This paper is the report of performance effects of various disks. The author, Marshall Kirk Mckusick, actually benchmarked various combinations of the three manufacturers disk controllers, and two pairs of Winchester disk drives on both VAX 11/780 and VAX/750 CPUs. The Emulex and Systems Industries disk controllers were interfaced to Fujitsu 2351A "Eagle" 404 Megabyte disk drives. The DEC UDA50 disk controller was interfaced to two DEC RA81 456 Megabyte Winchester disk drives.
DEC UDA50 disk controller is a new controller design which is part of a larger, long range storage architecture referred to as "DSA" or Digital Storage Architecture. Emulex SC750/SC780 disk controllers interface to the CMI bus of a VAX 11/750 and the SBI bus of a 11/VAX 780. System Industries 9900 disk controller is an evolution of the S.I. 9400 first introduced as a UNIBUS SMD interface. DEC RA81 disk drive is a rack-mountable 456 Megabyte Winchester disk drive manufactured by DEC. Fujitsu 2351A disk drive is a Winchester disk drive with an SMD controller interface.
The author uses two methods to approach to evaluate the performance of the target peripherals in an environment as much like 4.2BSD UNIX systems as possible. Two different approaches are indirect and direct.
The indirect approach used by DEC involves two steps. First, the environment in which performance is to be evaluated is parameterized. The second stage involves simulating this mixture of I/O activities with the devices to be tested and nothing the total volume of transactions processed per unit time by each system.
The direct approach uses the standard structured file system mechanism present in the 4.2BSD UNIX OS to create the sequence of locations and sizes of reads and writes to the benchmarked equipment. The author chooses this methodology and simply creates, write, and read files, as they would be by user's activities.
The author shows numbers of test results, compares drives, and concludes his benchmark results. He draws from the tests that buffering can be effective in smoothing the effects of lower bus speeds and bus contention and placing more intelligence in the controller seems to hinder UNIX system performance more than it helps.
The paper was written in 1983 so information was a little bit old and useless for recent systems. However, it helped me a lot to understand I/O system in UNIX system but I still don't know and have never heard any hard drives and controllers.
I was reading white paper on IBM website last night about 75 gigabytes hard drive and its specifications. Compare to this hard drive and controller and the old UNIX I/O devices such as Fujitsu 2351A "Eagle" 404 Megabyte disk drives and DEC RA81 456 Megabyte Winchester disk drive, actual capacity of disk drives is incompatible. However, there is a point to benchmark these old drives and controllers in the past. Memory was expensive and users of super computers want to effectively share all resources. That¡¯s why the author did benchmarks these devices.