分片故障转移请求

用于对数据库分片进行故障转移的 REST API 请求

Redis 企业软件
方法 路径 描述
发布 /v1/shards/actions/failover 故障转移多个分片
发布 /v1/shards/{uid}/actions/failover 故障转移特定分片

故障转移多个分片

POST /v1/shards/actions/failover

对shard_uids，并将其副本提升为主分片。此请求是异步的。

集群会自动管理故障转移以确保高可用性。仅将此故障转移 REST API 请求用于测试和计划内维护。

所需权限

权限名称 角色
failover_shard admin
cluster_member
db_member

请求

HTTP 请求示例

POST /v1/shards/actions/failover

示例 JSON 正文

{
    "shard_uids": ["2","4","6"]
}

Request headers

Key Value Description
Host cnm.cluster.fqdn Domain name
Accept application/json Accepted media type

Request body

The request body is a JSON object that can contain the following fields:

Field Type Description
shard_uids array of strings List of primary shard UIDs to fail over. The shards must belong to the same database.
dead_uids array of strings Primary shards to avoid stopping. Optional.
dead_nodes array of strings Nodes that should not be drained or used for promoted replica shards. Optional.
dry_run boolean Determines whether the failover is actually done. If true, will just do a dry run. If the dry run succeeds, the request returns a 200 OK status code. Otherwise, it returns a JSON object with an error code and description. Optional.
force_rebind boolean Rebind after promotion. Optional.
redis_version_upgrade string New version of the promoted primary shards. Optional.

Response

Returns a JSON object with an action_uid. You can track the action's progress with a GET /v1/actions/<action_uid> request.

Example JSON body

{
    "action_uid": "e5e24ddf-a456-4a7e-ad53-4463cd44880e",
    "description": "Failover was triggered"
}

Status codes

Code Description
200 OK No error.
400 Bad Request Shard is a replica or the specified failover shards are not in the same database.
404 Not Found A list of shard UIDs is required and not given, or a specified shard does not exist.
409 Conflict Database is currently busy.

Error codes

When errors are reported, the server may return a JSON object with error_code and message field that provide additional information. The following are possible error_code values:

Code Description
db_busy Database is currently busy.
failover_shards_different_bdb All failover shards should be in the same database.
shard_is_slave Shard is a replica.
shard_not_exist Shard does not exist.
shard_uids_required List of shard UIDs is required and not given.

Fail over shard

POST /v1/shards/{int: uid}/actions/failover

Performs failover on the primary shard with the specified shard_uid, and promotes its replica shard to a primary shard. This request is asynchronous.

The cluster automatically manages failover to ensure high availability. Use this failover REST API request only for testing and planned maintenance.

Required permissions

Permission name Roles
failover_shard admin
cluster_member
db_member

Request

Example HTTP request

POST /v1/shards/1/actions/failover

Example JSON body

{
    "force_rebind": true
}

Request headers

Key Value Description
Host cnm.cluster.fqdn Domain name
Accept application/json Accepted media type

URL parameters

Field Type Description
uid integer The unique ID of the shard to fail over.

Request body

The request body is a JSON object that can contain the following fields:

Field Type Description
dead_uid string Primary shard to avoid stopping. Optional.
dead_nodes array of strings Nodes that should not be drained or used for promoted replica shards. Optional.
dry_run boolean Determines whether the failover is actually done. If true, will just do a dry run. If the dry run succeeds, the request returns a 200 OK status code. Otherwise, it returns a JSON object with an error code and description. Optional.
force_rebind boolean Rebind after promotion. Optional.
redis_version_upgrade string New version of the promoted primary shards. Optional.

Response

Returns a JSON object with an action_uid. You can track the action's progress with a GET /v1/actions/<action_uid> request.

Example JSON body

{
    "action_uid": "e5e24ddf-a456-4a7e-ad53-4463cd44880e",
    "description": "Failover was triggered"
}

Status codes

Code Description
200 OK No error.
400 Bad Request Shard is a replica.
404 Not Found Specified shard does not exist.
409 Conflict Database is currently busy.

Error codes

When errors are reported, the server may return a JSON object with error_code and message field that provide additional information. The following are possible error_code values:

Code Description
db_busy Database is currently busy.
shard_is_slave Shard is a replica.
shard_not_exist Shard does not exist.
RATE THIS PAGE
Back to top ↑