[bmwg] comments on draft-novak-bmwg-ipflow-meth-04

[Al Morton <acmorton at att.com>]

At 07:18 AM 10/9/2009, Jan Novak (janovak) wrote:
> I have just submitted a new IP Flow monitoring draft
> version.
> It addresses comments/discussions from the Stockholm meeting and
> finalises
> the changes towards the subject clarity (hopefully) which did
> not get into the draft in time for the last meeting.


Hi Jan and Benoit,

I have a few comments and suggestions on your draft, below.

Al (as a participant)

============================================================================

Some suggestions for:
> 2.2.5 Flow Export Rate
>
>    Definition:
>       Number of Flow Records that expire from the Cache (as defined by
>       the Flow Expiration term) within a time interval and are exported
>       to the Collector.

Number of Flow Records that expire from the Cache (as defined by
the Flow Expiration term) and are exported to the Collector
within a time interval.


>    Discussion:
>
>       The Flow Export Rate is measured using Flow Export data collected

s/collected/observed/

>       at the Collector by counting the exported Flow Records

during the measurement interval.

> . The
>       obtained value is an average of the instant export rates over the
>       measurement interval (see sections 5.1.3 and 5.2.3).

The value obtained is an average of the instantaneous export rates
observed during the measurement interval (see sections 5.1.3 and 5.2.3).

>       ...The lowest
>       possible measurement interval (if the requirement is to measure
>       rather instant DUT load than average load) is limited by the
>       export capabilities of the particular Flow monitoring
>       implementation.

The smallest possible measurement interval (if attempting to measure
instantaneous export rate rather than average export rate on the DUT)
is limited by the export capabilities of the particular Flow monitoring
implementation.

====================================
Suggestion for:
> 2.2.6 Cache Overflow
>
>    Definition:
>       Cache Overflow is a state of the Cache when all the available Flow
>       Record entries in the Cache are already used/occupied e.g. there is
>       no space in the Cache for new incoming Flow Records.
... and the new records will be discarded?

====================================
Suggestion for:
> 3.  Flow Monitoring Performance Metric
>
> 3.1 The Definition
>
>    Flow Monitoring Throughput
>
>    Definition:
>       The maximum Flow Export Rate the DUT can sustain without losing a
>       single Flow Record expired from the Cache and without dropping any
>       forwarded packets.

...and without dropping any packets in the forwarding plane (see Figure 1).

>       The measured Flow Export Rate MUST include the Control
>       Information.

The measured Flow Export Rate MUST include BOTH the Data stream and
the Control Information, as defined in section 2 of [RFC5470].

(maybe this last sentence should be in the Flow Export Rate Definition?)

===============================================
> 3.3 Test Concept

(this is a very helpful section now)

>    Traffic in the Figure 1 represents the test traffic sent to the DUT
>    and forwarded by the DUT. When testing devices which do not act as
>    network devices (appliances - probes, sniffers, replicators) the
>    forwarding plane is missing.

suggest:
... the forwarding plane is simply an Observation Point,
as defined in section 2 of [RFC5470].

>    The Flow monitoring enabled (see section 4.3) on the DUT (and
>    represented in the Figure 1 by the Flow Monitoring Plane) uses the
>    traffic information provided by the Forwarding Plane and configured
>    Flow Keys to create the Flow Records representing the traffic
>    forwarded by the DUT.

... forwarded (or observed) by the DUT.

<stuff omitted>
>    The Flow Monitoring Plane handles Flow Records which reflect the
>    forwarded traffic. The metric that measures the Flow monitoring
>    activity is Flow Export Rate.

...performance is Flow Export Rate.

=======================================================
> 3.4 The Measurement Procedure Overview
>
>    The measurement procedure is fully specified in sections 4, 5 and 6.
>    This section provides an overview of principles for the measurements.
>
>    The basic measurement procedure of performance characteristics of a
>    DUT with Flow monitoring enabled is a conventional non drop rate

s/non drop rate/Throughput/  (non drop rate is not a defined term in BMWG)

>    measurement with gradually increasing traffic rates until losses are
>    seen as fully specified by [RFC2544].

=============================================================
some typos below:
> 3.4.2 Forwarding Plane Performance Measurement
>
>    The Forwarding Plane (see Figure 1) performance metrics are fully
>    specified by [RFC2544] and MUST be measured accordingly. A detailed
>    traffic analysis (see the bullets below) with relation to Flow
>    monitoring must be performed prior of any [RFC2544] measurement.

s/measurement/measurements/

>    Mmainly the Flow Export Rate caused by the test traffic during an

s/Mmainly/Mainly/

>    [RFC2544] measurement MUST be known and noted.


General comment on Forwarding Plane Measurement
(possibly addressed in later sections of the draft):
The test load used for flow monitoring stress
will likely be different from
that specified in RFC2544 - we need to take care that
the Throughput benchmark will still be meaningful,
and than may mean some compromises between traffic that
stresses the Flow Monitoring Plane and the Forwarding Plane.

(stopping at the end of 3.4 for now)