After performing a Planned Migration from Production to Recovery, running a Reprotect results in the following transient UI errors, even though the Recovery Plan ultimately finishes successfully:
Error - Failed to sync data on replica consistency group [Name]. Invalid device. Refer to IBM Storage Virtualize Family SRA troubleshooting
Error - Data could not be synchronized in replication consistency group 'Name'. Invalid device. Refer to IBM Storage Virtualize Family SRA troubleshooting.
VMware Live Site Recovery 9.0.x
IBM Storage Virtualize Family SRA 4.2.0 20250228
To remediate and prevent these errors, implement the following steps.
NOTE: It is recommended to verify specific value adjustments with IBM SRA Support1. Allow the Protection Window to Expire (Short-Term Workaround)
Wait 15 to 20 minutes without modifying the recovery plan. This allows the internal IBM storage array volume protection timer to expire naturally since active host I/O has stopped.
Once the time window clears, select your Recovery Plan in SRM and re-run the task (e.g., Cleanup or Reprotect). The subsequent execution should pass the volume checks.
2. Adjust IBM Volume Protection Settings (Long-Term Fix)
If this error continuously impacts automated testing or disaster recovery windows, work with your storage administration team to reduce or disable the global volume protection period on the IBM Spectrum Virtualize/SVC cluster.
This can be adjusted via the IBM CLI: chsystem -vdiskprotectiontime <minutes>
NOTE: IBM pre-requisites document that vdisk protection must be disabled before performing operations in non-preconfigured environments.3. Throttle SRA Command Concurrency
To prevent future HTTP 429 (Too Many Requests) rate-limiting blocks, lower the maximum number of concurrent storage operations sent by SRM.
In the vSphere Client, navigate to SRM Advanced Settings and reduce the value for storageProvider.maxSimultaneousStorageOps.
Consult IBM Storage Support to determine the optimal customized values for storageProvider.maxSimultaneousStorageOps and storageProvider.hostRescanDelay for your specific array model.
IBM SRA Support has acknowledged this issue and is working to resolve it in a future release of SRA. Submit a case with IBM SRA Support, if you have further questions about this.