From 67588ee5d75f59ed42e72f322bf0665e3f9a1dd7 Mon Sep 17 00:00:00 2001 From: Chris Chen Date: Tue, 31 Oct 2023 10:27:03 +0800 Subject: [PATCH] Add operator port (#2320) * Add operator port * Update port desc * Update docs-2.0-zh/nebula-operator/8.custom-cluster-configurations/8.1.custom-conf-parameter.md * Update port --------- Co-authored-by: abby.huang <78209557+abby-cyber@users.noreply.github.com> --- .../8.1.custom-conf-parameter.md | 115 ++++++++++++++++-- .../8.1.custom-conf-parameter.md | 108 +++++++++++++++- mkdocs.yml | 2 +- 3 files changed, 210 insertions(+), 15 deletions(-) diff --git a/docs-2.0-en/nebula-operator/8.custom-cluster-configurations/8.1.custom-conf-parameter.md b/docs-2.0-en/nebula-operator/8.custom-cluster-configurations/8.1.custom-conf-parameter.md index 8ca11e3f7e8..f75c4564bde 100644 --- a/docs-2.0-en/nebula-operator/8.custom-cluster-configurations/8.1.custom-conf-parameter.md +++ b/docs-2.0-en/nebula-operator/8.custom-cluster-configurations/8.1.custom-conf-parameter.md @@ -1,6 +1,6 @@ -# Customize configuration parameters for a NebulaGraph cluster +# Customize parameters for a NebulaGraph cluster -Meta, Storage, and Graph services in a NebulaGraph Cluster have their own configuration settings, which are defined in the YAML file of the NebulaGraph cluster instance as `config`. These settings are mapped and loaded into the corresponding service's ConfigMap in Kubernetes. At the time of startup, the configuration present in the ConfigMap is mounted onto the directory `/usr/local/nebula/etc/` for every service. +Meta, Storage, and Graph services in a NebulaGraph cluster have their own configuration settings, which are defined in the YAML file of the NebulaGraph cluster instance as `config`. These settings are mapped and loaded into the corresponding service's ConfigMap in Kubernetes. At the time of startup, the configuration present in the ConfigMap is mounted onto the directory `/usr/local/nebula/etc/` for every service. !!! note @@ -11,6 +11,7 @@ The structure of `config` is as follows. ```go Config map[string]string `json:"config,omitempty"` ``` + ## Prerequisites You have created a NebulaGraph cluster. For how to create a cluster with Kubectl, see [Create a cluster with Kubectl](../3.deploy-nebula-graph-cluster/3.1create-cluster-with-kubectl.md). @@ -25,7 +26,7 @@ The following example uses a cluster named `nebula` and the cluster's configurat kubectl edit nebulaclusters.apps.nebula-graph.io nebula ``` -2. Add `enable_authorize` and `auth_type` under `spec.graphd.config`. +2. Customize parameters under the `spec.graphd.config` field. In the following sample, the `enable_authorize` and `auth_type` parameters are used for demonstration purposes. ```yaml apiVersion: apps.nebula-graph.io/v1alpha1 @@ -55,20 +56,118 @@ The following example uses a cluster named `nebula` and the cluster's configurat "auth_type": "password" ... ``` + + The parameters that can be added under the `config` field are listed in detail in the [Meta service configuration parameters](../../5.configurations-and-logs/1.configurations/2.meta-config.md), [Storage service configuration parameters](../../5.configurations-and-logs/1.configurations/4.storage-config.md), and [Graph service configuration parameters](../../5.configurations-and-logs/1.configurations/3.graph-config.md) topics. + + !!! note + + * To update cluster configurations without incurring pod restart, ensure that all parameters added under the `config` field support runtime dynamic modification. Check the **Whether supports runtime dynamic modifications** column of the parameter tables on the aforementioned parameter details pages to see if a parameter supports runtime dynamic modification. + * If one or more parameters that do not support runtime dynamic modification are added under the `config` field, pod restart is required for the parameters to take effect. + To add the `config` for the Meta and Storage services, add `spec.metad.config` and `spec.storaged.config` respectively. 3. Run `kubectl apply -f nebula_cluster.yaml` to push your configuration changes to the cluster. - After customizing the parameters `enable_authorize` and `auth_type`, the configurations in the corresponding ConfigMap (`nebula-graphd`) of the Graph service will be overwritten. + After customizing the parameters, the configurations in the corresponding ConfigMap (`nebula-graphd`) of the Graph service will be overwritten. + + +## Customize port configurations -## Modify cluster configurations online +You can add the `port` and `ws_http_port` parameters under the `config` field to customize port configurations. For details about these two parameters, see the Networking configurations section in [Meta service configuration parameters](../../5.configurations-and-logs/1.configurations/2.meta-config.md), [Storage service configuration parameters](../../5.configurations-and-logs/1.configurations/4.storage-config.md), and [Graph service configuration parameters](../../5.configurations-and-logs/1.configurations/3.graph-config.md). -Cluster configurations are modified online by calling the HTTP interface, without the need to restart the cluster Pod. +!!! note + + * Pod restart is required for the `port` and `ws_http_port` parameters to take effect. + * It is NOT recommnended to modify the `port` parameter after the cluster is started. + +1. Modifiy the cluster configuration file. + + ```yaml + apiVersion: apps.nebula-graph.io/v1alpha1 + kind: NebulaCluster + metadata: + name: nebula + namespace: default + spec: + graphd: + config: + port: "3669" + ws_http_port: "8080" + resources: + requests: + cpu: "200m" + memory: "500Mi" + limits: + cpu: "1" + memory: "1Gi" + replicas: 1 + image: vesoft/nebula-graphd + version: {{nebula.tag}} + metad: + config: + ws_http_port: 8081 + resources: + requests: + cpu: "300m" + memory: "500Mi" + limits: + cpu: "1" + memory: "1Gi" + replicas: 1 + image: vesoft/nebula-metad + version: {{nebula.tag}} + dataVolumeClaim: + resources: + requests: + storage: 2Gi + storageClassName: local-path + storaged: + config: + ws_http_port: 8082 + resources: + requests: + cpu: "300m" + memory: "500Mi" + limits: + cpu: "1" + memory: "1Gi" + replicas: 1 + image: vesoft/nebula-storaged + version: {{nebula.tag}} + dataVolumeClaims: + - resources: + requests: + storage: 2Gi + storageClassName: local-path + enableAutoBalance: true + reference: + name: statefulsets.apps + version: v1 + schedulerName: default-scheduler + imagePullPolicy: IfNotPresent + imagePullSecrets: + - name: nebula-image + enablePVReclaim: true + topologySpreadConstraints: + - topologyKey: kubernetes.io/hostname + whenUnsatisfiable: "ScheduleAnyway" + ``` -It should be noted that only when all configuration items in `config` are the parameters that can be dynamically modified at runtime, can the operation of online modifications be triggered. If the configuration items in `config` contain parameters that cannot be dynamically modified, then the cluster configuration will be updated by restarting the Pod. +2. Run the `kubectl apply -f nebula_cluster.yaml` to push your configuration changes to the cluster. -For information about the parameters that can be dynamically modified for each service, see the parameter table column of **Whether supports runtime dynamic modifications** in [Meta service configuration parameters](../../5.configurations-and-logs/1.configurations/2.meta-config.md), [Storage service configuration parameters](../../5.configurations-and-logs/1.configurations/4.storage-config.md), and [Graph service configuration parameters](../../5.configurations-and-logs/1.configurations/3.graph-config.md), respectively. +3. Verify that the configuration takes effect. + ```bash + kubectl get svc + ``` + Sample response: + ``` + NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE + nebula-graphd-headless ClusterIP None 3669/TCP,8080/TCP 10m + nebula-graphd-svc ClusterIP 10.102.13.115 3669/TCP,8080/TCP 10m + nebula-metad-headless ClusterIP None 9559/TCP,8081/TCP 11m + nebula-storaged-headless ClusterIP None 9779/TCP,8082/TCP,9778/TCP 11m + ``` diff --git a/docs-2.0-zh/nebula-operator/8.custom-cluster-configurations/8.1.custom-conf-parameter.md b/docs-2.0-zh/nebula-operator/8.custom-cluster-configurations/8.1.custom-conf-parameter.md index bf696f1e47f..2c83dbc0683 100644 --- a/docs-2.0-zh/nebula-operator/8.custom-cluster-configurations/8.1.custom-conf-parameter.md +++ b/docs-2.0-zh/nebula-operator/8.custom-cluster-configurations/8.1.custom-conf-parameter.md @@ -26,7 +26,7 @@ Config map[string]string `json:"config,omitempty"` kubectl edit nebulaclusters.apps.nebula-graph.io nebula ``` -2. 在 YAML 文件的`spec.graphd.config`配置项中，添加`enable_authorize`和`auth_type`。 +2. 在 YAML 文件的`spec.graphd.config`配置项中，添加需要修改的参数。下文以 `enable_authorize`和`auth_type`为例。 ```yaml apiVersion: apps.nebula-graph.io/v1alpha1 @@ -56,19 +56,115 @@ Config map[string]string `json:"config,omitempty"` "auth_type": "password" ... ``` + 在`config`字段下可配置的参数详情，请分别参见 [Meta 服务配置参数](../../5.configurations-and-logs/1.configurations/2.meta-config.md)、[Storage 服务配置参数](../../5.configurations-and-logs/1.configurations/4.storage-config.md)、[Graph 服务配置参数](../../5.configurations-and-logs/1.configurations/3.graph-config.md)。 + + !!! note + + * 若要在集群运行时动态修改参数配置且不触发 Pod 重启，请确保当前修改的参数全部支持运行时动态修改。参数是否支持运行时动态修改，请查看上述参数详情页各个表格中**是否支持运行时动态修改**一列。 + * 若本次修改的参数包含一个或多个不支持运行时动态修改的参数，则会触发 Pod 重启。 如果需要为 Meta 服务和 Storage 服务配置`config`，则在`spec.metad.config`和`spec.storaged.config`中添加对应的配置项。 3. 执行`kubectl apply -f nebula_cluster.yaml`使上述更新生效。 - 在自定义参数`enable_authorize`和`auth_type`后，Graph 服务对应的 ConfigMap（`nebula-graphd`）中的配置将被覆盖。 + 在修改参数值后，Graph 服务对应的 ConfigMap（`nebula-graphd`）中的配置将被覆盖。 +### 配置自定义端口 -## 在线修改集群配置 +您可以在`config`字段中添加`port`和`ws_http_port`参数，从而配置自定义的端口。这两个参数的详细信息，请参见[Meta 服务配置参数](../../5.configurations-and-logs/1.configurations/2.meta-config.md)、[Storage 服务配置参数](../../5.configurations-and-logs/1.configurations/4.storage-config.md)、[Graph 服务配置参数](../../5.configurations-and-logs/1.configurations/3.graph-config.md)的 networking 配置一节。 -在线修改集群配置主要通过调用`HTTP`接口对集群的配置参数进行修改，而无需重启集群 Pod。 +!!! note -需要注意的是，只有当`config`中的配置项均为可在运行时动态修改的参数时，才能触发在线修改集群配置操作。如果`config`中的配置项包含不可动态修改的参数，那么集群配置将通过重启 Pod 的方式进行更新。 + * 自定义`port`和`ws_http_port`参数配置后，会触发 Pod 重启，并在重启后生效。 + * 在集群启动后，不建议修改`port`参数。 -关于各个服务可动态修改的参数，分别参见 [Meta 服务配置参数](../../5.configurations-and-logs/1.configurations/2.meta-config.md)、[Storage 服务配置参数](../../5.configurations-and-logs/1.configurations/4.storage-config.md)、[Graph 服务配置参数](../../5.configurations-and-logs/1.configurations/3.graph-config.md)的配置参数表格中**是否支持运行时动态修改**一列。 +1. 修改集群配置文件。 + ```yaml + apiVersion: apps.nebula-graph.io/v1alpha1 + kind: NebulaCluster + metadata: + name: nebula + namespace: default + spec: + graphd: + config: + port: "3669" + ws_http_port: "8080" + resources: + requests: + cpu: "200m" + memory: "500Mi" + limits: + cpu: "1" + memory: "1Gi" + replicas: 1 + image: vesoft/nebula-graphd + version: {{nebula.tag}} + metad: + config: + ws_http_port: 8081 + resources: + requests: + cpu: "300m" + memory: "500Mi" + limits: + cpu: "1" + memory: "1Gi" + replicas: 1 + image: vesoft/nebula-metad + version: {{nebula.tag}} + dataVolumeClaim: + resources: + requests: + storage: 2Gi + storageClassName: local-path + storaged: + config: + ws_http_port: 8082 + resources: + requests: + cpu: "300m" + memory: "500Mi" + limits: + cpu: "1" + memory: "1Gi" + replicas: 1 + image: vesoft/nebula-storaged + version: {{nebula.tag}} + dataVolumeClaims: + - resources: + requests: + storage: 2Gi + storageClassName: local-path + enableAutoBalance: true + reference: + name: statefulsets.apps + version: v1 + schedulerName: default-scheduler + imagePullPolicy: IfNotPresent + imagePullSecrets: + - name: nebula-image + enablePVReclaim: true + topologySpreadConstraints: + - topologyKey: kubernetes.io/hostname + whenUnsatisfiable: "ScheduleAnyway" + ``` + +2. 执行`kubectl apply -f nebula_cluster.yaml`使上述更新生效。 + +3. 验证配置已经生效。 + + ```bash + kubectl get svc + ``` + + 返回示例： + + ``` + NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE + nebula-graphd-headless ClusterIP None 3669/TCP,8080/TCP 10m + nebula-graphd-svc ClusterIP 10.102.13.115 3669/TCP,8080/TCP 10m + nebula-metad-headless ClusterIP None 9559/TCP,8081/TCP 11m + nebula-storaged-headless ClusterIP None 9779/TCP,8082/TCP,9778/TCP 11m + ``` \ No newline at end of file diff --git a/mkdocs.yml b/mkdocs.yml index 499ca89a062..9d83e3a8e14 100644 --- a/mkdocs.yml +++ b/mkdocs.yml @@ -582,7 +582,7 @@ nav: - Deploy clusters with Helm: nebula-operator/3.deploy-nebula-graph-cluster/3.2create-cluster-with-helm.md - Connect to NebulaGraph databases: nebula-operator/4.connect-to-nebula-graph-service.md - Configure clusters: - - Custom configuration parameters for a NebulaGraph cluster: nebula-operator/8.custom-cluster-configurations/8.1.custom-conf-parameter.md + - Customize parameters for a NebulaGraph cluster: nebula-operator/8.custom-cluster-configurations/8.1.custom-conf-parameter.md - Storage: - Dynamically expand persistent volumes: nebula-operator/8.custom-cluster-configurations/storage/8.6.dynamic-expand-pv.md - Reclaim PVs: nebula-operator/8.custom-cluster-configurations/storage/8.2.pv-reclaim.md