'Kbytes Read' column in 'Show capture detail' output has abnormal values. (4329730)

'Kbytes Read' column in 'show capture detail' shows abnormal values.

For example,

sp_ctrl> show capture detail

---
Log reader threads:

Thread Instance Host Oracle Log Redo Log Log Offset Kbytes Read
------ ------------ ------------ ---------- -------- ------------ ------------
1 inst1 host1 157406 157304 3751952 1056636240
2 inst2 host2 175326 175163 3826184 5210556136
3 inst3 host3 132655 132564 3804520 1407648151
4 inst4 host4 195223 195041 97226296 8674385375 <===

Log reader threads:

Thread Instance Host Oracle Log Redo Log Log Offset Kbytes Read
------ ------------ ------------ ---------- -------- ------------ ------------
1 inst1 host1 157406 157304 11139440 1056643518
2 inst2 host2 175326 175163 7431312 5210559658
3 inst3 host3 132655 132564 7101408 1407651388
4 inst4 host4 195223 195042 77541088 8698352288 <===
---

Kbytes Read value for inst4 instance is incorrect.

Difference in each value is as follows.

(8698352288 - 8674385375)/1024/1024 = 22.86GB

Size of each Redo log is 1GB and Redo logs processed for inst4 between two outputs above is from 195041 until 195042.
These values are incorrect as they show capture had processed over 22GB logs whereas capture had only processed one redo log.

Miscalculation is caused by Oracle 10g log parallelism.

WORKAROUND:

Check the 'Offset' column value(in Bytes) instead of the 'Kbytes Read' column.

STATUS:

Waiting for fix in a future release of SharePlex For Oracle.

The latest version of SharePlex for Oracle can be downloaded at: 

https://support.quest.com/es-es/shareplex/download-new-releases