We’re migrating product documentation to docs.vmware.com, starting with Carbon Black Cloud. Learn more.

How to interpret the results.txt output from CBPTest

How to interpret the results.txt output from CBPTest

Executing CBPTest Tool - Disk Performance Testing Tool results in a log file Results.txt. To interpret these results:

NOTE:  There has been an update to the script used and the way the results are displayed.   Per the comment from Matt Larsen at the bottom of the page:

There is a change in the output. In order to increase accuracy, multiple test files are used. So you will see individual results per file, and a total.  The INDIVIDUAL results should be nearly identical, and that is the value we use - not the total.

  1. Open Results.txt a text editor such as Notepad
  2. Scroll to the first set of throughput results and look for the total MB/s number under "Write IO" on line 92. This is the Sequential Writes result.
    Example screenshot:
    CBPTest_SequentialWrites.png
  3. Scroll to the second set of throughput results and look for the total MB/s number under "Read IO" on line 197. This is the Random Reads result.
    Example screenshot:
    CBPTest_RandomReads.png
  4. Scroll to the third set of throughput results and look for the total MB/s number under "Write IO" on line 316. This is the Random Writes result.
    Example screenshot:
    CBPTest_RandomWrites.png
  5. Compare the results to the requirements listed on Page 15 of Bit9 Security Platform v7.2.3 - SQL Server Configuration for Bit9SqlIo. Note that CBPTest is a newer test that has replaced Bit9SqlIo, but the required performance metrics are identical. Screenshot:
    Screen Shot 2016-09-07 at 22.14.55.png

In the example shown above, the storage is capable of supporting a Cb Protection instance with up to 500 endpoints, with "Random Writes" being the limiting factor.

Labels (1)
Comments

Here's a quick and dirty shell script for those on *nix/OSX to grab and print those three fields:

#!/usr/bin/env bash

# Useage:

# ./grab_cbptest_data.sh Results.txt

echo -n "Sequential Writes MB/s: "

head -92 $1 | tail -1 | cut -d "|" -f 3 | sed -e 's/^[[:space:]]*//'

echo -n "Random Reads MB/s: "

head -197 $1 | tail -1 | cut -d "|" -f 3 | sed -e 's/^[[:space:]]*//'

echo -n "Random Writes MB/s: "

head -316 $1 | tail -1 | cut -d "|" -f 3 | sed -e 's/^[[:space:]]*//'

There is a change in the output. In order to increase accuracy, multiple test files are used. So you will see individual results per file, and a total.  The INDIVIDUAL results should be nearly identical, and that is the value we use - not the total. 

In v2.1 the data will be parsed out.

Hey mlarsen​ any news on v2.1.

Thanks.

Article Information
Author:
pvz
Creation Date:
‎09-07-2016
Views:
2160
Contributors