Performance - Hard (disk|drive)
Table of Contents
1 - Concept
Many factors are involved in the performance of a disk system such as:
- the level of utilization,
- the rate of throughput,
- the amount of disk space available,
- the presence and development of a wait queue,
- the interruptions generated by the disk system and paging activity (which influence other resources principally processor or memory).
Certain conditions might result in increased response time and cause performance to slow.
2 - Articles Related
3 - Interrelation of the factors
Many of these factors are interrelated.
For example, if utilization is high, transfer rates ( throughput ) might peak, and a queue might begin to form.
4 - Disk Space
Although disk space doesn't directly affect the disks transfer rate, when extremely low, disk space can also have an influence on response time because applications that read and write data can't do so efficiently.
5 - Material
Use the bus, controller, cabling, and disk technologies that produce the best throughput that is practical and affordable. Most workstations perform adequately with the most moderately priced disk components.
If your configuration contains different types of disks, controllers, and buses, the differences in their designs can have an influence on throughput rates.
6 - Disk configurations
6.1 - Volume-set
In addition, the use of certain kinds of volume-set configurations can offer performance benefits. For example, using striped volumes can provide better performance because they increase throughput by enabling multiple disks to service sequential or clustered I/O requests. (Striped volumes are not fault tolerant.) System Monitor supports monitoring volume sets with the same performance objects and counters provided for individual disks. Notice that hardware-based RAID devices report all activity to a single physical disk and do not show distribution of disk operations among the individual disks in the array.
6.2 - File system
6.3 - Disable I/O PerfCounter
By default, Task Manager continuously measures data for process I/O operations that you can select and display on the Processes tab in Task Manager. In a multiprocessor environment, this data is shared by the processors on which the process runs.
When a process that generates considerable disk and network I/O, such as a database service, runs on several processors, updating the shared measurements of process I/O and global I/O operations can slow the system.
You can improve the performance of I/O-intensive operations on SMP systems if you configure the system to bypass the global I/O counters and Task Manager process I/O counters.
To do so, add the CountOperations entry to the registry as a REG_DWORD in HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Session Manager\I/O System. (If the I/O System subkey is not present, add it before creating the entry.)
Set the entry value to 0. When so configured, Task Manager no longer provides per-process I/O measurements.
6.4 - Misalignment with MBR limitation
Windows 2000 has an internal structure called the master boot record (MBR) that limits the maximum number of hidden sectors to 63. This characteristic of the MBR causes the default starting sector for disks that report more than 63 sectors per track to be the 64th sector.
As a result, when programs transfer data to or from disks that have more than 63 sectors per track, misalignment can occur at the track level, with allocations beginning at a sector other than the starting sector. This misalignment can defeat system optimizations of I/O operations designed to avoid crossing track boundaries.
Additional disk-design factors make proper alignment even more difficult to achieve.
For example, track information reported by disks is not always accurate.
Diskpar.exe (Windows 2000 Resource Kit program) shows how you can use Windows 2000 APIs to obtain and set partition information. By applying the same functions used in this tool, you can avoid performance loss due to disk misalignment on disks with large track sizes and alignment optimizations.