On Site Recovery failback, replicated devices are not represented to original hosts. (HPE SRA)

Products

• VMware Live Recovery

Issue/Introduction

Site Recovery failover completes successfully, but during failback, replicated devices are mounted on multiple clusters rather than the original location.

Environment

Site Recovery manager 8.x

 

Cause

SRA parameter ’SaveExportInformation’ is set in the ‘sra.settings’ file.

“When the flag is enabled, SRA uses the original Host mapping, configured before the Failover, to export storage LUNs during the Failback process.
When the flag is disabled, SRA continues to use the Host list shared by SRM to export storage LUNs.”

It is disabled by default.

HPE SRA have the same parameter named ‘SaveExportInformation’ with the same scope. The default value is True. SRA stores the export information in the repository to try to keep original exports.

 

Resolution

Currently, this issue is only reported with HPE SRA

SRM does not make any assumptions about the format of the SRA configuration or the number of configuration files. "SaveExportInformation" is a SRA configuration parameter. SRA configuration parameters should be set by Storage vendor.

Additional Information

https://www.ibm.com/docs/en/spectrumaccel-sra/3.2.0?topic=uscf-updating-sra-configuration-file-from-sra-container-volume-preferred-method

https://support.hpe.com/hpesc/public/docDisplay?docId=sd00004131en_us&docLocale=en_US&page=GUID-66936794-1B81-408C-9A42-6258689C0345.html