Kris20030907 opened a new pull request, #7050:
URL: https://github.com/apache/rocketmq-dashboard/pull/7050
Closes #7049
## Solution
`clearRuntimeSessionId` now stamps `gmt_modified` explicitly from the
caller's
clock — the same convention the alert mappers use (`@Param("now")
LocalDateTime
now`): repositories take the time, they do not read it themselves.
```java
conversationMapper.update(null, new UpdateWrapper<RmqAiConversation>()
.eq("id", id)
.set("runtime_session_id", null)
.set("gmt_modified", now));
```
`AiRunExecutor.forgetRuntimeSession` passes `LocalDateTime.now(clock)`, the
same
service clock `rememberRuntimeSession` stamps with. The DDL's
`ON UPDATE CURRENT_TIMESTAMP` never gets a chance to rewrite the column in
the
database session's timezone, so the list ordering
(`gmt_modified desc, id`) sees one clock again.
A schema-level alternative — dropping `ON UPDATE CURRENT_TIMESTAMP` — was
considered and rejected: it needs a migration for existing installs, and
every
other write site already stamps the column explicitly.
## Tests
-
`AiTimelineRepositoryTest.forgettingTheRuntimeSessionShouldStampGmtModifiedFromTheCallerClockTest`:
the captured `UpdateWrapper` must carry `gmt_modified` bound to the
caller's
timestamp. Red before the fix.
- The existing forget-session tests keep asserting the clear itself
(`runtime_session_id` named and bound to null); `AiRunExecutorTest`'s two
verifications follow the new signature.
## Verification
Full server suite: 3736 tests pass. Frontend suites: not affected
(server-only
change).
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
