100% safe, reversible and non-destructive
Instead of installing a proprietary file system on a set of disks, Tiger Pool creates a virtual volume into which any number of live volumes are “pooled” together. The result is a unified namespace in which a logical merge of all available data can be found.
The total size of the namespace is equivalent to the sum of all volumes sizes and the available free space is equivalent to the sum of all available free spaces. There are no restrictions as to how much content can be stored or moved to a given folder. On the surface the pool just looks like a single large drive into which you would have copied the content of all the volumes it actually contains.
Scales to multiple petabytes
Any file system can be at risk of some form of corruption. Since Tiger Pool does not create or maintain metadata, but rather combines multiple volumes together, it is never at risk of losing anything. If any of the pool members get corrupted (hardware malfunction, power failure, etc.), only the data of that member will disappear from the pool until it can be restored. Meanwhile, the rest of the pool remains functional.
When cumulating petabytes of data, this could be the difference between restoring for just hours compared to weeks.
Transparent live migration of files between pool members
Files stored in the pool are stored on a pool member. Tiger Pool makes it easy to transparently migrate data from one pool member to another.
Meanwhile, the pool can remain in use. Only the files that are currently being accessed will not be moved.
Supports standard NTFS file systems with Active Directory
Growing a pool is as easy as formatting a new disk array using NTFS or Tiger Store
and adding it to the existing pool.
Decommissioning an aging array is as simple as bringing a new array into the pool, and initiating a migration between the two pool members. Both operations are totally risk-free and can be performed between pool members of completely different characteristics.
Supports SMB shares for easy deployments
Once created, the pool can be accessed as a local drive and can be re-shared over SMB for easy access by other users.
Relies on simple data distribution policies
When reading from a pool, all physical volumes it contains are solicited so content can be presented to the users as a unified volume. When writing to the pool, a simple policy dictates on which pool member the new data is to be written.
Any given file stored into the pool will always reside on one of the pool member. As there is no distribution of content, it is important that all pool members are properly RAIDed to ensure data protection and integrity.
Tiger Pool has been highly optimized to ensure virtually no delays get introduced.
Accessing the pool is just as fast as accessing the underlying file systems. As such, the most demanding operations, such as color grading of image file sequences, can take place on the pool.