k3s 集群部署 rook ceph
在此之前这篇博客 介绍了如何在 kubernetes 集群上部署 rook ceph 集群，该博客适用的 kubernetes 集 群版本为 v1.15.9，在新版本的 Cloudpods 中集群采用 k3s 部署，并且 k3s 的版本为 v1.28.5，此前博客中的文档已不适用，本文介绍如何在 cloudpods k3s 集群中部署 rook ceph。
部署准备
在 cloupods k3s 集群中部署 rook ceph v1.17.9 版本， 首先准备好 k3s 搭建的 cloudpods 集群，并且集群内都有可用的裸盘给 Ceph 使用, eg：
$ kubectl get nodes
NAME STATUS ROLES AGE VERSION
wyq-test-cluster-2 Ready <none> 39h v1.28.5+k3s1
wyq-test-cluster Ready control-plane,master 40h v1.28.5+k3s1
wyq-test-cluster-3 Ready <none> 39h v1.28.5+k3s1
$ lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS
sda 8:0 0 100G 0 disk
├─sda1 8:1 0 1G 0 part /boot
└─sda2 8:2 0 99G 0 part /
sdb 8:16 0 300G 0 disk
然后给对应的节点打上 role=storage-node 的标签：
# 打标签
$ kubectl label node wyq-test-cluster role=storage-node
$ kubectl label node wyq-test-cluster-2 role=storage-node
$ kubectl label node wyq-test-cluster-3 role=storage-node
# 查看标签对应的节点
$ kubectl get nodes -L role
NAME STATUS ROLES AGE VERSION ROLE
wyq-test-cluster-3 Ready <none> 39h v1.28.5+k3s1 storage-node
wyq-test-cluster-2 Ready <none> 39h v1.28.5+k3s1 storage-node
wyq-test-cluster Ready control-plane,master 40h v1.28.5+k3s1 storage-node
部署 Rook 前准备
下载 Rook 相关代码：
# clone rook 源码
$ git clone --single-branch --branch v1.17.9 https://github.com/rook/rook.git
$ cd rook/deploy/examples
部署前替换镜像，由于rook ceph 中默认的镜像可能国内无法拉取，所以同步了必要的镜像到 cloudpods aliyun registry, 需要的可以按需替换，对应关系如下：
docker.io/rook/ceph:v1.17.9
gcr.io/k8s-staging-sig-storage/objectstorage-sidecar:v20240513-v0.1.0-35-gefb3255
quay.io/ceph/ceph:v19.2.3
quay.io/ceph/cosi:v0.1.2
quay.io/cephcsi/cephcsi:v3.14.2
quay.io/csiaddons/k8s-sidecar:v0.12.0
registry.k8s.io/sig-storage/csi-attacher:v4.8.1
registry.k8s.io/sig-storage/csi-node-driver-registrar:v2.13.0
registry.k8s.io/sig-storage/csi-provisioner:v5.2.0
registry.k8s.io/sig-storage/csi-resizer:v1.13.2
registry.k8s.io/sig-storage/csi-snapshotter:v8.2.1
--------------------------
registry.cn-beijing.aliyuncs.com/yunionio/rook:ceph-v1.17.9
registry.cn-beijing.aliyuncs.com/yunionio/rook:objectstorage-sidecar-v20240513-v0.1.0-35-gefb3255
registry.cn-beijing.aliyuncs.com/yunionio/ceph:v19.2.3
registry.cn-beijing.aliyuncs.com/yunionio/rook:cosi-v0.1.2
registry.cn-beijing.aliyuncs.com/yunionio/rook:cephcsi-v3.14.2
registry.cn-beijing.aliyuncs.com/yunionio/rook:k8s-sidecar-v0.12.0
registry.cn-beijing.aliyuncs.com/yunionio/rook:csi-attacher-v4.8.1
registry.cn-beijing.aliyuncs.com/yunionio/rook:csi-node-driver-registrar-v2.13.0
registry.cn-beijing.aliyuncs.com/yunionio/rook:csi-provisioner-v5.2.0
registry.cn-beijing.aliyuncs.com/yunionio/rook:csi-resizer-v1.13.2
registry.cn-beijing.aliyuncs.com/yunionio/rook:csi-snapshotter-v8.2.1
需要替换镜像的文件：
$ git diff operator.yaml
diff --git a/deploy/examples/operator.yaml b/deploy/examples/operator.yaml
index 99a220742..031545824 100644
--- a/deploy/examples/operator.yaml
+++ b/deploy/examples/operator.yaml
@@ -119,12 +119,12 @@ data:
# The default version of CSI supported by Rook will be started. To change the version
# of the CSI driver to something other than what is officially supported, change
# these images to the desired release of the CSI driver.
- # ROOK_CSI_CEPH_IMAGE: "quay.io/cephcsi/cephcsi:v3.14.2"
- # ROOK_CSI_REGISTRAR_IMAGE: "registry.k8s.io/sig-storage/csi-node-driver-registrar:v2.13.0"
- # ROOK_CSI_RESIZER_IMAGE: "registry.k8s.io/sig-storage/csi-resizer:v1.13.2"
- # ROOK_CSI_PROVISIONER_IMAGE: "registry.k8s.io/sig-storage/csi-provisioner:v5.2.0"
- # ROOK_CSI_SNAPSHOTTER_IMAGE: "registry.k8s.io/sig-storage/csi-snapshotter:v8.2.1"
- # ROOK_CSI_ATTACHER_IMAGE: "registry.k8s.io/sig-storage/csi-attacher:v4.8.1"
+ ROOK_CSI_CEPH_IMAGE: "registry.cn-beijing.aliyuncs.com/yunionio/rook:cephcsi-v3.14.2"
+ ROOK_CSI_REGISTRAR_IMAGE: "registry.cn-beijing.aliyuncs.com/yunionio/rook:csi-node-driver-registrar-v2.13.0"
+ ROOK_CSI_RESIZER_IMAGE: "registry.cn-beijing.aliyuncs.com/yunionio/rook:csi-resizer-v1.13.2"
+ ROOK_CSI_PROVISIONER_IMAGE: "registry.cn-beijing.aliyuncs.com/yunionio/rook:csi-provisioner-v5.2.0"
+ ROOK_CSI_SNAPSHOTTER_IMAGE: "registry.cn-beijing.aliyuncs.com/yunionio/rook:csi-snapshotter-v8.2.1"
+ ROOK_CSI_ATTACHER_IMAGE: "registry.cn-beijing.aliyuncs.com/yunionio/rook:csi-attacher-v4.8.1"
# To indicate the image pull policy to be applied to all the containers in the csi driver pods.
# ROOK_CSI_IMAGE_PULL_POLICY: "IfNotPresent"
@@ -511,7 +511,7 @@ data:
CSI_ENABLE_CSIADDONS: "false"
# Enable watch for faster recovery from rbd rwo node loss
ROOK_WATCH_FOR_NODE_FAILURE: "true"
- # ROOK_CSIADDONS_IMAGE: "quay.io/csiaddons/k8s-sidecar:v0.12.0"
+ ROOK_CSIADDONS_IMAGE: "registry.cn-beijing.aliyuncs.com/yunionio/rook:k8s-sidecar-v0.12.0"
# The CSI GRPC timeout value (in seconds). It should be >= 120. If this variable is not set or is an invalid value, it's default to 150.
CSI_GRPC_TIMEOUT_SECONDS: "150"
@@ -617,7 +617,7 @@ spec:
serviceAccountName: rook-ceph-system
containers:
- name: rook-ceph-operator
- image: docker.io/rook/ceph:v1.17.9
+ image: registry.cn-beijing.aliyuncs.com/yunionio/rook:ceph-v1.17.9
args: ["ceph", "operator"]
securityContext:
runAsNonRoot: true
开始部署 rook ceph:
首先根据自己的环境，修改 rook 提供的 cluster.yaml 里面的内容：
下面是修改后的 cluster.yaml 的 diff 内容：
diff --git a/deploy/examples/cluster.yaml b/deploy/examples/cluster.yaml
index f9820f384..d48a966a1 100644
--- a/deploy/examples/cluster.yaml
+++ b/deploy/examples/cluster.yaml
@@ -21,7 +21,7 @@ spec:
# versions running within the cluster. See tags available at https://hub.docker.com/r/ceph/ceph/tags/.
# If you want to be more precise, you can always use a timestamp tag such as quay.io/ceph/ceph:v19.2.3-20250717
# This tag might not contain a new Ceph version, just security fixes from the underlying operating system, which will reduce vulnerabilities
- image: quay.io/ceph/ceph:v19.2.3
+ image: registry.cn-beijing.aliyuncs.com/yunionio/ceph:v19.2.3
# Whether to allow unsupported versions of Ceph. Currently Reef and Squid are supported.
# Future versions such as Tentacle (v20) would require this to be set to `true`.
# Do not set to true in production.
@@ -113,7 +113,7 @@ spec:
# Requires a kernel that supports msgr v2 (kernel 5.11 or CentOS 8.4 or newer).
requireMsgr2: false
# enable host networking
- #provider: host
+ provider: host
# enable the Multus network provider
#provider: multus
#selectors:
@@ -175,22 +175,22 @@ spec:
# To control where various services will be scheduled by kubernetes, use the placement configuration sections below.
# The example under 'all' would have all services scheduled on kubernetes nodes labeled with 'role=storage-node' and
# tolerate taints with a key of 'storage-node'.
- # placement:
- # all:
- # nodeAffinity:
- # requiredDuringSchedulingIgnoredDuringExecution:
- # nodeSelectorTerms:
- # - matchExpressions:
- # - key: role
- # operator: In
- # values:
- # - storage-node
- # podAffinity:
- # podAntiAffinity:
- # topologySpreadConstraints:
- # tolerations:
- # - key: storage-node
- # operator: Exists
+ placement:
+ all:
+ nodeAffinity:
+ requiredDuringSchedulingIgnoredDuringExecution:
+ nodeSelectorTerms:
+ - matchExpressions:
+ - key: role
+ operator: In
+ values:
+ - storage-node
+ podAffinity:
+ podAntiAffinity:
+ topologySpreadConstraints:
+ tolerations:
+ - key: storage-node
+ operator: Exists
# The above placement information can also be specified for mon, osd, and mgr components
# mon:
# Monitor deployments may contain an anti-affinity rule for avoiding monitor
@@ -262,8 +262,8 @@ spec:
mgr: system-cluster-critical
#crashcollector: rook-ceph-crashcollector-priority-class
storage: # cluster level storage configuration and selection
- useAllNodes: true
- useAllDevices: true
+ useAllNodes: false
+ useAllDevices: false
#deviceFilter:
config:
# crushRoot: "custom-root" # specify a non-default root label for the CRUSH map
@@ -287,6 +287,16 @@ spec:
# config: # configuration can be specified at the node level which overrides the cluster level config
# - name: "172.17.4.301"
# deviceFilter: "^sd."
+ nodes:
+ - name: "wyq-test-cluster"
+ devices:
+ - name: "sdb"
+ - name: "wyq-test-cluster2"
+ devices:
+ - name: "sdb"
+ - name: "wyq-test-cluster3"
+ devices:
+ - name: "sdb"
# Whether to always schedule OSD pods on nodes declared explicitly in the "nodes" section, even if they are
# temporarily not schedulable. If set to true, consider adding placement tolerations for unschedulable nodes.
scheduleAlways: false
diff --git a/deploy/examples/toolbox.yaml b/deploy/examples/toolbox.yaml
index 5559e064a..812716e96 100644
--- a/deploy/examples/toolbox.yaml
+++ b/deploy/examples/toolbox.yaml
@@ -19,7 +19,7 @@ spec:
serviceAccountName: rook-ceph-default
containers:
- name: rook-ceph-tools
- image: quay.io/ceph/ceph:v19
+ image: registry.cn-beijing.aliyuncs.com/yunionio/ceph:v19.2.3
command:
- /bin/bash
- -c
开始创建 rook ceph 集群：
kubectl create -f crds.yaml -f common.yaml -f operator.yaml
kubectl create -f cluster.yaml
# 查看 rook-ceph namespace 里面的 pod 健康状况，等待 pod running
$ kubectl -n rook-ceph get pods
NAME READY STATUS RESTARTS AGE
rook-ceph-operator-99bcd958c-29t7f 1/1 Running 0 12h
csi-rbdplugin-tnnw4 3/3 Running 0 12h
csi-cephfsplugin-rnprg 3/3 Running 0 12h
csi-rbdplugin-g57qr 3/3 Running 0 12h
csi-cephfsplugin-7ft2q 3/3 Running 0 12h
csi-cephfsplugin-l84w5 3/3 Running 0 12h
csi-rbdplugin-r5722 3/3 Running 0 12h
rook-ceph-mon-a-68644647b6-9qdk8 2/2 Running 0 12h
rook-ceph-mon-b-b5fd4569d-fhr9j 2/2 Running 0 12h
rook-ceph-mon-c-7fc4667b4c-pvr9r 2/2 Running 0 12h
csi-cephfsplugin-provisioner-64d55dff4f-86qk9 6/6 Running 0 12h
csi-rbdplugin-provisioner-d5d7f5d5-qfxm5 6/6 Running 0 12h
csi-rbdplugin-provisioner-d5d7f5d5-frw88 6/6 Running 0 12h
csi-cephfsplugin-provisioner-64d55dff4f-5wk4x 6/6 Running 0 12h
rook-ceph-crashcollector-wyq-test-cluster-795564c75b-vqv5f 1/1 Running 0 12h
rook-ceph-exporter-wyq-test-cluster-7d955fff7d-8tqj8 1/1 Running 0 12h
rook-ceph-mgr-a-6cc8d446b7-j2jbx 3/3 Running 0 12h
rook-ceph-mgr-b-f7897c956-98w7q 3/3 Running 0 12h
rook-ceph-tools-56fbc74755-flwz5 1/1 Running 0 12h
rook-ceph-osd-prepare-wyq-test-cluster-6x2s4 0/1 Completed 0 12h
rook-ceph-osd-prepare-wyq-test-cluster-2-gxzlj 0/1 Completed 0 12h
rook-ceph-osd-prepare-wyq-test-cluster-3-c8mqr 0/1 Completed 0 12h
rook-ceph-osd-0-557d7678b-xfjnl 2/2 Running 0 12h
rook-ceph-crashcollector-wyq-test-cluster-3-5594748df6-4bksb 1/1 Running 0 12h
rook-ceph-crashcollector-wyq-test-cluster-2-5cc578c887-7tns8 1/1 Running 0 12h
rook-ceph-exporter-wyq-test-cluster-2-6cc44967cc-jh9bf 1/1 Running 0 12h
rook-ceph-exporter-wyq-test-cluster-3-57d757879d-lfmvd 1/1 Running 0 12h
rook-ceph-osd-2-755d54f95f-z42mw 2/2 Running 0 12h
rook-ceph-osd-1-6897fb8bfd-nhlw4 2/2 Running 0 12h
kubectl -n rook-ceph get cephcluster
NAME DATADIRHOSTPATH MONCOUNT AGE PHASE MESSAGE HEALTH EXTERNAL FSID
rook-ceph /var/lib/rook 3 13h Ready Cluster created successfully HEALTH_OK 0fa8e642-7e3c-4096-9d24-4781199632c3
ceph 集群部署完后，我们需要部署 toolbox.yaml pod 获取集群连接信息：
$ kubectl apply -f toolbox.yaml
deployment.apps/rook-ceph-tools created
$ kubectl -n rook-ceph get pods | grep tools
rook-ceph-tools-666bf4b894-knfvk 1/1 Running 0 2m9s
# 进入 toolbox pod
$ kubectl -n rook-ceph exec -it $(kubectl -n rook-ceph get pod -l "app=rook-ceph-tools" -o jsonpath='{.items[0].metadata.name}') -- bash
# 查看ceph.conf 和 keyring
$ cat /etc/ceph/ceph.conf
[global]
mon_host = 10.168.26.185:6789,10.168.26.184:6789,10.168.26.187:6789
[client.admin]
keyring = /etc/ceph/keyring
$ cat /etc/ceph/keyring
[client.admin]
key = AQCwoNdp1+ziORAAwRmOqnJg8ZAdUInn1MRyRA==
至此 rook ceph 集群以创建完毕，后续添加到 cloudpods 集群可以继续查看此博客 。
支持的 ceph 版本
rook ceph v1.17 版本支持的 ceph 版本为：
- Squid (v19.2.0+)
- Reef (v18.2.0+)
如要部署不同的 ceph 版本需要替换 cluster.yaml 中的 ceph 镜像版本。