ceph-filesystem-crd.md b/Documentation/ceph-filesystem-crd.md index f0a6a6d60ca3..de34aff5f457 100644 --- a/Documentation/ceph-filesystem-crd.md +++ b/Documentation/ceph-filesystem-crd.md @@ -86,6 +86,8 @@ spec: replicated: size: 3 dataPools: + - replicated: + size: 3 - erasureCoded: dataChunks: 2 codingChunks: 1 @@ -94,6 +96,8 @@ spec: activeStandby: true ``` +**IMPORTANT**: For erasure coded pools, we have to create a replicated pool as the default data pool and an erasure-coded pool as a secondary pool. + (These definitions can also be found in the [`filesystem-ec.yaml`](https://github.com/rook/rook/blob/{{ branchName }}/deploy/examples/filesystem-ec.yaml) file. Also see an example in the [`storageclass-ec.yaml`](https://github.com/rook/rook/blob/{{ branchName }}/deploy/examples/csi/cephfs/storageclass-ec.yaml) for how to configure the volume.) diff --git a/deploy/examples/csi/cephfs/storageclass-ec.yaml b/deploy/examples/csi/cephfs/storageclass-ec.yaml index 6c792812921a..d3975cec844f 100644 --- a/deploy/examples/csi/cephfs/storageclass-ec.yaml +++ b/deploy/examples/csi/cephfs/storageclass-ec.yaml @@ -14,7 +14,10 @@ parameters: # Ceph pool into which the volume shall be created # Required for provisionVolume: "true" - pool: myfs-ec-data0 + + # For erasure coded pools, we have to create a replicated pool as the default data pool and an erasure-coded + # pool as a secondary pool. + pool: myfs-ec-data1 # The secrets contain Ceph admin credentials. These are generated automatically by the operator # in the same namespace as the cluster. diff --git a/deploy/examples/filesystem-ec.yaml b/deploy/examples/filesystem-ec.yaml index 205adfd13a25..fcf15dcc17f1 100644 --- a/deploy/examples/filesystem-ec.yaml +++ b/deploy/examples/filesystem-ec.yaml @@ -17,6 +17,10 @@ spec: size: 3 # The list of data pool specs dataPools: + # For erasure coded pools, we have to create a replicated pool as the default + # data pool and an erasure-coded pool as a secondary pool. + - replicated: + size: 3 # You need at least three `bluestore` OSDs on different nodes for this config to work - erasureCoded: dataChunks: 2