PowerMax 2500 & 8500: Unexpected invalids during DR Test in a mainframe environment

Summary: Unexpected invalids during Disaster Recovery (DR) Test or Geographically Dispersed Disaster Restart (GDDR) Test.

This article applies to This article does not apply to This article is not tied to any specific product. Not all product versions are identified in this article.
Check out other resources

Symptoms

Unexpected invalids during a DR or GDDR test

Exposure: Mainframe (CKD) only, R2 SnapVX Target, PowerMax 2500 & 8500 only
 

No host reads or writes.
All LPARs are shut down.
All SDDF sessions are clear.
R2 Local Replication (LREP)
Remote Data Compare (RDC)
No customer process issues identified
Symmetrix Quadrilateral Asynchronous Replication (SQAR) configuration
Symmetrix Triangular Asynchronous Replication (STAR) configuration
GDDR failure
No errors posted internally.

Cause

This is due to a microcode issue. 
 

Below describes the sequence of events:

  • Remote Data Compare (RDC) is running on the SRDF/A leg
  • Snapshots are configured
    • V3 uses a different mechanism and does not perform copy-on-read behavior
    • SnapVX is rearchitected between V3 and V4
    • V4 defaults to copy-on-read and copy-on-write when intercepted
  • RDC work creates an intercept copy as the track status has changed to perform the checks between sites in an SRDF/A group
  • SRDF/A is brought down, and SDDF cycling stops
  • The SDDF bits involved keeping their cleared protection
  • The SDDF STAR D0 sees the cleared protections and is required to change them to invalids
    • Only one device is seen to have invalids due to the slowness of how RDC operates (checking tracks, which is random at the time)

      Resolution

      Contact the Support Center and quote this article ID.
       

      Workaround:
      Support can disable RDC in the array.
       

      Permanent Solution:
      PMOS-66457 addresses this issue in a future 10.2.0.x release

      PMOS-66461 addresses this issue in a future 10.3.x.0 release

      Fix: In the R2 Read IO flow, engineering removes the Snap Intercept Copy On Read bit when RA Snap Intercept is called. This prevents any copies during SRDF R2 reads where there is a target device as an R2.

      Affected Products

      Geographically Dispersed Disaster Restart (GDDR), Mainframe Enablers, PowerMax 2500, PowerMax 8500, TimeFinder/Snap, VMAX Local Replication
      Article Properties
      Article Number: 000328948
      Article Type: Solution
      Last Modified: 02 يونيو 2026
      Version:  7
      Find answers to your questions from other Dell users
      Support Services
      Check if your device is covered by Support Services.