Repeated 'IO error to recovery disk' alerts after a protected VM disk is resized to a non whole GB size

Repeated 'IO error to recovery disk' alerts after a protected VM disk is resized to a non whole GB size

Machine-distilled from a resolved support ticket on 2026-09-28. Source ref: ee9553f4c9f3. Verify before relying on it.

Applies to: Zerto on vSphere where a protected VM disk has been expanded while the VM is protected. Not confirmed against specific Zerto versions.

Symptom: A single VPG repeatedly enters and leaves an error state, and a very large number of alert emails are generated stating 'IO error to recovery disk' for a recovery volume of the protected VM on the recovery VRA. Other VPGs on the same platform are unaffected.

Cause: A disk on the protected VM was expanded to a size that is not a whole number of gigabytes (for example a size entered as a fractional terabyte value). Zerto could not correctly mirror the resize to the recovery disk, so the source and recovery disk sizes no longer lined up and writes to the recovery disk failed. In a similar earlier case, a backlog of VM snapshots on the protected VM also contributed.

Resolution: 1. If only one VPG is affected, treat it as a VPG specific issue rather than a platform or storage fault, and check the VM or Zerto logs for a recent disk resize (a SCSI volume resize event).

2. Confirm the new size of the resized disk. If it is not a whole number of GB, resize the protected VM disk to the next whole GB value in vSphere.

3. If the protected VM has an accumulation of snapshots, consolidate or remove them before resizing.

4. Check that the source and recovery disk sizes match, and realign them if they do not.

5. If the VPG still does not recover after the sizes are realigned, run a force sync on the VPG.

6. Advise that protected VM disks should always be expanded in whole GB increments.