This post is more than 5 years old
1 Rookie
•
33 Posts
0
736
October 20th, 2015 05:00
Is RPA local or remote copy a crash consistent image when cg replicate normally
If I had three ARRAY, first is Production array, second is CDP/CRR array, and use ARRAY sync/async replication solution to replicate the CDP/CRR volume from second ARRAY to the third ARRAY, and recoverpoint CG replicate and distribute journal normally, could I get a crash consistent image in the third ARRAY when the Production array and CDP/CRR array both damaged?
No Events found!
forshr
2 Intern
•
1.1K Posts
0
October 21st, 2015 07:00
Yes, RP is constantly distributing snapshots to the journal and replica so there will always be a crash consistent image on the replica as long as its not in a long resync situation. Of course it may be not up to date due to journal or replication lag.
forshr
2 Intern
•
1.1K Posts
0
October 20th, 2015 11:00
This would constitute cascaded replication and is not supported by RecoverPoint. You need to establish the remote copy to the third array from the Prod array in the same CG as the copy to the second array. As long as full RP has been installed, as opposed to RP/SE, and EX or CL license files are applied then you have a 1:4 fan-out replication capability.
Regards,
Rich Forshaw
Consultant Corporate Systems Engineer - RecoverPoint & VPLEX (EMEA)
Data Protection and Availability Solutions
EMC Europe Limited
Mobile: 44 (0) 7730 781169 44%20(0)%207730%20781169>
E-mail: richard.forshaw@emc.com
Twitter: @rw4shaw
liuchuanwen
1 Rookie
•
33 Posts
0
October 20th, 2015 18:00
hi Forshr, thanks for your response. Could I ask another question:
If I lose DR journal for any reason at a time, maybe the array which store Journal crashed, but the array which store replication volume still work. Could the replication volume be accessed with crash consistent? As I know, the Journal is distributed to replication volume in turn, so when Journal lost, the replication volume should be an image of a point in time(although we don't know what is the exact time, that depend on the journal Lag).
As I understand, although there is an image in the replication volume, but it can't be guaranteed to be an crash consistent image, so whether the data in replication volume can be used is depend on luck. Am I correct?