Routine load from Kafka issue with Unix timestamp

[dorsegal]

Steps to reproduce the behavior (Required)

Follow this - https://docs.starrocks.io/docs/loading/RoutineLoad/#submit-a-routine-load-job-1

I am doing the same but my timestamp is 1715553720000 which throws error and from logs the value is NULL.
If I do 1715553720 it works as expected.

StarRocks version (Required)

3.3.5

[dorsegal]

I have managed to do a workaroound.

COLUMNS(
temp_timestamp,
timestamp = from_unixtime(cast(temp_timestamp as bigint)/1000)
)

But now I am facing something else I see tasks submitted and messages are being consumed but only for several minutes. after then that it stopped.

TaskId                              |TxnId |TxnStatus|JobId |CreateTime         |LastScheduledTime  |ExecuteStartTime   |Timeout|BeId  |DataSourceProperties                                                                                                                                                                                                                |Message                  |Warehouse        |
------------------------------------+------+---------+------+-------------------+-------------------+-------------------+-------+------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+-------------------------+-----------------+
49fefa24-18c6-411a-a0fd-7d078dfa549e|157835|COMMITTED|373030|2024-10-30 17:27:45|2024-10-30 17:27:55|2024-10-30 17:27:55|60     |373236|Progress:{"0":263060,"33":191049,"3":137429,"36":216789,"6":143667,"39":151216,"9":123804,"42":213118,"12":123741,"45":174647,"15":147513,"48":153462,"18":103136,"21":133956,"24":151699,"27":124039,"30":254126},LatestOffset:null|task submitted to execute|default_warehouse|
bbda35ff-77ef-4984-ad0c-6d4c18a6c598|157836|COMMITTED|373030|2024-10-30 17:27:45|2024-10-30 17:27:55|2024-10-30 17:27:55|60     |371107|Progress:{"1":227331,"34":138045,"4":110393,"37":89110,"7":145714,"40":164551,"10":87523,"43":233408,"13":113599,"46":143718,"16":148728,"49":116000,"19":96077,"22":103162,"25":144535,"28":61957,"31":172068},LatestOffset:null   |task submitted to execute|default_warehouse|
6ecccbcb-b192-443e-beee-62460c59e985|157837|COMMITTED|373030|2024-10-30 17:27:45|2024-10-30 17:27:55|2024-10-30 17:27:55|60     |372851|Progress:{"32":216134,"2":309174,"35":388152,"5":226989,"38":171677,"8":219902,"41":247702,"11":235816,"44":209313,"14":223932,"47":164389,"17":242534,"20":164987,"23":324165,"26":213792,"29":185581},LatestOffset:null           |task submitted to execute|default_warehouse|

did more digging and saw many of those
WARN (leaderCheckpointer|27260) [RoutineLoadJob.unprotectUpdateState():1321] routine load job 373030-data_from_kafka changed to PAUSED with reason: null

Also Kafka consumer group is empty