### Problem In https://gitlab.com/gitlab-com/ops-sub-department/section-ops-request-for-help/-/issues/264 and https://gitlab.com/gitlab-com/ops-sub-department/section-ops-request-for-help/-/issues/293 we are seeing errors due to the IID tracking (handled by `InternalId`) not reflecting the latest CI Pipeline IID after a project is imported, even though the IID tracking should be flushed after import https://gitlab.com/gitlab-com/ops-sub-department/section-ops-request-for-help/-/issues/264#note_1718285528. ### Workaround A workaround is to reset the IID tracking for CI Pipelines for the project https://gitlab.com/gitlab-com/ops-sub-department/section-ops-request-for-help/-/issues/264#note_1730359496. --- To verify if you're affected, check your logs for an error like this: ``` PG::UniqueViolation: ERROR: duplicate key value violates unique constraint "index_ci_pipelines_on_project_id_and_iid" DETAIL: Key (project_id, iid)=(52897123, 756) already exists. ``` The project ID of your affected project would be `52897123` in this example. To resolve the issue, [connect to the PostgreSQL database](https://docs.gitlab.com/omnibus/settings/database.html#connecting-to-the-postgresql-database) and run the following queries: ```sql SELECT * FROM internal_ids WHERE project_id = 52897123 and usage = 5; SELECT COUNT(*) FROM ci_pipelines WHERE project_id = 52897123; ``` Make sure to replace `52897123` with your project ID. The value `5` for `usage` in the first query is a constant and [refers to CI pipelines](https://gitlab.com/gitlab-org/gitlab/-/blob/v16.10.0-ee/app/models/concerns/enums/internal_id.rb?ref_type=tags#L13). The `last_value` in the response of the first query will differ from the count in the second query response, 349 vs. 8217 in this example: ``` id | project_id | usage | last_value | namespace_id ----------+------------+-------+------------+-------------- 25674456 | 52897123 | 5 | 349 | (1 row) count ------- 8217 (1 row) ``` To resolve this, delete the row from the `internal_ids` table via its ID, `25674456` in this example: ```sql DELETE FROM internal_ids WHERE id = 25674456; ``` This ID will be recreated when next needed, deleting it poses no risk. --- Alternatively, you can use [a Rails console session](https://docs.gitlab.com/ee/administration/operations/rails_console.html#starting-a-rails-console-session) to achieve the same result: ``` InternalId.flush_records!(usage: :ci_pipelines, project_id: <project_id>) ```