ReFS recovery using data recovery software
If the ReFS volume has failed and you cannot access the data, the only option for you is to try to recover data.
The point is that the ReFS developers designed ReFS as a filesystem which does not fail,
so ReFS does not have any built-in means of repairing. Nevertheless, if you read this article, you have already discovered that
ReFS can fail as well as other filesystems.
So to recover data from a failed ReFS volume you need to use one of the data recovery software which is capable of ReFS recovery.
Unfortunately, there are a small number of tools which are able to recover data from it,
not to mention the total absence of free ones. If your ReFS volume failed, we recommend you try the following software:
ReFS recovery takes a long time
ReFS filesystem was designed to be used on large storage,
so ReFS is equipped with various techniques to handle large volume of data quickly and efficiently.
Among these techniques is the usage of B-tree structures to provide fast access to the data.
However, when recovering data, it is impossible to use those structures which provide fast access during the normal operation,
since they are most likely damaged. It means that to recover data from a ReFS volume,
it is needed to analyze all the data stored on the storage.
So expect that ReFS data recovery will take a lot of time even in case of a typical size of a ReFS volume.
Storage Spaces failure
When the ReFS developers were creating the ReFS filesystem, they planned that it would work in cooperation with Storage Spaces.
It means that if you are not able to access the data on a ReFS volume, ReFS failure apart,
you need to take into account possible Storage Space failure as well.
Other ReFS data recovery specifics
ReFS filesystem is designed in such a way that modified metadata is not written to the original location.
Instead, the ReFS driver writes the data to the new location, increases the number of generation of the current record, and updates
all the pointers related to the previous record.
The copies of the previous records are stored on the disk until the disk space is really needed for a new record.
The "copy on write" feature of the ReFS filesystem not only helps ReFS avoid "torn writes",
but in addition significantly increases the chance to recover data successfully.
Since there is no official documentation about the ReFS filesystem and most likely will never be any,
all the algorithms used to recover data are based on "reverse engineering".
It means that information on how the ReFS filesystem works is obtained from the looking at the test volumes.
If your specific case differs from those cases which were used in the analysis of ReFS filesystem,
it may happen that the software cannot recover some part of your data.
In any case you should either try some other data recovery tools in the hope that they utilize other algorithms or
contact one of their technical support teams for help. Unfortunately,
such a situation will continue for quite a long time since ReFS is a "young" filesystem.