Are there plans to implement a prorietary client to minimise network usage
by implementing true incremental/delta backups? a.la rsync algorithm or
By splitting file data into delta's (compared across and referenced by
complete data set)
I suspect online disk requirements can be reduced much much more.
Lastly, should consider the structure of files on disk to support
with a HSM backend (for tape storage) inline with data retention.
As opposed to writing your own.
Their my thoughts, happy to code and contribute, however looking for
direction and an
understanding of what is already in the works.