2748460 - Restored db state differs between offset and timestamp in PITR restore -- SAP IQ

IQ 16.0 SP 11 PL20 on Linux (Behaviour is the same in 16.1 as well).

During testing of PITR backup/restore process when iq_log_backup_interval has been set, restoring a PITR log backup based on offset and timestamp show different snapshots being restored. for test we are updating a table every min.

When using " recover until timestamp '2018-08-31 17:35:23.123' ",

we get these rows:
>>
...
        83 2018-08-31 17:33:00.001              1442320
        84 2018-08-31 17:34:00.000              1443664
        85 2018-08-31 17:35:00.001              1444176
        86 2018-08-31 17:36:00.001              1444688
        87 2018-08-31 17:37:00.001              1445200
        88 2018-08-31 17:38:00.000              1445779
        89 2018-08-31 17:39:00.001              1446315
        90 2018-08-31 17:40:00.001              1446827
        91 2018-08-31 17:41:00.001              1447339
<<

When using " recover until offset 1444700 ", we get those (as expected)

>>

...
        83 2018-08-31 17:33:00.001              1442320
        84 2018-08-31 17:34:00.000              1443664
        85 2018-08-31 17:35:00.001              1444176
<<

So we actually get 6 rows beyond the expectation (17:36 - 17:41, including)

• SAP IQ16.0
• SAP IQ16.1

PITR, "Point In Time Recovery" , "Point-In-Time-Recovery" , , KBA , BC-SYB-IQ , Sybase IQ , Problem