harbor删除保留策略遇到的问题
harbor版本：v2.10
所做操作
在harbor api控制台做了如下操作，本意为针对项目id为66的项目配置镜像删除策略为保留最近拉取的3个
实际 curl 命令如下
curl -X 'PUT' \
'https://<url>/api/v2.0/retentions/1' \
-H 'accept: application/json' \
-H 'authorization: Basic xxx' \
-H 'Content-Type: application/json' \
-H 'X-Harbor-CSRF-Token: xxx' \
-d '{
"id": 0,
"algorithm": "or",
"rules": [
{
"id": 0,
"priority": 1,
"disabled": false,
"action": "retain",
"tag_selectors": [
{
"kind": "latestPulled",
"decoration": "pulled",
"pattern": "**",
"extras": "{\"count\": \"3\"}"
}
],
"scope_selectors": {
"repository": [
{
"kind": "doublestar",
"pattern": "**"
}
]
}
}
],
"trigger": {
"kind": "Schedule",
"settings": {
"cron": "0 0 0 * * *"
},
"references": {}
},
"scope": {
"level": "project",
"ref": 66
}
}'
返回如下
执行完成后，在harbor控制台查看，发现一直在转圈，误以为是有问题，实际上是因为项目数量很多，有900多个
误操作
误以为配置有问题，想要删除，执行了如下操作
实际curl命令如下
curl -X 'DELETE' \
'https://<url>/api/v2.0/retentions/1' \
-H 'accept: application/json' \
-H 'authorization: Basic xxx' \
-H 'X-Harbor-CSRF-Token: xxx'
返回如下
报错
执行完删除后在项目的 策略 选项卡中就报错 internal server error
查看 harbor-core 容器日志报错如下
$ docker logs -f harbor-core
......
2024-12-04T08:04:56Z [ERROR] [/lib/http/error.go:57]: {"errors":[{"code":"UNKNOWN","message":"unknown: no such Retention policy with id 1"}]}
2024-12-04T08:04:57Z [ERROR] [/lib/http/error.go:57]: {"errors":[{"code":"UNKNOWN","message":"unknown: no such Retention policy with id 1"}]}
2024-12-04T08:04:57Z [ERROR] [/lib/http/error.go:57]: {"errors":[{"code":"UNKNOWN","message":"unknown: no such Retention policy with id 1"}]}
......
解决方法
进入pg数据库
$ docker exec -it harbor-db psql -U postgres
psql (15.8)
Type "help" for help.
No entry for terminal type "xterm";
using dumb terminal settings.
切换到 registry 数据库
postgres=# \c registry
You are now connected to database "registry" as user "postgres".
查看表
registry=# \dt
List of relations
Schema | Name | Type | Owner
--------+-----------------------------+-------+----------
public | access | table | postgres
public | alembic_version | table | postgres
public | artifact | table | postgres
public | artifact_accessory | table | postgres
public | artifact_blob | table | postgres
public | artifact_reference | table | postgres
public | artifact_trash | table | postgres
public | audit_log | table | postgres
public | blob | table | postgres
public | cve_allowlist | table | postgres
public | data_migrations | table | postgres
public | execution | table | postgres
public | harbor_label | table | postgres
public | harbor_user | table | postgres
public | immutable_tag_rule | table | postgres
public | job_log | table | postgres
public | job_queue_status | table | postgres
public | label_reference | table | postgres
public | notification_policy | table | postgres
public | oidc_user | table | postgres
public | p2p_preheat_instance | table | postgres
public | p2p_preheat_policy | table | postgres
public | permission_policy | table | postgres
public | project | table | postgres
public | project_blob | table | postgres
public | project_member | table | postgres
public | project_metadata | table | postgres
public | properties | table | postgres
public | quota | table | postgres
public | quota_usage | table | postgres
public | registry | table | postgres
public | replication_policy | table | postgres
public | report_vulnerability_record | table | postgres
public | repository | table | postgres
public | retention_policy | table | postgres
public | robot | table | postgres
public | role | table | postgres
public | role_permission | table | postgres
public | sbom_report | table | postgres
public | scan_report | table | postgres
public | scanner_registration | table | postgres
public | schedule | table | postgres
public | schema_migrations | table | postgres
public | system_artifact | table | postgres
public | tag | table | postgres
public | task | table | postgres
public | user_group | table | postgres
public | vulnerability_record | table | postgres
(48 rows)
retention_policy 表是用于存储保留策略的定义，项目与保留策略的绑定关系存在于 project_metadata 表中
查看 project_metadata 表并指定有问题的项目id
registry=# SELECT * FROM project_metadata WHERE project_id = 66;
id | project_id | name | value | creation_time | update_time
----+------------+--------------+-------+----------------------------+----------------------------
48 | 48 | public | false | 2024-10-12 06:56:59.045553 | 2024-10-12 06:56:59.045553
98 | 48 | retention_id | 1 | 2024-12-04 04:17:51.363924 | 2024-12-04 04:17:51.363924
(2 rows)
查看 project_metadata 表并指定无问题的项目id
registry=# SELECT * FROM project_metadata WHERE project_id = 1;
id | project_id | name | value | creation_time | update_time
----+------------+--------+-------+----------------------------+----------------------------
1 | 1 | public | true | 2024-09-27 09:48:08.122988 | 2024-09-27 09:48:08.122988
(1 row)
对比后发现有问题的项目id条目中多了一条记录
98 | 66 | retention_id | 1 | 2024-12-04 04:17:51.363924 | 2024-12-04 04:17:51.363924
删除这条记录
registry=# DELETE FROM project_metadata WHERE project_id = 66 AND name = 'retention_id';
DELETE 1
再次查看，多余的记录已经被删除
registry=# SELECT * FROM project_metadata WHERE project_id = 66;
id | project_id | name | value | creation_time | update_time
----+------------+--------+-------+----------------------------+----------------------------
48 | 48 | public | false | 2024-10-12 06:56:59.045553 | 2024-10-12 06:56:59.045553
(1 row)
返回harbor控制台刷新即可
