none
FSCT message while testing server with McAfee vscan RRS feed

  • Question

  • What is FSCT saying after an hour or so running - it then continually reports:

    After running the test for a couple of hours, the client finally continually prints this message over-and-over:

    Unable to get next file from the file set. The file lists are empty.
    Unable to get next file from the file set. The file lists are empty.
    Unable to get next file from the file set. The file lists are empty.
    Unable to get next file from the file set. The file lists are empty.
    Unable to get next file from the file set. The file lists are empty.
    Unable to get next file from the file set. The file lists are empty.
    ...

    The vscan client (client to the 136 NexentaStor server) runs and prints this - which seems normal:

    PS C:\FSCT> .\fsct.exe run controller /server dell-53-135 /password Nexenta1 /volumes \\dell-53-135\fsroot1 /clients fsct-client /min_users 100 /max_users 100 /duration 432000000 /workload HomeFolders
    server's name: dell-53-135
    server's password: Nexenta1
    volumes: \\dell-53-135\fsroot1
    clients' names: fsct-client
    min users: 100
    max users: 100
    test's duration: 432000000
    workload to use: HomeFolders
    Executing: rd /s /q running
    Executing: rd /s /q running complete.
    Executing: md running\signals
    Executing: md running\signals complete.
    Running the test (workload: HomeFolders).
    Running iteration for 432000000 seconds with 100 users [warmup: 30s].
    Performance Counter thread started for dell-53-135

     

    Sunday, November 10, 2019 10:02 PM

All replies

  • Wednesday, November 13, 2019 8:29 AM
  • Thanks to pointer by 'Alrea' realized using the /duration 4300000 (five days) is not the right way to setup long running test.  The FSCT document 2011 says this about that in bold:

    • For the most stable results, we recommend that you reboot the file server before each test in order to put the server in a more clean state. If you are planning for comparisons on different hardware, you should consider running with the same number of users and the same incremental (step) size in the run command. For the test duration, we recommend a run time of 10 minutes. If the time is too short, the results will have higher variance, and if the run time is too long, the workload may run out of files and start to report errors.

    Thanks

    Wednesday, November 20, 2019 12:04 PM