Installation manual

94
3.2 ATTO Disk Benchmark
The ATTO Utilities for Windows suite includes a utility to measure peak and sustained throughput for disk
reads and disk writes.
The ATTO Disk Benchmark measures peak and
sustained throughput for disk reads and writes.
1 Launch the application.
2 Select the drive letter for the disk to benchmark.
3 Select the transfer sizes to test.
4 Select the I/O option.
5 Click the Start button.
6 Wait for benchmark to run through the desired
transfer sizes.
7The Test Results Display at the bottom of the
window is updated as the test progresses. The
y-axis of the graph represents the transfer sizes
in the selected range. The x-axis represents the
transfer speeds in MB/sec. I/O speeds in
KB/sec. for each transfer size are displayed
textually to the right of the graph.
8 Click the Stop button to stop the test.
When the test completes, the results can be
saved or printed.
If errors were detected, a dialog box displays the
errors in a table with the following four columns
and a button:
• Benchmark Transfer Size: transfer size at
which the error occurred.
• Buffer Index: index into the data block at
which the error occurred.
• Actual Value: the value read from the file.
• Expected Value: the value written to the file.
• Log to File: Logs the error table to a *.log
file and closes the dialog. The file is given
the same name as the test file and saved in
the same directory. If the test was not
previously saved, errors are logged to the
generic file Bench32Error.log in the root of
the test drive. If the log file already exists, the
new errors are appended to the previously
recorded errors. This is the only way to save
detected errors. They are not saved in the
test document file.
If the I/O comparison option was selected and
errors were not detected, the message No
errors detected is displayed.
Note
Additional information on using Bench32 is
available by accessing the
Help menu in the
application.
Benchmark fields
The benchmark fields include.
• Drive: Select the logical drive to benchmark. A
test can be performed on any system drive.
• Transfer Size: Select the range of transfer sizes
used for reading and writing data to the test
drive. Transfer speeds are displayed for each
size in the range. If the first size is greater than
the second size, the test is not performed for any
transfer size.
• Total Length: Select the total size of the data file
to be created on the test drive. This file is
deleted when testing completes.
• Direct I/O: If this option is checked, file I/O on
the test drive is performed with no system
buffering or caching. Combine this option with
Overlapped I/O for maximum asynchronous
performance.
• Radio Button Group
• Overlapped I/O performs queued I/O. Upon
selection, the Queue Depth option displays to
select the maximum number of read or write
commands which may be executed
simultaneously.
• I/O Comparison compares the data read from
the test file to the data written on a per block
basis. You can select the data pattern for
comparison from the Test Pattern dropdown
box.
• Run Continuously runs the test continuously
for a specified number of minutes. The test stops
before the specified time if any errors are
detected.
• Neither: Select if you do not want to perform
overlapped I/O or I/O comparisons.
The following fields do not affect the benchmark but
are informational, providing documentation of the test
environment.