Cofio Software

Performance

ViStor attacks many aspects of performance, in order to give it a combined level of speed when applied as a solution to a network's storage strategy.But to help you understand on a more detailed level, we have summarized several performance areas in which ViStor holds an exceptional edge against its competitors.

Restores Come Fast - Restores of your mission critical data is far faster on ViStor disk than tape, because it is coming from fast disk, rather than a tape media that must be found, loaded, and accessed prior to any data transfer.

Reduced Backup Windows - Similarly, ViStor backups to disk are faster as well, resulting in smaller backup windows for the network.

More Virtual Drives - Multiple virtual tape drives supported by ViStor drive down network backup windows by servicing numerous backup streams from numerous systems simultaneously. Need to go faster? Add another virtual tape drive.

Multi-Threaded Transfers - The ViStor VTL utilizes a multi-threaded architecture, which maximizes multiprocessor capabilities and simultaneously optimizes concurrent data transfers. This means that while the ViStor sends out multiple threads to several CPUs for crunching, it is also making sure the multiple data transfers are happening at the same time for maximum speed and efficiency.

Multiple Network Interfaces - ViStor sends multiple streams of data on multiple network adapters. If you are working with a ViStor server system that has 4 network cards engaged in high data transfers during a backup or restore, ViStor will support this high rate of concurrent data exchange.

Multiple RAID Sets - ViStor can further optimize backup and restore duties by reading and writing to multiple RAID sets at the same time. If your ViStor server holds numerous RAID sets, the most efficient use of it would entail multiple stream support, in the event that multiple requests must be serviced simultaneously.

Kernel Centric - ViStor removes the overhead of kernel resident memory being transferred into user space and back into kernel as part of its normal operations. This eliminates the multiple calls back and forth between layers that would normally result from other architecture styles, as well as the context switching overhead. Performance is greatly improved.

CPU Offload - A dedicated ViStor server can easily offload the main backup server, freeing up CPU cycles for managing faster backups.

Optimized Cache - When caching write IO in resident memory, ViStor holds the cache available for instant transfer or processing, resulting in faster response to IO calls.