docs(api-reference): daily api-review 2026-10-29 — drop license release_limited, add rum cumulative, monit absent_counts_as_normal

api-reference/monitors.openapi.en.json

@@ -9616,6 +9616,10 @@
           "recovery": {
             "$ref": "#/components/schemas/AlertRuleRecoveryQueryV2",
             "description": "Recovery evaluation config. Required (with non-empty `recovery.expr`) only when `recovery_mode` is `expression_match` or `recovery_query_match`; must be omitted for the other modes."
+          },
+          "absent_counts_as_normal": {
+            "type": "boolean",
+            "description": "Count an active instance that is missing from every condition query of a successful round as one normal round, instead of freezing the instance in its current state. Only valid with `recovery_mode` `condition_clear` or `expression_match`; omitting it keeps the freeze-on-absence behavior of existing rules. The matched datasources must be served by monit-edge v0.77.0 or later, otherwise saving the rule fails with `InvalidParameter`."
           }
         }
       },

api-reference/monitors.openapi.zh.json

@@ -9616,6 +9616,10 @@
           "recovery": {
             "$ref": "#/components/schemas/AlertRuleRecoveryQueryV2",
             "description": "恢复判定配置。仅 `recovery_mode` 为 `expression_match` 或 `recovery_query_match` 时必填（`recovery.expr` 非空），其余模式不允许传入。"
+          },
+          "absent_counts_as_normal": {
+            "type": "boolean",
+            "description": "把在某轮成功评估中未出现在任何条件查询里的活动实例计为一次正常轮次，而不是把它冻结在当前状态。仅 `recovery_mode` 为 `condition_clear` 或 `expression_match` 时可用；省略该字段时保持既有规则「数据缺失即冻结」的行为。命中的数据源须由 monit-edge v0.77.0 及以上版本服务，否则保存规则会以 `InvalidParameter` 失败。"
           }
         }
       },

api-reference/on-call.openapi.en.json

@@ -17694,8 +17694,7 @@
                         "created_at": 0,
                         "updated_at": 0
                       }
-                    ],
-                    "release_limited": false
+                    ]
                   }
                 }
               }
@@ -20596,8 +20595,7 @@
         "description": "People with active fixed or temporary On-call licenses.",
         "required": [
           "total",
-          "items",
-          "release_limited"
+          "items"
         ],
         "properties": {
           "total": {
@@ -20611,10 +20609,6 @@
             "items": {
               "$ref": "#/components/schemas/LicensePersonItem"
             }
-          },
-          "release_limited": {
-            "type": "boolean",
-            "description": "Whether the account has used up its license release quota for the current 30-day window. While `true`, releasing a seat fails with error code `NoLicense`."
           }
         }
       },

api-reference/on-call.openapi.zh.json

@@ -17694,8 +17694,7 @@
                         "created_at": 0,
                         "updated_at": 0
                       }
-                    ],
-                    "release_limited": false
+                    ]
                   }
                 }
               }
@@ -20596,8 +20595,7 @@
         "description": "当前账户中持有固定或临时 On-call 许可的人员。",
         "required": [
           "total",
-          "items",
-          "release_limited"
+          "items"
         ],
         "properties": {
           "total": {
@@ -20611,10 +20609,6 @@
             "items": {
               "$ref": "#/components/schemas/LicensePersonItem"
             }
-          },
-          "release_limited": {
-            "type": "boolean",
-            "description": "当前 30 天窗口内的许可释放配额是否已用尽。为 `true` 时释放席位会返回错误码 `NoLicense`。"
           }
         }
       },

api-reference/openapi.en.json

@@ -13798,10 +13798,6 @@
             },
             "type": "array"
           },
-          "release_limited": {
-            "description": "Whether the account has used up its license release quota for the current 30-day window. While `true`, releasing a seat fails with error code `NoLicense`.",
-            "type": "boolean"
-          },
           "total": {
             "description": "Number of people holding an active license.",
             "format": "int64",
@@ -13810,8 +13806,7 @@
         },
         "required": [
           "total",
-          "items",
-          "release_limited"
+          "items"
         ],
         "type": "object"
       },
@@ -20760,8 +20755,12 @@
       "RumDataQueryDefinition": {
         "description": "One RUM data query definition.",
         "properties": {
+          "cumulative": {
+            "description": "Set to `true` to get a time series of a single `COUNT` or `SUM` as the increments of its running total: each bucket holds only what the bucket adds, so the running sum ends at the aggregate over the range, aligned the way that aggregate is. Requires `format` `time_series` and exactly one select field, `COUNT(...)` or `SUM(...)`, with an alias; a statement with `SELECT DISTINCT`, `GROUP BY`, `HAVING` or `LIMIT` makes the whole request fail with `InvalidParameter`. `COUNT(DISTINCT x)` is accepted and counts each `x` in its first bucket only. The query always runs unsampled, so the response carries no `sampling`.",
+            "type": "boolean"
+          },
           "disable_sampling": {
-            "description": "When `true`, the query always runs on the full data and the response never carries `sampling`. When `false` (default), the server may sample aggregate queries over large time windows.",
+            "description": "When `true`, the query always runs on the full data and the response never carries `sampling`. When `false` (default), the server may sample aggregate queries over large time windows. A query that would have been sampled otherwise also takes one of the account's exact-query slots: it queues while too many such queries run for the account, and if no slot frees up before its deadline that query's `error` entry is set with code `BadRequest`.",
             "type": "boolean"
           },
           "dql": {
@@ -30335,6 +30334,10 @@
           "recovery": {
             "$ref": "#/components/schemas/AlertRuleRecoveryQueryV2",
             "description": "Recovery evaluation config. Required (with non-empty `recovery.expr`) only when `recovery_mode` is `expression_match` or `recovery_query_match`; must be omitted for the other modes."
+          },
+          "absent_counts_as_normal": {
+            "type": "boolean",
+            "description": "Count an active instance that is missing from every condition query of a successful round as one normal round, instead of freezing the instance in its current state. Only valid with `recovery_mode` `condition_clear` or `expression_match`; omitting it keeps the freeze-on-absence behavior of existing rules. The matched datasources must be served by monit-edge v0.77.0 or later, otherwise saving the rule fails with `InvalidParameter`."
           }
         }
       },
@@ -53756,8 +53759,7 @@
                         "created_at": 0,
                         "updated_at": 0
                       }
-                    ],
-                    "release_limited": false
+                    ]
                   }
                 }
               }

api-reference/openapi.zh.json

@@ -13798,10 +13798,6 @@
             },
             "type": "array"
           },
-          "release_limited": {
-            "description": "当前 30 天窗口内的许可释放配额是否已用尽。为 `true` 时释放席位会返回错误码 `NoLicense`。",
-            "type": "boolean"
-          },
           "total": {
             "description": "持有有效许可的人员数量。",
             "format": "int64",
@@ -13810,8 +13806,7 @@
         },
         "required": [
           "total",
-          "items",
-          "release_limited"
+          "items"
         ],
         "type": "object"
       },
@@ -20760,8 +20755,12 @@
       "RumDataQueryDefinition": {
         "description": "单个 RUM 数据查询定义。",
         "properties": {
+          "cumulative": {
+            "description": "设为 `true` 时，把单个 `COUNT` 或 `SUM` 的时间序列按累计值的增量返回：每个时间桶只填该桶新增的部分，累计和正好等于整个区间的聚合值，且与聚合值一样对齐。要求 `format` 为 `time_series`，并且只有一个带别名的 `COUNT(...)` 或 `SUM(...)` 字段；语句含 `SELECT DISTINCT`、`GROUP BY`、`HAVING` 或 `LIMIT` 时整个请求以 `InvalidParameter` 失败。支持 `COUNT(DISTINCT x)`，该值只在首次出现的桶里计一次。该查询始终不采样，因此响应不带 `sampling`。",
+            "type": "boolean"
+          },
           "disable_sampling": {
-            "description": "为 `true` 时始终查询全量数据，响应不会出现 `sampling`。为 `false`（默认）时，服务端可能对大时间跨度的聚合查询进行抽样。",
+            "description": "为 `true` 时始终查询全量数据，响应不会出现 `sampling`。为 `false`（默认）时，服务端可能对大时间跨度的聚合查询进行抽样。本会被抽样的查询还会占用该账户的一个 exact 查询槽位：同时运行的这类查询过多时它会排队，若在超时前没有空出槽位，该查询的 `error` 条目会以 `BadRequest` 返回。",
             "type": "boolean"
           },
           "dql": {
@@ -30335,6 +30334,10 @@
           "recovery": {
             "$ref": "#/components/schemas/AlertRuleRecoveryQueryV2",
             "description": "恢复判定配置。仅 `recovery_mode` 为 `expression_match` 或 `recovery_query_match` 时必填（`recovery.expr` 非空），其余模式不允许传入。"
+          },
+          "absent_counts_as_normal": {
+            "type": "boolean",
+            "description": "把在某轮成功评估中未出现在任何条件查询里的活动实例计为一次正常轮次，而不是把它冻结在当前状态。仅 `recovery_mode` 为 `condition_clear` 或 `expression_match` 时可用；省略该字段时保持既有规则「数据缺失即冻结」的行为。命中的数据源须由 monit-edge v0.77.0 及以上版本服务，否则保存规则会以 `InvalidParameter` 失败。"
           }
         }
       },
@@ -53756,8 +53759,7 @@
                         "created_at": 0,
                         "updated_at": 0
                       }
-                    ],
-                    "release_limited": false
+                    ]
                   }
                 }
               }

api-reference/rum.openapi.en.json

@@ -1223,8 +1223,12 @@
       "RumDataQueryDefinition": {
         "description": "One RUM data query definition.",
         "properties": {
+          "cumulative": {
+            "description": "Set to `true` to get a time series of a single `COUNT` or `SUM` as the increments of its running total: each bucket holds only what the bucket adds, so the running sum ends at the aggregate over the range, aligned the way that aggregate is. Requires `format` `time_series` and exactly one select field, `COUNT(...)` or `SUM(...)`, with an alias; a statement with `SELECT DISTINCT`, `GROUP BY`, `HAVING` or `LIMIT` makes the whole request fail with `InvalidParameter`. `COUNT(DISTINCT x)` is accepted and counts each `x` in its first bucket only. The query always runs unsampled, so the response carries no `sampling`.",
+            "type": "boolean"
+          },
           "disable_sampling": {
-            "description": "When `true`, the query always runs on the full data and the response never carries `sampling`. When `false` (default), the server may sample aggregate queries over large time windows.",
+            "description": "When `true`, the query always runs on the full data and the response never carries `sampling`. When `false` (default), the server may sample aggregate queries over large time windows. A query that would have been sampled otherwise also takes one of the account's exact-query slots: it queues while too many such queries run for the account, and if no slot frees up before its deadline that query's `error` entry is set with code `BadRequest`.",
             "type": "boolean"
           },
           "dql": {

api-reference/rum.openapi.zh.json

@@ -1223,8 +1223,12 @@
       "RumDataQueryDefinition": {
         "description": "单个 RUM 数据查询定义。",
         "properties": {
+          "cumulative": {
+            "description": "设为 `true` 时，把单个 `COUNT` 或 `SUM` 的时间序列按累计值的增量返回：每个时间桶只填该桶新增的部分，累计和正好等于整个区间的聚合值，且与聚合值一样对齐。要求 `format` 为 `time_series`，并且只有一个带别名的 `COUNT(...)` 或 `SUM(...)` 字段；语句含 `SELECT DISTINCT`、`GROUP BY`、`HAVING` 或 `LIMIT` 时整个请求以 `InvalidParameter` 失败。支持 `COUNT(DISTINCT x)`，该值只在首次出现的桶里计一次。该查询始终不采样，因此响应不带 `sampling`。",
+            "type": "boolean"
+          },
           "disable_sampling": {
-            "description": "为 `true` 时始终查询全量数据，响应不会出现 `sampling`。为 `false`（默认）时，服务端可能对大时间跨度的聚合查询进行抽样。",
+            "description": "为 `true` 时始终查询全量数据，响应不会出现 `sampling`。为 `false`（默认）时，服务端可能对大时间跨度的聚合查询进行抽样。本会被抽样的查询还会占用该账户的一个 exact 查询槽位：同时运行的这类查询过多时它会排队，若在超时前没有空出槽位，该查询的 `error` 条目会以 `BadRequest` 返回。",
             "type": "boolean"
           },
           "dql": {