Intel Server RAID Q1 2009 boot camp training lab workbook

Intel

Server RAID Boot Camp Training Q1 2009

Answers

Lab 1-1. If we change 8 KB random read pattern to 16 KB random read, which performance

metric changes more – IOPS or MB/s?

There is almost no change in IOPS, because reading 16 KB versus 8 KB adds very small delay

compared to moving heads from one location to another. Performance measured in MB/s is

almost double, because with almost the same amount of IOPS, the amount of data

transferred in every I/O is 2X times larger.

Lab 1-2. If we change 1 MB sequential read pattern to 2 MB sequential read, which

performance metric changes more – IOPS or MB/s?

There is almost no change in MB/s, because with both 1 MB and 2 MB block sizes the

performance is bottlenecked by sustained sequential throughput of individual HDDs.

Performance in IOPS becomes two times smaller, because it takes two times longer time to

read each block.

Lab 2-1. Which RAID Read Policy between No Read Ahead mode and Adaptive Read Ahead

mode provides better sequential read performance and why?

We can get better sequential read performance in Adaptive Read Ahead mode. That is

because the RAID controller recognizes sequential access pattern and requests data from all

HDDs in parallel instead of one-by-one.

Lab 2-2. Why does Write Back mode provide larger performance benefits than Write

Through mode during a sequential write performance test?

In Write Back mode, the RAID controller acknowledges write I/O requests immediately after

the data loads into the controller cache. The application can continue working without

waiting for the data to be physically written to the hard drives.

Lab 2-3. Why is SATA 2.0 disk write cache unnecessary if we set RAID Write Policy to Write

Back mode during a sequential write performance test?

In Write Back mode, the RAID controller can effectively utilize queuing capabilities (NCQ/TCQ)

of the HDD, which makes caching at the disk level unnecessary.

Intel Corporation