Large filesystem backup

Marc Fisseler OpenText Employee

Status: Under Consideration

Under Consideration

See status update history

over 4 years ago

Data Protector should provide a “file system connector” to back up a large amount of data in a very efficient and fast way. (GPFS, NTFS, cluster technologies, …, big data lakes). It is critical to have parallel streams and additional technologies to maximize backup performance.

Goal is to have minimum read, no (minimum) tree-walk, minimum transmit data for very large file systems.

This feature should be supported by all main deduplication device providers and in addition by a Data Protector software deduplication target (VSA).

Tags:

Other

backup

Ulrich Windl over 3 years ago

Assuming even big filesystems have some (logical) structure, I'd also think manually creating different subtrees of it might do what is being asked for. However if there's only on disk behind, I wonder if the performance will be any better when using multiple streams on the same filesystem/disk.
- Cancel
- Vote Up 0 Vote Down
- More
- Cancel
AlexH over 4 years ago

Would gladly test a first implementation with one of our GPFS filesystem.
- Cancel
- Vote Up 0 Vote Down
- More
- Cancel
Jesko over 4 years ago

Hi,
so far an often used way is to manually split the data on one filesystem so it is backed up with multiple streams. It would be great if DP could do this on its own, by just specifying the number of streams for the Filesystem.

Regards
Jesko
- Cancel
- Vote Up 0 Vote Down
- More
- Cancel

Menu

Data Protector Ideas

Labels

Large filesystem backup

Tags:

Resources