Any workaround to progress saving for mask attack to restore file?
I have lost progress to command prompt freezes and other problems, and I am experimenting with persistent spot instances on EC2 as I have limited access to hardware and space due to my current situation.

From what I can tell, using a mask attack causes restore point saves to be stuck at zero for the entire mask process.  An .hcmask file with multiple masks seems to only save progress at each mask's exhaustion.

I also found this which supports the theory somewhat, and the response looks like it will not be addressed:

However I could be misinterpreting things, so if someone could clear up what I have found I would appreciate it.

One workaround I have found is to use MP64 to generate a .dict file of the mask but it has several limitations.  The first is file size of the generated dictionaries.  While not a huge deal locally, for an ec2 instance with a 25 GB hdd it can be a real problem.

The biggest limitation, however, is lack of support for markov chains.  The hash I am working on is a tough nut and is probably using a long password.  For longer masks I cannot use the full keyspace.  Not only would the file size generated by mp64 of the full key space be unmanageable, but the time to process it with hcat64 be impossible..  MP64 does not support markov chains.

DOes anyone have any input of how I can work around this problem?  Is there any way to save progress of large masks when cudahashcaat will be exited and restored often?  I have looked though the software options many times and cannot seem to find an answer.

THank you.
Your -n value is too high.