Article number
Affected Versions
Source Hypervisor
Target Hypervisor

All VPGs Replicating to/from a ZCA Cannot Maintain Stable Replication or Stuck in a Ongoing Sync

Viewed 386 times


An administrator may notice as further VPGs are added to a ZCA that performance/stability will start to deteriorate, leading to higher RPOs, higher RTOs, and some VPGs stuck in ongoing syncs.

Root Cause

Likely an over saturated ZCA (too many mirror disks/operations ongoing) as the VRA was reporting high CPU utilization. This can also be seen from the ZCA's Task Manager Performance tab for real time monitoring of resources. However, Task Manager reports on all resources for the ZCA, not the resources strictly for the VRA service.


VPG slow/stuck sync to Azure/AWS or RPO is continuously increasing.

ZCA's VRA logs show high CPU utilization (see below) as does the Task Manager Performance tab on the ZCA server itself.

ActiveCompressor,CompressionLevelsController::checkCpu,CPU is high cpuUsage=99 

Scaling up the ZCA resources does not provide relief.

Planner within Analytics reports that more than 1 ZCA is required for the current workload assigned to it (available in 7.0 Update 1 and higher).


To resolve this issue, follow the below steps:

  1. Create a net-new ZCA (or more if required).

  2. Pair the new ZCA(s) to the same site that the original ZCA is paired to.

  3. Balance out the VPGs between the two or more ZCAs in terms of number of mirrors/VMs and overall change rate. This will require exporting the VPG(s), destroying the VPG(s), then importing the VPG(s) via Diagnostic Utility (or recreating manually) to the new ZCA(s).

Additionally, if on Zerto version 7.0 Update 1 or later, Analytics' Planner feature is available to use to help determine how many ZCAs are required to protect the selected workload. This can be used on a reaction and proactive basis (as in, estimate of ZCA count can be determined even if VMs selected are not currently protected).