docs: document ADMIN function events

docs/reference/sql/greptime-private/events.md

@@ -13,7 +13,7 @@ operation succeeded.
 
 The table is created when the first event is recorded. If no event has been
 recorded, or event recording is disabled, querying it returns a table-not-found
-error. Configure recording in [Lifecycle event recorder](/user-guide/deployments-administration/configuration.md#lifecycle-event-recorder).
+error. Configure recording in [Event recording](/user-guide/deployments-administration/configuration.md#event-recording).
 
 ```sql
 USE greptime_private;

docs/user-guide/deployments-administration/configuration.md

@@ -506,9 +506,9 @@ default_ratio = 1.0
 
 How to use distributed tracing, please reference [Tracing](/user-guide/deployments-administration/monitoring/tracing.md#tutorial-use-jaeger-to-trace-greptimedb)
 
-### Lifecycle event recorder
+### Event recording
 
-The lifecycle event recorder persists procedure lifecycle events in the `greptime_private.events` system table. Configure it in `standalone` mode or on Metasrv in a distributed deployment:
+Recorded events are stored in the `greptime_private.events` system table.
 
 ```toml
 [event_recorder]
@@ -530,11 +530,17 @@ create_database, alter_database, drop_database,
 create_flow, drop_flow,
 create_table, create_logical_tables, alter_table, alter_logical_tables,
 drop_table, undrop_table, purge_dropped_table, truncate_table,
-create_view, drop_view
+create_view, drop_view, admin_function
 ```
 
 `undrop_table` and `purge_dropped_table` require GreptimeDB Enterprise.
 
+In distributed deployments, the Frontend supports the following event type:
+
+```text
+admin_function
+```
+
 Metasrv supports the following event types:
 
 ```text

docs/user-guide/deployments-administration/monitoring/events/event-data-model.md

@@ -50,6 +50,20 @@ not guaranteed to appear in the order shown:
 | `Failed`         | The Procedure reached a failed terminal state. Inspect `procedure_error` for failure details.                                                                                            |
 | `Poisoned`       | The Procedure cannot proceed. Inspect `procedure_error` for the failure details.                                                                                                         |
 
+## ADMIN function event columns
+
+An `admin_function` event records the result returned by an `ADMIN` statement.
+
+| Column                   | Meaning                                                                                                                                |
+| ------------------------ | -------------------------------------------------------------------------------------------------------------------------------------- |
+| `actor`                  | The current database user that executed the `ADMIN` function.                                                                        |
+| `admin_function_name`    | The name of the executed ADMIN function.                                                                                              |
+| `admin_function_status`  | The execution status: `Succeeded` or `Failed`.                                                                                        |
+| `admin_function_output`  | JSON containing `result` when the function succeeds or `error` when it fails.                                                         |
+
+The event `payload` contains the function arguments. If the result is a
+Procedure ID, query the Procedure events for that ID to track its progress.
+
 ## Query JSON fields
 
 See the [JSON functions](/reference/sql/functions/json.md) reference for

docs/user-guide/deployments-administration/monitoring/events/overview.md

@@ -14,9 +14,7 @@ acknowledgement.
 
 ## Configure event recording
 
-The event recorder can be configured in standalone deployments or on Metasrv in distributed deployments. See [Lifecycle event recorder](/user-guide/deployments-administration/configuration.md#lifecycle-event-recorder) for configuration options and the supported type list.
-
-Standalone deployments record supported local DDL Procedure events. Distributed deployments with Metasrv can additionally record operational event types.
+See [Event recording](/user-guide/deployments-administration/configuration.md#event-recording) for configuration options and the event types supported by standalone, Frontend, and Metasrv.
 
 ## Query events
 

docs/user-guide/deployments-administration/monitoring/events/query-events.md

@@ -54,6 +54,36 @@ cluster. It varies with workload and does not define the configured or
 source-supported types. See [DDL events](/user-guide/deployments-administration/monitoring/events/ddl-events.md)
 for the supported local DDL event types.
 
+## Query ADMIN function events
+
+An `admin_function` event records the function name, the current database user,
+the immediate status, the input arguments, and the immediate output:
+
+```sql
+SELECT timestamp,
+       actor,
+       admin_function_name,
+       admin_function_status,
+       json_to_string(payload) AS payload,
+       json_to_string(admin_function_output) AS output
+FROM greptime_private.events
+WHERE type = 'admin_function'
+  AND timestamp >= now() - INTERVAL '1' hour
+ORDER BY timestamp DESC
+LIMIT 20;
+```
+
+For a successful function, `output` contains a `result`. For a failed function,
+it contains an `error`:
+
+```text
+| actor | admin_function_name | admin_function_status | output        |
+| root  | flush_table         | Succeeded              | {"result":0} |
+| root  | unknown_function     | Failed                 | {"error":"..."} |
+```
+
+The `actor` value comes from the current protocol session user.
+
 Combine an event type with a database and object name to avoid unrelated rows:
 
 ```sql

i18n/zh/docusaurus-plugin-content-docs/current/reference/sql/greptime-private/events.md

@@ -7,7 +7,7 @@ description: greptime_private 数据库中的 events 表。
 
 `events` 表保存 GreptimeDB 运行期间产生的事件。单机部署记录受支持的本地 DDL Procedure 事件；带有 Metasrv 的分布式部署还可以记录运维事件。事件会异步写入，写入失败不会影响原操作的执行结果。因此，不能以事件是否出现作为操作成功的依据。
 
-该表会在首次记录事件时自动创建。如果还没有事件记录，或已禁用事件记录，查询会提示表不存在。配置方法请参阅[生命周期事件记录器](/user-guide/deployments-administration/configuration.md#生命周期事件记录器)。
+该表会在首次记录事件时自动创建。如果还没有事件记录，或已禁用事件记录，查询会提示表不存在。配置方法请参阅[事件记录配置](/user-guide/deployments-administration/configuration.md#事件记录配置)。
 
 ```sql
 USE greptime_private;

i18n/zh/docusaurus-plugin-content-docs/current/user-guide/deployments-administration/configuration.md

@@ -508,9 +508,9 @@ default_ratio = 1.0
 
 如何使用分布式追踪，请参考 [Tracing](/user-guide/deployments-administration/monitoring/tracing.md#教程使用-jaeger-追踪-greptimedb-调用链路)
 
-### 生命周期事件记录器
+### 事件记录配置
 
-生命周期事件记录器会将 Procedure 的生命周期事件持久化到 `greptime_private.events` 系统表。在单机模式下使用 `standalone` 配置；在分布式部署中，在 Metasrv 上配置：
+记录的事件保存在 `greptime_private.events` 系统表中。
 
 ```toml
 [event_recorder]
@@ -532,11 +532,17 @@ create_database, alter_database, drop_database,
 create_flow, drop_flow,
 create_table, create_logical_tables, alter_table, alter_logical_tables,
 drop_table, undrop_table, purge_dropped_table, truncate_table,
-create_view, drop_view
+create_view, drop_view, admin_function
 ```
 
 `undrop_table` 和 `purge_dropped_table` 仅 GreptimeDB 企业版支持。
 
+在分布式部署中，Frontend 支持以下事件类型：
+
+```text
+admin_function
+```
+
 Metasrv 支持以下事件类型：
 
 ```text

i18n/zh/docusaurus-plugin-content-docs/current/user-guide/deployments-administration/monitoring/events/event-data-model.md

@@ -46,6 +46,20 @@ Procedure 事件还有以下列：
 | `Failed`         | Procedure 到达失败终态。请检查 `procedure_error` 中的失败详情。                                                                                                     |
 | `Poisoned`       | Procedure 无法继续。请检查 `procedure_error` 中的失败详情。                                                                                                         |
 
+## 管理函数事件列
+
+`admin_function` 事件记录 `ADMIN` 语句返回的结果。
+
+| 列                      | 含义                                                                                                            |
+| ----------------------- | --------------------------------------------------------------------------------------------------------------- |
+| `actor`                 | 执行 `ADMIN` 语句的当前数据库用户。                                                                              |
+| `admin_function_name`   | 执行的管理函数名称。                                                                                             |
+| `admin_function_status` | 执行状态：`Succeeded` 或 `Failed`。                                                                             |
+| `admin_function_output` | JSON 数据。函数成功时包含 `result`，函数失败时包含 `error`。                                                     |
+
+事件的 `payload` 包含函数参数。如果返回结果是 Procedure ID，可使用该 ID 查询
+Procedure 事件以查看执行进度。
+
 ## 查询 JSON 字段
 
 详细信息请参阅 [JSON 函数](/reference/sql/functions/json.md)。在事件查询中，

i18n/zh/docusaurus-plugin-content-docs/current/user-guide/deployments-administration/monitoring/events/overview.md

@@ -11,9 +11,7 @@ GreptimeDB 会异步写入事件记录，通常每 5 秒刷新一次。刚完成
 
 ## 配置事件记录
 
-单机部署可以在本地配置事件记录器，分布式部署则可以在 Metasrv 上配置。配置选项和支持的事件类型请参阅[生命周期事件记录器](/user-guide/deployments-administration/configuration.md#生命周期事件记录器)。
-
-单机部署会记录受支持的本地 DDL Procedure 事件。带有 Metasrv 的分布式部署还可以记录更多运维事件。
+配置选项以及 standalone、Frontend 和 Metasrv 支持的事件类型请参阅[事件记录配置](/user-guide/deployments-administration/configuration.md#事件记录配置)。
 
 ## 查询事件
 

i18n/zh/docusaurus-plugin-content-docs/current/user-guide/deployments-administration/monitoring/events/query-events.md

@@ -47,6 +47,35 @@ ORDER BY type;
 该结果仅反映最近一小时内实际出现的事件类型，不能作为已配置或受支持类型的完整清单。
 支持的本地 DDL 事件类型请参阅 [DDL 事件](/user-guide/deployments-administration/monitoring/events/ddl-events.md)。
 
+## 查询管理函数事件
+
+`admin_function` 事件记录管理函数名称、当前数据库用户、立即执行状态、输入参数
+和立即返回结果：
+
+```sql
+SELECT timestamp,
+       actor,
+       admin_function_name,
+       admin_function_status,
+       json_to_string(payload) AS payload,
+       json_to_string(admin_function_output) AS output
+FROM greptime_private.events
+WHERE type = 'admin_function'
+  AND timestamp >= now() - INTERVAL '1' hour
+ORDER BY timestamp DESC
+LIMIT 20;
+```
+
+管理函数执行成功时，`output` 包含 `result`；执行失败时，包含 `error`：
+
+```text
+| actor | admin_function_name | admin_function_status | output        |
+| root  | flush_table         | Succeeded              | {"result":0} |
+| root  | unknown_function     | Failed                 | {"error":"..."} |
+```
+
+`actor` 的值来自当前协议会话用户。
+
 将事件类型、数据库和对象名称组合，可以避免混入无关事件：
 
 ```sql

i18n/zh/docusaurus-plugin-content-docs/version-1.2/reference/sql/greptime-private/events.md

@@ -7,7 +7,7 @@ description: greptime_private 数据库中的 events 表。
 
 `events` 表保存 GreptimeDB 运行期间产生的事件。单机部署记录受支持的本地 DDL Procedure 事件；带有 Metasrv 的分布式部署还可以记录运维事件。事件会异步写入，写入失败不会影响原操作的执行结果。因此，不能以事件是否出现作为操作成功的依据。
 
-该表会在首次记录事件时自动创建。如果还没有事件记录，或已禁用事件记录，查询会提示表不存在。配置方法请参阅[生命周期事件记录器](/user-guide/deployments-administration/configuration.md#生命周期事件记录器)。
+该表会在首次记录事件时自动创建。如果还没有事件记录，或已禁用事件记录，查询会提示表不存在。配置方法请参阅[事件记录配置](/user-guide/deployments-administration/configuration.md#事件记录配置)。
 
 ```sql
 USE greptime_private;

i18n/zh/docusaurus-plugin-content-docs/version-1.2/user-guide/deployments-administration/configuration.md

@@ -508,9 +508,9 @@ default_ratio = 1.0
 
 如何使用分布式追踪，请参考 [Tracing](/user-guide/deployments-administration/monitoring/tracing.md#教程使用-jaeger-追踪-greptimedb-调用链路)
 
-### 生命周期事件记录器
+### 事件记录配置
 
-生命周期事件记录器会将 Procedure 的生命周期事件持久化到 `greptime_private.events` 系统表。在单机模式下使用 `standalone` 配置；在分布式部署中，在 Metasrv 上配置：
+记录的事件保存在 `greptime_private.events` 系统表中。
 
 ```toml
 [event_recorder]

i18n/zh/docusaurus-plugin-content-docs/version-1.2/user-guide/deployments-administration/monitoring/events/overview.md

@@ -11,7 +11,7 @@ GreptimeDB 会异步写入事件记录，通常每 5 秒刷新一次。刚完成
 
 ## 配置事件记录
 
-单机部署可以在本地配置事件记录器，分布式部署则可以在 Metasrv 上配置。配置选项和支持的事件类型请参阅[生命周期事件记录器](/user-guide/deployments-administration/configuration.md#生命周期事件记录器)。
+配置选项以及 standalone 和 Metasrv 支持的事件类型请参阅[事件记录配置](/user-guide/deployments-administration/configuration.md#事件记录配置)。
 
 单机部署会记录受支持的本地 DDL Procedure 事件。带有 Metasrv 的分布式部署还可以记录更多运维事件。
 

versioned_docs/version-1.2/reference/sql/greptime-private/events.md

@@ -13,7 +13,7 @@ operation succeeded.
 
 The table is created when the first event is recorded. If no event has been
 recorded, or event recording is disabled, querying it returns a table-not-found
-error. Configure recording in [Lifecycle event recorder](/user-guide/deployments-administration/configuration.md#lifecycle-event-recorder).
+error. Configure recording in [Event recording](/user-guide/deployments-administration/configuration.md#event-recording).
 
 ```sql
 USE greptime_private;

versioned_docs/version-1.2/user-guide/deployments-administration/configuration.md

@@ -506,9 +506,9 @@ default_ratio = 1.0
 
 How to use distributed tracing, please reference [Tracing](/user-guide/deployments-administration/monitoring/tracing.md#tutorial-use-jaeger-to-trace-greptimedb)
 
-### Lifecycle event recorder
+### Event recording
 
-The lifecycle event recorder persists procedure lifecycle events in the `greptime_private.events` system table. Configure it in `standalone` mode or on Metasrv in a distributed deployment:
+Recorded events are stored in the `greptime_private.events` system table.
 
 ```toml
 [event_recorder]

versioned_docs/version-1.2/user-guide/deployments-administration/monitoring/events/overview.md

@@ -14,7 +14,7 @@ acknowledgement.
 
 ## Configure event recording
 
-The event recorder can be configured in standalone deployments or on Metasrv in distributed deployments. See [Lifecycle event recorder](/user-guide/deployments-administration/configuration.md#lifecycle-event-recorder) for configuration options and the supported type list.
+See [Event recording](/user-guide/deployments-administration/configuration.md#event-recording) for configuration options and the event types supported by standalone and Metasrv.
 
 Standalone deployments record supported local DDL Procedure events. Distributed deployments with Metasrv can additionally record operational event types.