Article number
Affected Versions
7.5 Update 1
Source Hypervisor
Target Hypervisor

Alert message: “IO error to recovery disk [{Datastore}] Z-VRA-hostname/xxxxxxxx/vmxxxxx/filename.vmdk of virtual machine Z-VRA-hostname.” might show when starting and/or stopping a failover test

Viewed 260 times


IO error alerts can be thrown during Failover Tests while mirror disks are attached and detached from the Testing Recovery VM and the VRA.
Note, it's important to look at the VRA and host logs to verify that there isn't any storage issue. If unsure on how to do so, call Zerto Support for additional help.

Root Cause

A bug  was opened with Development regarding IO alerts being thrown while starting and/or stopping Failover Tests.  Development acknowledged that the underlying cause of the alerts thrown during this operation is because the VRA is trying to apply journal data to the mirror disk while it's in the process of transferring (detaching/attaching) between the Test VM(s) and the VRA during Failover Test. The alerts thrown are cosmetic because the VRA can't write what it's trying to for a short period of time when this is occurring. When the disks become available, it proceeds with writing to them.  The alerts typically takes one minute or less to go away.


A user may experience alerts generated in the ZVR GUI regarding IO error to recovery disks when they start or stop a FOT.

For example the following flow may occur:

12/11/2019 1:22:33 PM - {VPG Name} - Stop Failover Test. Completed successfully: Stop Test for VPG {VPG Name}

Alert turned on at 12/11/2019 1:22:34 PM: IO error to recovery disk [{Datastore}] Z-VRA-hostname/xxxxxxxx/vmxxxxx/filename.vmdk of virtual machine Z-VRA-hostname.



These alerts can be ignored.  At this time Development is not planning to change this behavior.