qlik replicate task is terminating with fatal error repository.c:5755

[paccar_sk]

Hi, when SQL server host if recycled replicate task should keep in trying to ping the host at a sent interval (defined in the configuration) this feature is not behaving consistently for the same data source, sometime  Qlik Replicate will restart the task without problem, sometime it will terminate with Fatal error (repository.c:5755) , Wondering what could be the reason and how can we diagnose beyond log file?

[lyka]

Hello,

If the error is an environmental error or a recoverable error, then replicate will retry. If its a fatal error, then replicate will not try to resume. What is the error when the task failed?

Thanks

Lyka

[paccar_sk]

Error Code is c 5755..

00021708: 2021-02-04T16:50:29 [METADATA_MANAGE ]E: RetCode: SQL_ERROR SqlState: 28000 NativeError: 18456 Message: [Microsoft][SQL Server Native Client 11.0][SQL Server]Login failed for user '********'. [1022502] (ar_odbc_conn.c:563)
00021708: 2021-02-04T16:50:29 [METADATA_MANAGE ]E: Failed to connect [1022515] (sqlserver_endpoint_imp.c:2077)
00021708: 2021-02-04T16:50:29 [METADATA_MANAGE ]E: Failed while preparing stream component '********. [1022515] (metadatamanager.c:846)
00021708: 2021-02-04T16:50:29 [METADATA_MANAGE ]E: Cannot create the source utility component [1022515] (metadatamanager.c:744)
00021708: 2021-02-04T16:50:29 [TASK_MANAGER ]E: Creating Metadata Manager's utility components failed [1022515] (replicationtask.c:3595)
00021708: 2021-02-04T16:50:29 [TASK_MANAGER ]W: Task '********' encountered a fatal error (repository.c:5755)

[lyka · Accepted solution]

Hello,

You can try this:

1) Add the following internal parameter to make the the 28000 error recoverable:  $info.query_syntax.error_code_connection_failure  and set the value to   00026,00904,10054,4060,01222,55,64,121,1231,233,109,18401,6005,40613,1205,10060,28000

2) Change the task Error Handling-->Environment Errors and set "Interval between retry attempts (seconds):" to 30. This will give the server a little longer before we retry the connection.

Thanks,

Lyka