管理 ETCD
- 先决条件
- 安装
- 备份
- 恢复

管理 ETCD

GreptimeDB 集群默认需要一个 etcd 集群用于元数据存储。让我们使用 Bitnami 的 etcd Helm chart 安装一个 etcd 集群。

先决条件

Kubernetes >= v1.23
kubectl >= v1.18.0
Helm >= v3.0.0

安装

1. helm upgrade --install etcd \
2. oci://registry-1.docker.io/bitnamicharts/etcd \
3. --version 10.2.12 \
4. --set replicaCount=3 \
5. --set auth.rbac.create=false \
6. --set auth.rbac.token.enabled=false \
7. --create-namespace \
8. -n etcd-cluster

备注

中国大陆用户如有网络访问问题，可直接使用阿里云 OCI 镜像仓库的方式安装 etcd 集群：

1. helm install etcd \
2. oci://greptime-registry.cn-hangzhou.cr.aliyuncs.com/charts/etcd \
3. --set image.registry=greptime-registry.cn-hangzhou.cr.aliyuncs.com \
4. --set image.tag=3.5.12 \
5. --set replicaCount=3 \
6. --set auth.rbac.create=false \
7. --set auth.rbac.token.enabled=false \
8. --create-namespace \
9. -n etcd-cluster

等待 etcd 集群运行：

1. kubectl get po -n etcd-cluster

Expected Output

1. NAME     READY   STATUS    RESTARTS   AGE
2. etcd-0   1/1     Running   0          64s
3. etcd-1   1/1     Running   0          65s
4. etcd-2   1/1     Running   0          72s

etcd initialClusterState 参数指定启动 etcd 节点时 etcd 集群的初始状态。它对于确定节点如何加入集群非常重要。该参数可以采用以下两个值:

new: 表示 etcd 集群是新的。所有节点将作为新集群的一部分启动，并且不会使用任何先前的状态。
existing: 表示该节点将加入一个已经存在的 etcd 集群，这种情况下必须确保 initialCluster 参数配置了当前集群所有节点的信息。

etcd 集群运行起来后，我们需要设置 initialClusterState 参数为 existing：

1. helm upgrade --install etcd \
2. oci://registry-1.docker.io/bitnamicharts/etcd \
3. --version 10.2.12 \
4. --set initialClusterState="existing" \
5. --set removeMemberOnContainerTermination=false \
6. --set replicaCount=3 \
7. --set auth.rbac.create=false \
8. --set auth.rbac.token.enabled=false \
9. --create-namespace \
10. -n etcd-cluster

等待 etcd 集群运行完毕，使用以下命令检查 etcd 集群的健康状态：

1. kubectl -n etcd-cluster \
2. exec etcd-0 -- etcdctl \
3. --endpoints etcd-0.etcd-headless.etcd-cluster:2379,etcd-1.etcd-headless.etcd-cluster:2379,etcd-2.etcd-headless.etcd-cluster:2379 \
4. endpoint status -w table

Expected Output

1. +----------------------------------------+------------------+---------+---------+-----------+------------+-----------+------------+--------------------+--------+
2. |                ENDPOINT                |        ID        | VERSION | DB SIZE | IS LEADER | IS LEARNER | RAFT TERM | RAFT INDEX | RAFT APPLIED INDEX | ERRORS |
3. +----------------------------------------+------------------+---------+---------+-----------+------------+-----------+------------+--------------------+--------+
4. | etcd-0.etcd-headless.etcd-cluster:2379 | 680910587385ae31 |  3.5.15 |   20 kB |     false |      false |         4 |      73991 |              73991 |        |
5. | etcd-1.etcd-headless.etcd-cluster:2379 | d6980d56f5e3d817 |  3.5.15 |   20 kB |     false |      false |         4 |      73991 |              73991 |        |
6. | etcd-2.etcd-headless.etcd-cluster:2379 | 12664fc67659db0a |  3.5.15 |   20 kB |      true |      false |         4 |      73991 |              73991 |        |
7. +----------------------------------------+------------------+---------+---------+-----------+------------+-----------+------------+--------------------+--------+

备份

在 bitnami etcd chart 中，使用共享存储卷 Network File System (NFS) 存储 etcd 备份数据。通过 Kubernetes 中的 CronJob 进行 etcd 快照备份，并挂载 NFS PersistentVolumeClaim (PVC)，可以将快照传输到 NFS 中。

添加以下配置，并将其命名为 etcd-backup.yaml 文件，注意需要将 existingClaim 修改为你的 NFS PVC 名称：

1. replicaCount: 3

3. auth:
4. rbac:
5. create: false
6. token:
7. enabled: false

9. initialClusterState: "existing"
10. removeMemberOnContainerTermination: false

12. disasterRecovery:
13. enabled: true
14. cronjob:
15. schedule: "*/30 * * * *"
16. historyLimit: 2
17. snapshotHistoryLimit: 2
18. pvc:
19. existingClaim: "${YOUR_NFS_PVC_NAME_HERE}"

重新部署 etcd 集群：

1. helm upgrade --install etcd \
2. oci://registry-1.docker.io/bitnamicharts/etcd \
3. --version 10.2.12 \
4. --create-namespace \
5. -n etcd-cluster --values etcd-backup.yaml

你可以看到 etcd 备份计划任务：

1. kubectl get cronjob -n etcd-cluster

Expected Output

1. NAME               SCHEDULE      TIMEZONE   SUSPEND   ACTIVE   LAST SCHEDULE   AGE
2. etcd-snapshotter   */30 * * * *   <none>    False     0        <none>          36s

1. kubectl get pod -n etcd-cluster

Expected Output

1. NAME                              READY   STATUS      RESTARTS   AGE
2. etcd-0                            1/1     Running     0          35m
3. etcd-1                            1/1     Running     0          36m
4. etcd-2                            0/1     Running     0          6m28s
5. etcd-snapshotter-28936038-tsck8   0/1     Completed   0          4m49s

1. kubectl logs etcd-snapshotter-28936038-tsck8 -n etcd-cluster

Expected Output

1. etcd-0.etcd-headless.etcd-cluster.svc.cluster.local:2379 is healthy: successfully committed proposal: took = 2.698457ms
2. etcd 11:18:07.47 INFO  ==> Snapshotting the keyspace
3. {"level":"info","ts":"2025-01-06T11:18:07.579095Z","caller":"snapshot/v3_snapshot.go:65","msg":"created temporary db file","path":"/snapshots/db-2025-01-06_11-18.part"}
4. {"level":"info","ts":"2025-01-06T11:18:07.580335Z","logger":"client","caller":"[email protected]/maintenance.go:212","msg":"opened snapshot stream; downloading"}
5. {"level":"info","ts":"2025-01-06T11:18:07.580359Z","caller":"snapshot/v3_snapshot.go:73","msg":"fetching snapshot","endpoint":"etcd-0.etcd-headless.etcd-cluster.svc.cluster.local:2379"}
6. {"level":"info","ts":"2025-01-06T11:18:07.582124Z","logger":"client","caller":"[email protected]/maintenance.go:220","msg":"completed snapshot read; closing"}
7. {"level":"info","ts":"2025-01-06T11:18:07.582688Z","caller":"snapshot/v3_snapshot.go:88","msg":"fetched snapshot","endpoint":"etcd-0.etcd-headless.etcd-cluster.svc.cluster.local:2379","size":"20 kB","took":"now"}
8. {"level":"info","ts":"2025-01-06T11:18:07.583008Z","caller":"snapshot/v3_snapshot.go:97","msg":"saved","path":"/snapshots/db-2025-01-06_11-18"}
9. Snapshot saved at /snapshots/db-2025-01-06_11-18

接下来，可以在 NFS 服务器中看到 etcd 备份快照：

1. ls ${NFS_SERVER_DIRECTORY}

Expected Output

1. db-2025-01-06_11-18  db-2025-01-06_11-20  db-2025-01-06_11-22

恢复

当您遇到 etcd 数据丢失或损坏（例如，存储在 etcd 中的关键信息被意外删除，或者发生无法恢复的灾难性集群故障）时，您需要执行 etcd 恢复。此外，恢复 etcd 还可用于开发和测试目的。

恢复前需要停止向 etcd 集群写入数据（停止 GreptimeDB Metasrv 对 etcd 的写入），并创建最新的快照文件用于恢复。

添加以下配置文件，命名为 etcd-restore.yaml。注意，existingClaim 是你的 NFS PVC 的名字，snapshotFilename 为 etcd 快照文件名：

1. replicaCount: 3

3. auth:
4. rbac:
5. create: false
6. token:
7. enabled: false

9. startFromSnapshot:
10. enabled: true
11. existingClaim: "${YOUR_NFS_PVC_NAME_HERE}"
12. snapshotFilename: "${YOUR_ETCD_SNAPSHOT_FILE_NAME}"

部署 etcd 恢复集群：

1. helm upgrade --install etcd-recover \
2. oci://registry-1.docker.io/bitnamicharts/etcd \
3. --version 10.2.12 \
4. --create-namespace \
5. -n etcd-cluster --values etcd-restore.yaml

等待 etcd 恢复集群运行后，重新部署 etcd 恢复集群：

1. helm upgrade --install etcd-recover \
2. oci://registry-1.docker.io/bitnamicharts/etcd \
3. --version 10.2.12 \
4. --set initialClusterState="existing" \
5. --set removeMemberOnContainerTermination=false \
6. --set replicaCount=3 \
7. --set auth.rbac.create=false \
8. --set auth.rbac.token.enabled=false \
9. --create-namespace \
10. -n etcd-cluster

接着，将 Metasrv 的 etcdEndpoints 改成新的 etcd recover 集群，本例中为 "etcd-recover.etcd-cluster.svc.cluster.local:2379"：

1. apiVersion: greptime.io/v1alpha1
2. kind: GreptimeDBCluster
3. metadata:
4. name: greptimedb
5. spec:
6. # 其他配置
7. meta:
8. etcdEndpoints:
9. - "etcd-recover.etcd-cluster.svc.cluster.local:2379"

然后重启 GreptimeDB Metasrv 完成 etcd 恢复。