The Import Method selected for the VPG is 'ZImport For Data Volumes' or 'AWS Import', where the "ImportInstance" AWS API call is in charge of importing the boot disk.
Additionally, if Page Files are used, they are not solely located on the boot disk.
Additionally, sometimes disk letters can swap (for example: volume D: becomes E: and E: becomes D:). If all volumes keep their letter labels as they were, then the instance will be able to use the Page Files on other disks even after Failover to AWS with ZImport For All Volumes.
To relieve the stuck Failover task:
1) Restart the ZCA services.
2) Monitor the VPG as it will go into an 'Error' status and eventually self resolve.
To avoid the Failover task getting stuck in future:
1) After the VPG is back in a healthy status, edit the VPG and change the Import Method to 'ZImport For All Volumes' for each Protected VM.
2) Configure ZertoTools for Windows across all Protected VMs in the VPG.
3) Await a checkpoint to be selectable in the Failover set-up with a timestamp from after the ZertoTools for Windows was successfully configured across all Protected VMs in the VPG.
4) Start another Failover with a checkpoint as described in Step 3 above.
Should the use of ZImport for Data Volumes or AWS Import be mandatory to use:
1) Review the ZVM log for a line like below that provides the Conversion Task ID. In the example below, the Conversion ID is 'import-i-064ac9bf568fb7079'.
91e766df,da7143c3,,19-04-03 13:31:46.53,I,75,AWSComputeProxy,CheckIfConversionTaskIsCompleteOrCancelled,Task import-i-064ac9bf568fb7079 state is active status Progress: 55% #bytes converted 64424509440 out of 64424509440,
2) Provide the Conversion ID to AWS Support to determine the cause of the 'ImportInstance' becoming stuck.