Best Practices and Considerations

Consider the following concepts before beginning I/O throughput tests:
  • SAS performs the Write and Read operations with the Windows file cache. The file sizes for the Write and Read tests should be larger than your host system file cache. They can be larger than the RAM in your host system. For example, if you run your test with a 1GB file size, and your system has 5GB of cache, then the test processes limited disk I/O on the Read test. The file that is written remains in the system file cache, and the subsequent Read operation reads it from cache instead of from disk. If the Read tests are satisfied with cache instead of disks, the Read performance metrics are misleading, and you cannot generate consistent, repeatable results.
  • An invocation of a single-instance test of either of these tools, one Write and Read test, shows how a single SAS job might perform at that time on the file system. If your system is already very busy with other users’ workloads, this action might be adequate.
  • A multiple-instance invocation of the tests shows what a concurrent SAS working load might experience, and is recommended for systems that are experiencing no other current jobs, or that have a very light workload.
  • Conduct several of these test sets at various times of the day, and over several busy days to get a comprehensive profile of I/O characteristics on systems that run a variety of workloads.
  • Perform a test during the same time of day and on the same file system or systems.
  • Specify a file that you want to create in the file system that you want to test. If you are using batch mode, you can use the redirection operator (>) to send the output to a different file system. Here is an example: a .bat file with sasiotest.exe […arguments…] > test.txt.
  • The output from this tool is the megabytes written per second and megabytes read per second. Use these metrics to help analyze and tune I/O for file systems that support SAS.