On fös, 2008-05-30 at 16:03 -0400, Josh Luthman wrote:

Would you mind posting what you did wrong here should someone come across the same situation and want to learn from your mistake?

Of course. :) I expected the data in the rrd file to reflect directly on the data that appears on the test page in Hobbit. When the datatype is DERIVE, the number shown on the test page in Hobbit - in this particular case flexlm and the data shown is [ MATLAB : 2 ] - then this reflects directly on <last_ds> field, the value is 2.

When the dataset is changed to GAUGE, the picture changes. The value in the <last_ds> field, is always unknown. The value field seems to be an integrated or averaged function of the data shown in the test in Hobbit.

Below is an example where the measured data is [ MATLAB : 1 ] . Apparently the <value> field is 99 .

    <ds>
            <name> MATLAB </name>
            <type> GAUGE </type>
            <minimal_heartbeat> 600 </minimal_heartbeat>
            <min> 0.0000000000e+00 </min>
            <max> NaN </max>

            <!-- PDP Status -->
            <last_ds> UNKN </last_ds>
            <value> 9.9000000000e+01 </value>
            <unknown_sec> 0 </unknown_sec>
    </ds>

What I really found out, is that it does not matter. :) Somehow Hobbit graps the correct value anyway. The bottom line is that you can not expect a simple relation between the rrd data and the measured data when GAUGE datatype is used. Hobbit will graph it anyway. :)

-- Kindest Regards, Anna Jonna Ármannsdóttir, %& A: Because people read from top to bottom. Unix System Aministration, Computing Services, %& Q: Why is top posting bad? University of Iceland.