Selected timeline reconstructed from the supplied September 8 debug log.
This is not the complete debug.log file.

Connection ID: 6180152

17:33:09
Start SQL statement:
insert into o_reha_ord_tmp select * from o_reha_ord order by ord_date;

17:34:36
End SQL statement for the previous INSERT, followed by:
insert into o_rtx_ord_tmp select * from o_rtx_ord order by ord_date;

17:35:18 - 17:35:39
TRUNCATE TABLE statements executed on the same connection:
- truncate table o_reha_ord_tmp;
- truncate table o_rtx_ord_tmp;
- truncate table o_consult_ord_tmp;
- truncate table o_drug_ord_tmp;

17:35:56
Start SQL statement:
insert into o_reha_ord_tmp select * from o_reha_ord_back order by ord_date;

17:36:01 - 17:36:07
cpimport bulk loads started for o_consult_ord_tmp and o_reha_ord_tmp.

17:36:08
writeenginesplit messages:
- Send EOD message to All PMs
- Received signal to terminate the process.

17:36:09
cpimport results:
- JobId 4538, o_consult_ord_tmp, 0 rows inserted, status SUCCESS
- JobId 5083, o_reha_ord_tmp, 0 rows inserted, status SUCCESS

ColumnStore then logged OnReceiveRollback and started bulk rollback for both
tables, restoring extents, extent maps, HWM values, and database files.

17:36:17 - 17:36:25
TRUNCATE TABLE statements executed again using connection ID 6180152:
- truncate table o_consult_ord_tmp;
- truncate table o_reha_ord_tmp;

17:36:34
The INSERT into o_reha_ord_tmp from o_reha_ord_back was started again using
the same connection.

Crash signature observed for the incident:
ha_mcs_impl_write_batch_row_(... cal_connection_info ...)+0x163
SIGSEGV / signal 11

Additional observation:
The September 8 operation window contained more than one active connection.
The complete server error log should be used to correlate the exact crashing
thread with this timeline.