Restoring a Large Number of Files from an Image Backup Fails. (4040188)

When restoring a large file set (several million) from a virtual machine image, it is possible to exhaust the available memory on the machine performing the restore. 

BUG 26465

Processing and sorting a large amount of meta-data, to allow for restores from tape devices, requires a large amount of information to be retained in memory.  

Currently there is no method of restoring a large number of files in a restricted amount of memory.  Alternatives methods are being investigated for a future release.  Until an alternative is available the following workarounds have been identified.

1. Restore smaller sets of files spreading the load across multiple jobs.
2. Increase the available memory on the restore server.
3. Restore the Virtual Machine image to the vCenter/ESXi host renaming it to create a copy. Then copy the files from the secondary machine to the original.

To help plan your restores, test restores of millions of files have been performed to provide estimates on the amount of memory required to restore an increasing number of files.  

Note: These numbers are approximates and will change depending on your data.