yuxiqian commented on code in PR #4444:
URL: https://github.com/apache/flink-cdc/pull/4444#discussion_r3449773133
##########
flink-cdc-connect/flink-cdc-source-connectors/flink-connector-sqlserver-cdc/src/test/java/org/apache/flink/cdc/connectors/sqlserver/table/SqlServerConnectorITCase.java:
##########
@@ -358,6 +360,18 @@ void testStartupFromTimestampWithServerTimeZone() throws
Exception {
@ParameterizedTest
@ValueSource(booleans = {true, false})
void testAllTypes(boolean parallelismSnapshot) throws Throwable {
+ // CI runs tests under a randomized -Duser.timezone (see
.github/workflows/utils.sh). On
+ // sub-hour offsets (e.g. GMT+11:30) Debezium 1.9.8 underflows
NanoOfSecond while converting
+ // the streamed TIME columns of full_types, which drops the +U record
and stalls the sink
+ // until the job-level timeout. The JVM default cannot be re-pinned
here because the
+ // SQL Server JDBC driver resolves the zone outside this thread's
TimeZone.getDefault(), so
+ // skip this type-coverage case on sub-hour offsets; timezone
behaviour is covered by
+ // SqlServerTimezoneITCase.
+ Assumptions.assumeThat(
+ TimeZone.getDefault().getRawOffset() % (int)
TimeUnit.HOURS.toMillis(1))
+ .as(
+ "Skipping on sub-hour timezone offset due to Debezium
TIME conversion limitation")
+ .isZero();
Review Comment:
It's suspicious to skip the entire test case. Shall we consider modifying
the `.github/workflows/utils.sh` and always generate time zones on the hour?
--
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]
