Silk Performance Explorer - Server Monitor option for Linux gives NULL (zero) value for all the counters

Performance Explorer executes a set of commands against the REXEC daemon running on the Linux server, so the first thing to do would be to ensure the rexec daemon is running on the Linux box.

You will find the default measures under PRE-DEFINED DATA SOURCES | CUSTOM DATA | REXEC DATA
*********

QUESTION
----------------------------------
In SilkPerformer"s Server Analysis Module, how can I use the Rexec command to get performance data of a Unix Server ?

The easiest way to do this would be to obtain a MIB file for the device or server you wish to monitor.
Some vendors supply a MIB with their products.

If you do not have a MIB file then you can make use of the Rexec command.


ANSWER
----------------------------------
Open a command prompt from Windows and get information on the device - the command will be:
Rexec "netstat -i " (include the " marks)
You will need an admin level login and password

This will return values for all of the devices on the unix box, normally something like :-

Name MTU Net/Dest Address IPkts IErrors Opkts Oerrs Collis Queue
lo0 8232 loopback locahost 39112 0 22333 0 0 0
hme0 1500 293813 0 333332 0 0 0

The hme0 is the Network Card of the Unix machine which should provide us with the information we require.
Let"s say for example we are interested in the number of InPackets (InPkts) and OutPackets (Opkts) during the loadtest.

We now need to incorporate this Rexec command into SilkPerformer"s SAM, so that we can retrieve this information and either write it to a report or create a graphical representation of it.

Open SilkPerformer
Go to RESULTS | MONITOR SERVER
Performance Explorer will open.
Click MONITOR | ADD DATA SOURCE
Choose Custom Data from the list
Then choose Rexec Data from the Custom Data tree
Choose Next
Enter the Unix server name and the port you are connecting to.
Enter an administrator"s username and password

The Rexec Measures dialog opens.

1)Enter a Measure Type " My Unix"
2)Enter a Measure Name " In Pkts"
3)Enter a Command netstat - I hme0 // please ensure you use a capital letter I for this netstat command
4)Enter a Field Index (looking at the table above it is column 5 for Inpkts - so enter 5)
5)Enter a Line to Skip - 1 //this skips the table headers and choose the hme0 data only

Although the table above shows that you would need to skip 2 rows to get to hme0, however in our command (step 3) we are specifying the name of the device explicitly therefore this will be the only one listed

Click Add when done.
Now follow steps 1 - 5 to add Opkts to our custom measures, you will need to change the Field Index
(Step 4) to a value of 7

Click Add when done.

You should now be able to obtain these measures from your server.