dm-operator/: support start a new dm cluster with dm-masters and dm-workers

cmd/controller-manager/main.go

@@ -29,6 +29,7 @@ import (
 	"github.com/pingcap/tidb-operator/pkg/controller/autoscaler"
 	"github.com/pingcap/tidb-operator/pkg/controller/backup"
 	"github.com/pingcap/tidb-operator/pkg/controller/backupschedule"
+	"github.com/pingcap/tidb-operator/pkg/controller/dmcluster"
 	"github.com/pingcap/tidb-operator/pkg/controller/periodicity"
 	"github.com/pingcap/tidb-operator/pkg/controller/restore"
 	"github.com/pingcap/tidb-operator/pkg/controller/tidbcluster"
@@ -184,6 +185,7 @@ func main() {
 		}
 
 		tcController := tidbcluster.NewController(kubeCli, cli, genericCli, informerFactory, kubeInformerFactory, autoFailover, pdFailoverPeriod, tikvFailoverPeriod, tidbFailoverPeriod, tiflashFailoverPeriod)
+		dcController := dmcluster.NewController(kubeCli, cli, genericCli, informerFactory, kubeInformerFactory, autoFailover)
 		backupController := backup.NewController(kubeCli, cli, informerFactory, kubeInformerFactory)
 		restoreController := restore.NewController(kubeCli, cli, informerFactory, kubeInformerFactory)
 		bsController := backupschedule.NewController(kubeCli, cli, informerFactory, kubeInformerFactory)
@@ -218,6 +220,7 @@ func main() {
 		}
 		klog.Infof("cache of informer factories sync successfully")
 
+		go wait.Forever(func() { dcController.Run(workers, ctx.Done()) }, waitDuration)
 		go wait.Forever(func() { backupController.Run(workers, ctx.Done()) }, waitDuration)
 		go wait.Forever(func() { restoreController.Run(workers, ctx.Done()) }, waitDuration)
 		go wait.Forever(func() { bsController.Run(workers, ctx.Done()) }, waitDuration)

docs/api-references/docs.md

@@ -682,6 +682,19 @@ Optional: Defaults to nil</p>
 </tr>
 <tr>
 <td>
+<code>hostNetwork</code></br>
+<em>
+bool
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>Whether Hostnetwork is enabled for DM cluster Pods
+Optional: Defaults to false</p>
+</td>
+</tr>
+<tr>
+<td>
 <code>affinity</code></br>
 <em>
 <a href="https://kubernetes.io/docs/reference/generated/kubernetes-api/v1.18/#affinity-v1-core">
@@ -696,6 +709,19 @@ Kubernetes core/v1.Affinity
 </tr>
 <tr>
 <td>
+<code>priorityClassName</code></br>
+<em>
+string
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>PriorityClassName of DM cluster Pods
+Optional: Defaults to omitted</p>
+</td>
+</tr>
+<tr>
+<td>
 <code>nodeSelector</code></br>
 <em>
 map[string]string
@@ -4119,8 +4145,8 @@ LeastRemainAvailableStoragePercent should between 5 and 90. If not set, the defa
 <td>
 <code>type</code></br>
 <em>
-<a href="#tidbclusterconditiontype">
-TidbClusterConditionType
+<a href="#dmclusterconditiontype">
+DMClusterConditionType
 </a>
 </em>
 </td>
@@ -4196,6 +4222,10 @@ string
 </table>
 <h3 id="dmclusterconditiontype">DMClusterConditionType</h3>
 <p>
+(<em>Appears on:</em>
+<a href="#dmclustercondition">DMClusterCondition</a>)
+</p>
+<p>
 <p>DMClusterConditionType represents a dm cluster condition value.</p>
 </p>
 <h3 id="dmclusterspec">DMClusterSpec</h3>
@@ -4362,6 +4392,19 @@ Optional: Defaults to nil</p>
 </tr>
 <tr>
 <td>
+<code>hostNetwork</code></br>
+<em>
+bool
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>Whether Hostnetwork is enabled for DM cluster Pods
+Optional: Defaults to false</p>
+</td>
+</tr>
+<tr>
+<td>
 <code>affinity</code></br>
 <em>
 <a href="https://kubernetes.io/docs/reference/generated/kubernetes-api/v1.18/#affinity-v1-core">
@@ -4376,6 +4419,19 @@ Kubernetes core/v1.Affinity
 </tr>
 <tr>
 <td>
+<code>priorityClassName</code></br>
+<em>
+string
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>PriorityClassName of DM cluster Pods
+Optional: Defaults to omitted</p>
+</td>
+</tr>
+<tr>
+<td>
 <code>nodeSelector</code></br>
 <em>
 map[string]string
@@ -6386,6 +6442,19 @@ Defaults to Kubernetes default storage class.</p>
 </tr>
 <tr>
 <td>
+<code>storageSize</code></br>
+<em>
+string
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>StorageSize is the request storage size for dm-master.
+Defaults to &ldquo;10Gi&rdquo;.</p>
+</td>
+</tr>
+<tr>
+<td>
 <code>dataSubDir</code></br>
 <em>
 string
@@ -17308,7 +17377,6 @@ string
 <h3 id="tidbclusterconditiontype">TidbClusterConditionType</h3>
 <p>
 (<em>Appears on:</em>
-<a href="#dmclustercondition">DMClusterCondition</a>, 
 <a href="#tidbclustercondition">TidbClusterCondition</a>)
 </p>
 <p>
@@ -18723,7 +18791,7 @@ DMSecurityConfig
 <tbody>
 <tr>
 <td>
-<code>podName</code></br>
+<code>name</code></br>
 <em>
 string
 </em>
@@ -18733,7 +18801,7 @@ string
 </tr>
 <tr>
 <td>
-<code>id</code></br>
+<code>addr</code></br>
 <em>
 string
 </em>
@@ -18743,7 +18811,7 @@ string
 </tr>
 <tr>
 <td>
-<code>state</code></br>
+<code>stage</code></br>
 <em>
 string
 </em>
@@ -18837,6 +18905,19 @@ string
 </tr>
 <tr>
 <td>
+<code>maxFailoverCount</code></br>
+<em>
+int32
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>MaxFailoverCount limit the max replicas could be added in failover, 0 means no failover.
+Optional: Defaults to 3</p>
+</td>
+</tr>
+<tr>
+<td>
 <code>storageClassName</code></br>
 <em>
 string
@@ -18850,6 +18931,19 @@ Defaults to Kubernetes default storage class.</p>
 </tr>
 <tr>
 <td>
+<code>storageSize</code></br>
+<em>
+string
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>StorageSize is the request storage size for dm-worker.
+Defaults to &ldquo;10Gi&rdquo;.</p>
+</td>
+</tr>
+<tr>
+<td>
 <code>dataSubDir</code></br>
 <em>
 string
@@ -18935,7 +19029,7 @@ Kubernetes apps/v1.StatefulSetStatus
 </tr>
 <tr>
 <td>
-<code>workers</code></br>
+<code>members</code></br>
 <em>
 <a href="#workermember">
 map[string]github.com/pingcap/tidb-operator/pkg/apis/pingcap/v1alpha1.WorkerMember
@@ -18945,6 +19039,16 @@ map[string]github.com/pingcap/tidb-operator/pkg/apis/pingcap/v1alpha1.WorkerMemb
 <td>
 </td>
 </tr>
+<tr>
+<td>
+<code>image</code></br>
+<em>
+string
+</em>
+</td>
+<td>
+</td>
+</tr>
 </tbody>
 </table>
 <hr/>

examples/dm/README.md

@@ -0,0 +1,65 @@
+# A Basic DM cluster
+
+> **Note:**
+>
+> This setup is for test or demo purpose only and **IS NOT** applicable for critical environment. Refer to the [Documents](https://pingcap.com/docs/stable/tidb-in-kubernetes/deploy/prerequisites/) for production setup.
+
+The following steps will create a DM cluster.
+
+**Prerequisites**: 
+- Has TiDB operator `v1.2.0` or higher version installed. [Doc](https://pingcap.com/docs/stable/tidb-in-kubernetes/deploy/tidb-operator/)
+- Has default `StorageClass` configured, and there are enough PVs (by default, 2 PVs are required) of that storageClass:
+
+  This could by verified by the following command:
+
+  ```bash
+  > kubectl get storageclass
+  ```
+
+  The output is similar to this:
+
+  ```bash
+  NAME                 PROVISIONER               AGE
+  standard (default)   kubernetes.io/gce-pd      1d
+  gold                 kubernetes.io/gce-pd      1d
+  ```
+
+  Alternatively, you could specify the storageClass explicitly by modifying `dm-cluster.yaml`.
+
+## Install
+
+The following commands is assumed to be executed in this directory.
+
+Install the cluster:
+
+```bash
+> kubectl -n <namespace> apply -f ./
+```
+
+Wait for cluster Pods ready:
+
+```bash
+watch kubectl -n <namespace> get pod
+```
+
+## Explore
+
+Explore the DM master interface:
+
+```bash
+> kubectl -n <namespace> port-forward svc/basic-dm-master 8261:8261
+> dmctl --master-addr 127.0.0.1:8261 list-member
+```
+
+## Destroy
+
+```bash
+> kubectl -n <namespace> delete -f ./
+```
+
+The PVCs used by DM cluster will not be deleted in the above process, therefore, the PVs will be not be released neither. You can delete PVCs and release the PVs by the following command:
+
+```bash
+> kubectl -n <namespace> delete pvc -l app.kubernetes.io/instance=basic,app.kubernetes.io/managed-by=tidb-operator
+```
+

examples/dm/dm-cluster.yaml

@@ -0,0 +1,26 @@
+apiVersion: pingcap.com/v1alpha1
+kind: DMCluster
+metadata:
+  name: basic
+  namespace: demo
+spec:
+  version: v2.0.0-beta.2
+  pvReclaimPolicy: Delete
+  discovery:
+    address: "http://basic-discovery.demo:10261"
+  master:
+    baseImage: pingcap/dm
+    replicas: 1
+    # if storageClassName is not set, the default Storage Class of the Kubernetes cluster will be used
+    # storageClassName: local-storage
+    storageSize: "1Gi"
+    requests: {}
+    config: {}
+  worker:
+    baseImage: pingcap/dm
+    replicas: 1
+    # if storageClassName is not set, the default Storage Class of the Kubernetes cluster will be used
+    # storageClassName: local-storage
+    storageSize: "1Gi"
+    requests: {}
+    config: {}

go.sum

@@ -408,6 +408,7 @@ github.com/golangplus/fmt v0.0.0-20150411045040-2a5d6d7d2995/go.mod h1:lJgMEyOkY
 github.com/golangplus/testing v0.0.0-20180327235837-af21d9c3145e/go.mod h1:0AA//k/eakGydO4jKRoRL2j92ZKSzTgj9tclaCrvXHk=
 github.com/google/btree v0.0.0-20180813153112-4030bb1f1f0c h1:964Od4U6p2jUkFxvCydnIczKteheJEzHRToSGK3Bnlw=
 github.com/google/btree v0.0.0-20180813153112-4030bb1f1f0c/go.mod h1:lNA+9X1NB3Zf8V7Ke586lFgjr2dZNuvo3lPJSGZ5JPQ=
+github.com/google/cadvisor v0.34.0 h1:No7G6U/TasplR9uNqyc5Jj0Bet5VSYsK5xLygOf4pUw=
 github.com/google/cadvisor v0.34.0/go.mod h1:1nql6U13uTHaLYB8rLS5x9IJc2qT6Xd/Tr1sTX6NE48=
 github.com/google/certificate-transparency-go v1.0.21/go.mod h1:QeJfpSbVSfYc7RgB3gJFj9cbuQMMchQxrWXz8Ruopmg=
 github.com/google/go-cmp v0.2.0/go.mod h1:oXzfMopK8JAjlY9xF4vHSVASa0yLyX7SntLO5aqRK0M=

hack/update-crd-groups.sh

@@ -33,6 +33,7 @@ export GO111MODULE=on
 go install github.com/pingcap/tidb-operator/cmd/to-crdgen
 
 to-crdgen generate tidbcluster > $crd_target
+to-crdgen generate dmcluster >> $crd_target
 to-crdgen generate backup >> $crd_target
 to-crdgen generate restore >> $crd_target
 to-crdgen generate backupschedule >> $crd_target