Utilization should be defined as busy time/total time, its dimensionless or percent. Doesn't matter if its network or CPU or disk, thats the standard definition of utilization.

Network capacity depends how you look at it, the wire speed gives you a bps (100Mbit/s, 1Gbit/s) that is easy to obtain and an absolute limit, but  not useful for calculating utilization because of protocol overhead.

Throughput in octets/s is reported by most NICs but the maximum possible octets/s does not equal the bps of the wire, due to inter packet gaps and framing of the low level prototcol. The maximum also depends upon average packet size.

So since the two measures are of different things, it does make sense that they are in different units. That way someone who is about to divide one into the other might be prompted to ask why they are different units and become more enlightened about the meaning.

It would be nice if there was a common standard way to calculate network Utilization (%busy) but I haven't seen it. I came up with a calculation that tried to include framing and gaps once. I think Orca uses it. (www.orcaware.com).

Adrian





On 9/26/05, Loukik Kudarimoti <Loukik.Kudarimoti@dante.org.uk> wrote:
Hello Gang,

It has come to our notice that the messages in v2 responses do not
specify units while giving back measurement data (ex: Bandwidth
Utilization and Capacity). Specifying such units is necessary and
messages should be enhanced to support this.

Also, there is scope for optimization in such specification. For
example, while returning back data

instead of

<perfsonar:datum timeValue="1127183580" value=" 7.620117028888889E7"
units ="bps"/>
<perfsonar:datum timeValue="1127183640" value="1.4989345286805557E7"
units ="bps"/>
<perfsonar:datum timeValue="1127183700" value=" 1.50811390446627E7" units
="bps"/>

we could specify it just once..maybe like this:

<perfsonar:data dataUnits="bps">
    <perfsonar:datum timeValue="1127183580" value=" 7.620117028888889E7"/>
    <perfsonar:datum timeValue="1127183640" value="1.4989345286805557E7"/>
    <perfsonar:datum timeValue="1127183700" value="1.50811390446627E7 "/>
</perfsonar:data>

or like this:

<perfsonar:data>
<perfsonar:units dataUnits="bps"/>
    <perfsonar:datum timeValue="1127183580" value="7.620117028888889E7 "/>
    <perfsonar:datum timeValue="1127183640" value="1.4989345286805557E7"/>
    <perfsonar:datum timeValue="1127183700" value="1.50811390446627E7"/>
</perfsonar:data>

or something else...

The second comment is: Choice of units.

After a chat with Jeff on this, I can list out two options

Jeff suggested that: Service uses the units that the data is already in
(for ex. in rrd tool, data is in octets per second. Hence service
continues to provide data in octets per second) and continues to return
data in the same units. However, the units in use should be clearly
specified using any of the above suitable methods.

An option that I would like to propose is usage of units used in common
practice. For example, bandwidth, as known to me, is more commonly
expressed in bps (and their factors).  A service should hence
*reasonably* strive to use the units that are in common practice. Either
way, specification of units using any of the suitable methods mentioned
previously is absolutely necessary.

However, it can be remarked that 'bps' is not necessarily common.
Netflow measurements might prefer to use something else which are in
common practice in the netflow domain. In such cases, specification of
units should suffice.

Nevertheless, if a service returns capacity and utilization in the same
message, it would be nice to have them both in the same units (unlike
the current case with Perfsonar prototype where capacity is bps and
utilization is octets per second)

Question here is: Which option is ideal? Should we provide capacity and
utilization in the same units in our prototype?

Loukik.







--
______________________________________________________________________

Loukik Kudarimoti
Network Engineer

DANTE - www.dante.net

Tel: +44 (0)1223 371 300
Fax: +44 (0)1223 371 371
Email: loukik.kudarimoti@dante.org.uk

City House, 126-130 Hills Road
Cambridge CB2 1PQ
UK
_____________________________________________________________________