Sarahjohn,
By default, Delphix uses the BACKUP INCREMENTAL FROM SCN syntax, which does not interfere with other incremental backups using either the BACKUP INCREMENTAL LEVEL 0|1 syntax or other jobs running the FROM SCN syntax, but the error you specified pertains to bug 5852050 (ORA-600 [krcrfi_nohist] from RMAN in RAC environment with Block Change Tracking -
Oracle Support Doc ID 5852050.8).
Block Change Tracking (BCT) is highly recommended when incremental backups are taking place, as a separate database file is allocated simply to maintain a list to keep track of which blocks have been changed, which relieves an RMAN incremental backup of having to scan entire datafiles to find changed blocks. BCT is recommended for use with Delphix.
The bug in question is due to an Oracle RAC environment including one or more datafiles which do not reside on shared clustered storage, thus corrupting the BCT list. Here is an excerpt from the Support article referenced above...
Enabling block change tracking in a RAC environment for which a datafile not on a shared disk results in ORA-00600 [krcrfi_nohist] when using the RMAN BACKUP INCREMENTAL is commanded. Rediscovery Notes: 1) Target is a RAC database 2) Database BLOCK CHANGE TRACKING enabled 3) RMAN BACKUP INCREMENTAL fails with ORA-600 [krcrfi_nohist] 4) The storage location for this datafile is not on a shared device then this bug has been encountered. Workaround: Disable block change tracking or move the offending datafile to a shared disk.
So, the ORA-00600 [krcrfi_nohist] error you cited is not caused by concurrency of two or more incremental backups running simultaneously, but rather from a misconfiguration of the RAC environment by the database administrator (i.e. datafile on non-shared storage).
In addition to the workarounds cited in the bug text, I'll add that the bug is confirmed in Oracle versions lower than v11.1, particularly in v10.2.0.2 of Oracle database. These are very old versions, which are also not terminal releases for the very old versions, so the most effective workaround would be to upgrade to the terminal release for the current version (i.e. 10.2.0.5.x for v10.2, 11.1.0.7.x for v11.1, 11.2.0.4.x for v11.2, etc).
So, please upgrade the database software, and correct the misconfigured datafiles to move them to shared cluster storage, and this particular problem will be resolved.
Regarding your original question (i.e. "Can RMAN run simultaeously two incremental backups on the same database?"), it is possible if you're using the BACKUP INCREMENTAL FROM SCN syntax as Delphix does by default. This is one major reason why Delphix uses it, in order to not interfere with "normal backups" by database administrators.
Please let me know if this helps?
Thanks!
-Tim