HPE 3PAR StoreServ Storage

RCIP DR Test Oracle Recovery Time

Frequent Advisor

RCIP DR Test Oracle Recovery Time

We did an RCIP-based DR Test recently.  When we recovered the Oracle DB on a server connected to the target array, it "lost" 17 minutes from the RCIP "stop" time.


Look at the "showrcopy":


SLU-HP-MES-FC-01 cli% showrcopy groups SLU_P800

Remote Copy System Information
Status: Started, Normal

Group Information

Name         Target              Status   Role       Mode     Options
SLU_P800     SOASPW-3PAR-7400-2N Stopped  Primary    Periodic Period 10m, auto_recover,over_per_alert
  LocalVV                       ID   RemoteVV                      ID   SyncStatus    LastSyncTime
  SLULD1P-2P_SLU_P800_orabin      15 SLULD1P-2P_SLU_P800_orabin    1510 Stopped       2015-10-15 10:35:59 BRT
  SLULD1P-2P_SLU_P800_oraarch     16 SLULD1P-2P_SLU_P800_oraarch   1509 Stopped       2015-10-15 10:35:56 BRT
  SLULD1P-2P_SLU_P800_oradata01   17 SLULD1P-2P_SLU_P800_oradata01 1511 Stale         2015-10-15 10:30:19 BRT
  SLULD1P-2P_SLU_P800_oradata02   18 SLULD1P-2P_SLU_P800_oradata02 1514 Stale         2015-10-15 10:28:55 BRT
  SLULD1P-2P_SLU_P800_oradata03   19 SLULD1P-2P_SLU_P800_oradata03 1515 Stopped       2015-10-15 10:36:39 BRT


We are on a 10 minute period.  We “stopped” the RCIP Group at 10:38 BRT, which was 2 minutes after the last synch time of “oradata03” at 10:36.  “oradata02” synched at 10:28, which was 10 minutes before, and is the “oldest” synched volume.  Why didn’t Oracle recover back to 10:28 AM?  It had the data.  Why did it recover back to 10:20, which was 8 minutes “older” than the oldest data volume that it had?