Frequently facing Logstream truncate issue

[suvbin]

Frequently facing the "SAPHANA log truncated" error.  May i know please any steps i can perform, inorder to avoid thiese kind of errors.

Thanks in advance.

[lyka]

Good Day!

If you are seeing a message like:

00008208: 2020-09-20T23:54:10 [SOURCE_CAPTURE  ]E:  Task cannot resume. Change Log was truncated, last read index=679946252, current start index=680286114 [1024208]  (saphana_endpoint_capture_trigger_based.c:1937)

this is related to the Log retention period (min):

We also have seen this in similar cases when using the same CDC changes table one task may delete rows that are not yet processed by the second task.

This will also occur if one task was stopped and the other is still running, there is no relation to stopping the DB or SAP components.

If the two Tasks do not share the same table set, it is recommended to use a different source endpoint that uses different CDC changes tables, (see the below work around)

The workaround for now is:

- Using two different SAP Hana Endpoints for each Task.

- Specify a different schema for the CDC table in each endpoint.

This requires the SAP HANA user to have permissions to to write to the second schema.

If you would like to troubleshoot this further, please open a support ticket.

Thanks

Lyka

[suvbin]

Thank you for the reply. Our approach is like , we have one logstream task(SAP) and two target tasks for it , one goes to Azure managed instance(full load ON, apply changes is ON) and the other goes to Azure synapse (full load ON,apply changes ON,store changes ON) . Both the targets have different schemas. If the logstream task fails, remaining two will also fail , i need to do reload target for all the tasks.

Any suggestions here please ?