feat: Metadata: 自定义指标查询路由过期时间根据关联集群获取 --story=128196686 #8575
Conversation
unique0lai
added
feat
There was a problem hiding this comment.
Pull Request Overview
This PR modifies the metadata query routing to use cluster-specific retention times instead of a global default. The change allows custom metric expiration times to be determined based on the associated VM cluster's configuration.
Key Changes:
- Introduces cluster-specific retention time logic for filtering expired time series metrics
- Refactors query optimization by replacing paginated filtering with direct
filter()calls - Adds type annotations to improve type safety
Reviewed Changes
Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.
| File | Description |
|---|---|
bkmonitor/metadata/models/storage.py |
Added type annotation for default_settings field in ClusterInfo model |
bkmonitor/metadata/models/space/space_table_id_redis.py |
Refactored _filter_ts_info to group metrics by cluster retention time; simplified query filtering logic |
bkmonitor/metadata/models/space/space.py |
Added type annotation for dimension_values field in SpaceResource model |
bkmonitor/metadata/models/space/constants.py |
Added DEFAULT_VM_RETENTION_TIME constant (60 days in seconds) |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| BKCI_SYSTEM_TABLE_ID_PREFIX = "system." | ||
|
|
||
| # VM 集群默认过期时间, 单位为秒, 默认60天 | ||
| DEFAULT_VM_RETENTION_TIME = 60 * 3600 * 24 |
There was a problem hiding this comment.
[nitpick] The calculation is incorrect. The comment states the default is 60 days, but the calculation 60 * 3600 * 24 equals 5,184,000 seconds which is 60 days. However, the comment says 'unit is seconds, default 60 days' which is ambiguous. If 60 represents days, the formula should be 60 * 24 * 3600. The current order is mathematically correct but unconventional. Consider reordering to 60 * 24 * 3600 for clarity (days * hours_per_day * seconds_per_hour).
| DEFAULT_VM_RETENTION_TIME = 60 * 3600 * 24 | |
| DEFAULT_VM_RETENTION_TIME = 60 * 24 * 3600 |
| cluster_type=models.ClusterInfo.TYPE_VM, | ||
| ).only("cluster_id", "default_settings") | ||
| vm_cluster_id_retention_time_map: dict[int, int] = { | ||
| data.cluster_id: data.default_settings.get("retention_time") or DEFAULT_VM_RETENTION_TIME |
There was a problem hiding this comment.
Using or operator can incorrectly fall back to the default if retention_time is 0 or other falsy values. If a cluster intentionally sets retention_time to 0, this code will ignore that setting. Use if data.default_settings.get('retention_time') is not None else DEFAULT_VM_RETENTION_TIME instead.
| data.cluster_id: data.default_settings.get("retention_time") or DEFAULT_VM_RETENTION_TIME | |
| data.cluster_id: data.default_settings.get("retention_time") if data.default_settings.get("retention_time") is not None else DEFAULT_VM_RETENTION_TIME |
There was a problem hiding this comment.
Pull Request Overview
Copilot reviewed 4 out of 4 changed files in this pull request and generated 1 comment.
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| for data in ts_group_fields: | ||
| group_id_field_map.setdefault(data["group_id"], set()).add(data["field_name"]) | ||
| for retention_days, group_id_list in retention_time_group_ids_map.items(): | ||
| # 计算该分组的 begin_time，使用集群的 retention_time（天数转换为秒数） |
There was a problem hiding this comment.
The comment states '天数转换为秒数' (converting days to seconds), but the code uses datetime.timedelta(days=retention_days) which takes days directly, not seconds. The comment should be updated to '使用集群的 retention_time(以天为单位)' (using cluster's retention_time in days).
| # 计算该分组的 begin_time，使用集群的 retention_time（天数转换为秒数） | |
| # 计算该分组的 begin_time，使用集群的 retention_time(以天为单位) |
There was a problem hiding this comment.
Pull Request Overview
Copilot reviewed 4 out of 4 changed files in this pull request and generated no new comments.
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
1 participant
No description provided.