Start a Conversation

This post is more than 5 years old

Solved!

Go to Solution

2931

May 20th, 2014 09:00

SRDF/A write consistency

During the transmission of data from source to target is the target bucket labeled receive consistent the entire time during transmission?

Also if the link breaks during the transmission cycle is the received bucket data applied to disk or does it stay the way it is waiting for the link to come back?  I am asking because if it was a DR situation the host would need access to the latest data which would be in cache in the receive bucket.

13 Posts

May 20th, 2014 11:00

Are you talking single group SRDF/A or SRDF/A MSC?

In general in SRDF/A only the data in the Apply delta set on the R2 is Consistent.

In single group SRDF/A, if there is a disaster then any data in the partial receive delta set will be dumped. The R2 will be consistent up to the last Apply delta set.

In SRDF/A MSC. After a disaster before performing any recovery operation a SRDF/A cleanup operation is done to determine the disposition of all  receive delta sets in the various MSC RDF/A groups. cleanup will do the right thing and ensure that the R2s have consistent data.

Please read the EMC Symmetrix Remote Data Facility (SRDF) Product Guide - Do a search on "cleanup".

This guide is available on the EMC support site support.emc.com.


13 Posts

May 20th, 2014 11:00

Thanks. I am having trouble calculating the RPO. If a session is set to 15 seconds then the N-1 is up to 15 seconds behind at the switch. The n-2 is up to 30 seconds behind at the switch. If the disaster happens right before another switch then the N-2 data could be up to 45 seconds behind if a DR happens because data is still being gathered in the capture set. Is that correct? I know we typically say if the cycle is set to 15 seconds then the RPO is 30 seconds. Thanks for helping on this.

Thanks

Charles Urgitus, EMCTAe VCP5-DCV SPEED EMCISAV2

Solutions Architect

EMC²

Office: (336) 778-0465

Cell: (336) 624-2187

E-mail: charles.urgitus@emc.com

www.emc.com

13 Posts

May 20th, 2014 13:00

You can give yourself a headache very quickly trying to manually figure this out.

But if you are cycle switching precisely at the minimum cycle time then your RPO at worst should be twice your minimum cycle time.

Remember that if the R2 array has a complete receive Delta set then the clean up process will be able to use that data.

The best way to determine your exact RPO is to really use the symrdf query command with the -rdfa option.

The output will tell your precisely how far back your R2 devices are.
e.g.

C:\>symrdf query –rdfa

Device Group (DG) Name             : srdfadg

DG's Type                          : RDF1

DG's Symmetrix ID                  : 000192606020    (Microcode Version: 5876)

RDFA Session Number                : 9

RDFA Cycle Number                  : 697

RDFA Session Status                : Active

RDFA Consistency Exempt Devices    : No

RDFA Minimum Cycle Time            : 00:00:15

RDFA Avg Cycle Time                : 00:00:15

Duration of Last cycle             : 00:00:15

RDFA Session Priority              : 33

Tracks not Committed to the R2 Side: 7486

Time that R2 is behind R1          : 00:00:25

R2 Image Capture Time              : Thu Aug 22 14:25:53 2013

R2 Data is Consistent              : True

No Events found!

Top