使用 Helm 部署 TDengine 集群

Helm 是 Kubernetes 的包管理器，上一节使用 Kubernetes 部署 TDengine 集群的操作已经足够简单，但 Helm 依然可以提供更强大的能力。

安装 Helm

1. curl -fsSL -o get_helm.sh \
2. https://raw.githubusercontent.com/helm/helm/master/scripts/get-helm-3
3. chmod +x get_helm.sh
4. ./get_helm.sh

Helm 会使用 kubectl 和 kubeconfig 的配置来操作 Kubernetes，可以参考 Rancher 安装 Kubernetes 的配置来进行设置。

安装 TDengine Chart

TDengine Chart 尚未发布到 Helm 仓库，当前可以从 GitHub 直接下载：

1. wget https://github.com/taosdata/TDengine-Operator/raw/3.0/helm/tdengine-3.0.2.tgz

获取当前 Kubernetes 的存储类：

1. kubectl get storageclass

在 minikube 默认为 standard.

之后，使用 helm 命令安装：

1. helm install tdengine tdengine-3.0.2.tgz \
2. --set storage.className=<your storage class name>

在 minikube 环境下，可以设置一个较小的容量避免超出磁盘可用空间：

1. helm install tdengine tdengine-3.0.2.tgz \
2. --set storage.className=standard \
3. --set storage.dataSize=2Gi \
4. --set storage.logSize=10Mi

部署成功后，TDengine Chart 将会输出操作 TDengine 的说明：

1. export POD_NAME=$(kubectl get pods --namespace default \
2. -l "app.kubernetes.io/name=tdengine,app.kubernetes.io/instance=tdengine" \
3. -o jsonpath="{.items[0].metadata.name}")
4. kubectl --namespace default exec $POD_NAME -- taos -s "show dnodes; show mnodes"
5. kubectl --namespace default exec -it $POD_NAME -- taos

可以创建一个表进行测试：

1. kubectl --namespace default exec $POD_NAME -- \
2. taos -s "create database test;
3. use test;
4. create table t1 (ts timestamp, n int);
5. insert into t1 values(now, 1)(now + 1s, 2);
6. select * from t1;"

配置 Values

TDengine 支持 values.yaml 自定义。

通过 helm show values 可以获取 TDengine Chart 支持的全部 values 列表：

1. helm show values tdengine-3.0.2.tgz

你可以将结果保存为 values.yaml，之后可以修改其中的各项参数，如 replica 数量，存储类名称，容量大小，TDengine 配置等，然后使用如下命令安装 TDengine 集群：

1. helm install tdengine tdengine-3.0.2.tgz -f values.yaml

全部参数如下：

1. # Default values for tdengine.
2. # This is a YAML-formatted file.
3. # Declare variables to be passed into helm templates.

5. replicaCount: 1

7. image:
8. prefix: tdengine/tdengine
9. #pullPolicy: Always
10. # Overrides the image tag whose default is the chart appVersion.
11. #  tag: "3.0.2.0"

13. service:
14. # ClusterIP is the default service type, use NodeIP only if you know what you are doing.
15. type: ClusterIP
16. ports:
17. # TCP range required
18. tcp: [6030, 6041, 6042, 6043, 6044, 6046, 6047, 6048, 6049, 6060]
19. # UDP range
20. udp: [6044, 6045]

23. # Set timezone here, not in taoscfg
24. timezone: "Asia/Shanghai"

26. resources:
27. # We usually recommend not to specify default resources and to leave this as a conscious
28. # choice for the user. This also increases chances charts run on environments with little
29. # resources, such as Minikube. If you do want to specify resources, uncomment the following
30. # lines, adjust them as necessary, and remove the curly braces after 'resources:'.
31. # limits:
32. #   cpu: 100m
33. #   memory: 128Mi
34. # requests:
35. #   cpu: 100m
36. #   memory: 128Mi

38. storage:
39. # Set storageClassName for pvc. K8s use default storage class if not set.
40. #
41. className: ""
42. dataSize: "100Gi"
43. logSize: "10Gi"

45. nodeSelectors:
46. taosd:
47. # node selectors

49. clusterDomainSuffix: ""
50. # Config settings in taos.cfg file.
51. #
52. # The helm/k8s support will use environment variables for taos.cfg,
53. # converting an upper-snake-cased variable like `TAOS_DEBUG_FLAG`,
54. # to a camelCase taos config variable `debugFlag`.
55. #
56. # See the variable list at https://www.taosdata.com/cn/documentation/administrator .
57. #
58. # Note:
59. # 1. firstEp/secondEp: should not be set here, it's auto generated at scale-up.
60. # 2. serverPort: should not be set, we'll use the default 6030 in many places.
61. # 3. fqdn: will be auto generated in kubernetes, user should not care about it.
62. # 4. role: currently role is not supported - every node is able to be mnode and vnode.
63. #
64. # Btw, keep quotes "" around the value like below, even the value will be number or not.
65. taoscfg:
66. # Starts as cluster or not, must be 0 or 1.
67. #   0: all pods will start as a separate TDengine server
68. #   1: pods will start as TDengine server cluster. [default]
69. CLUSTER: "1"

71. # number of replications, for cluster only
72. TAOS_REPLICA: "1"

75. # TAOS_NUM_OF_RPC_THREADS: number of threads for RPC
76. #TAOS_NUM_OF_RPC_THREADS: "2"

78. #
79. # TAOS_NUM_OF_COMMIT_THREADS: number of threads to commit cache data
80. #TAOS_NUM_OF_COMMIT_THREADS: "4"

82. # enable/disable installation / usage report
83. #TAOS_TELEMETRY_REPORTING: "1"

85. # time interval of system monitor, seconds
86. #TAOS_MONITOR_INTERVAL: "30"

88. # time interval of dnode status reporting to mnode, seconds, for cluster only
89. #TAOS_STATUS_INTERVAL: "1"

91. # time interval of heart beat from shell to dnode, seconds
92. #TAOS_SHELL_ACTIVITY_TIMER: "3"

94. # minimum sliding window time, milli-second
95. #TAOS_MIN_SLIDING_TIME: "10"

97. # minimum time window, milli-second
98. #TAOS_MIN_INTERVAL_TIME: "1"

100. # the compressed rpc message, option:
101. #  -1 (no compression)
102. #   0 (all message compressed),
103. # > 0 (rpc message body which larger than this value will be compressed)
104. #TAOS_COMPRESS_MSG_SIZE: "-1"

106. # max number of connections allowed in dnode
107. #TAOS_MAX_SHELL_CONNS: "50000"

109. # stop writing logs when the disk size of the log folder is less than this value
110. #TAOS_MINIMAL_LOG_DIR_G_B: "0.1"

112. # stop writing temporary files when the disk size of the tmp folder is less than this value
113. #TAOS_MINIMAL_TMP_DIR_G_B: "0.1"

115. # if disk free space is less than this value, taosd service exit directly within startup process
116. #TAOS_MINIMAL_DATA_DIR_G_B: "0.1"

118. # One mnode is equal to the number of vnode consumed
119. #TAOS_MNODE_EQUAL_VNODE_NUM: "4"

121. # enbale/disable http service
122. #TAOS_HTTP: "1"

124. # enable/disable system monitor
125. #TAOS_MONITOR: "1"

127. # enable/disable async log
128. #TAOS_ASYNC_LOG: "1"

130. #
131. # time of keeping log files, days
132. #TAOS_LOG_KEEP_DAYS: "0"

134. # The following parameters are used for debug purpose only.
135. # debugFlag 8 bits mask: FILE-SCREEN-UNUSED-HeartBeat-DUMP-TRACE_WARN-ERROR
136. # 131: output warning and error
137. # 135: output debug, warning and error
138. # 143: output trace, debug, warning and error to log
139. # 199: output debug, warning and error to both screen and file
140. # 207: output trace, debug, warning and error to both screen and file
141. #
142. # debug flag for all log type, take effect when non-zero value\
143. #TAOS_DEBUG_FLAG: "143"

145. # generate core file when service crash
146. #TAOS_ENABLE_CORE_FILE: "1"

扩容

关于扩容可参考上一节的说明，有一些额外的操作需要从 helm 的部署中获取。

首先，从部署中获取 StatefulSet 的名称。

1. export STS_NAME=$(kubectl get statefulset \
2. -l "app.kubernetes.io/name=tdengine" \
3. -o jsonpath="{.items[0].metadata.name}")

扩容操作极其简单，增加 replica 即可。以下命令将 TDengine 扩充到三节点：

1. kubectl scale --replicas 3 statefulset/$STS_NAME

使用命令 show dnodes 和 show mnodes 检查是否扩容成功。

缩容

Helm - 图1warning

缩容操作并没有完整测试，可能造成数据风险，请谨慎使用。

获取需要缩容的 dnode 列表，并手动 Drop。

1. kubectl --namespace default exec $POD_NAME -- \
2. cat /var/lib/taos/dnode/dnodeEps.json \
3. | jq '.dnodeInfos[1:] |map(.dnodeFqdn + ":" + (.dnodePort|tostring)) | .[]' -r
4. kubectl --namespace default exec $POD_NAME -- taos -s "show dnodes"
5. kubectl --namespace default exec $POD_NAME -- taos -s 'drop dnode "<you dnode in list>"'

清理集群

Helm 管理下，清理操作也变得简单：

1. helm uninstall tdengine

但 Helm 也不会自动移除 PVC，需要手动获取 PVC 然后删除掉。