github-actions[bot] commented on code in PR #65050:
URL: https://github.com/apache/doris/pull/65050#discussion_r3499666914
##########
regression-test/suites/query_p0/system/test_query_sys_rowsets.groovy:
##########
@@ -28,6 +28,11 @@ suite("test_query_sys_rowsets", "query,p0") {
def rowsets_table_name = """ test_query_sys_rowsets.test_query_rowset """
sql """ drop table if exists ${rowsets_table_name} """
+ SimpleDateFormat sdf = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss")
+ // Use a lower bound before table creation to keep the timestamp filter
stable
+ // across second-level boundary races.
+ def rowsetFilterStartTime = sdf.format(new Date(System.currentTimeMillis()
- 60000L)).toString();
Review Comment:
Can we apply the same timestamp-boundary stabilization to the sibling
cloud-only rowsets scan test as part of this fix?
`test_query_sys_scan_rowsets.groovy` still creates the table, captures `now`
immediately afterwards, and later filters with `NEWEST_WRITE_TIMESTAMP >=
'${now}'` while `test_query_sys_scan_rowsets.out` expects `rowsets4` to exclude
version `0-1`. Since `NEWEST_WRITE_TIMESTAMP` is exposed from Unix-second
rowset metadata, if the create rowset and captured `now` land in the same
second the `>=` predicate can include `0-1`, leaving that sibling regression
flaky for the same boundary reason this patch fixes here.
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
