Disk throughput is used to measure the utilized (actual) bandwidth between the host and its storage devices. Throughput is the measurement of bandwidth consumption of a storage device or subsystem from its host(s)
In an Optical Prime project this means an aggregated quantity of read in write data transfers reported in x Bytes/per second.
Optical Prime measures throughput in terms of the total bandwidth consumed for reads and writes at a host level. There are other factors to consider in when measuring throughput; however DPACK distills this measurement leaving little guesswork to understanding what the Host OS needs to fulfill front-end bandwidth requirements. For storage design, it is important to consider how disk devices, arrays, and SANs can handle the throughput needs of the host so that latency is kept to a minimum.
General rules of thumb
Avg. Throughput = IOPS * Block size
Related To:
Throughput
Block (IO) Size
Queue Depth
Latency
Sequential v. Random I/O