Transcript PPT Version
Benchmarking Methodology
for Routing Experts 101
Al
Morton
bmwg chair
Crash Course Syllabus
Who we are
What we do
What we’ve accomplished
What we’re working on that might
interest you
Where we’re headed…
2
Who are we?
3
What we do in BMWG
Attributes of a BENCHMARK
a standardized test that serves as a basis for
performance evaluation and comparison.
Unambiguous method, repeatable
Reveal valuable characteristics to users and
provide insights to improve internetworking
technology
that permit cross-vendor comparison with
externally observable measurements
4
Lab Test Configurations:
DUT=Device Under Test
RFC 2544
Benchmarking Methodology
March 1999
+------------+
|
|
+------------| tester
|<-------------+
|
|
|
|
|
+------------+
|
|
|
|
+------------+
|
|
|
|
|
+----------->|
DUT
|--------------+
|
|
+------------+
Figure 1
+--------+
+------------+
+----------+
|
|
|
|
|
|
| sender |-------->|
DUT
|--------->| receiver |
|
|
|
|
|
|
+--------+
+------------+
+----------+
Figure 2
5
What BMWG’s
Benchmarks are NOT:
Functional tests
Evaluated with PASS/FAIL criteria
Performed on live networks, or performed
with live network traffic
– Assigned IPv4 address space for Benchmarking
198.18.0.0/15 - This block has been allocated for
use in benchmark tests of network interconnect
devices. Its use is documented in [RFC2544].
– About to request dedicated IPv6 space
– Want to talk Live Network Measurements?
PMOL meets tomorrow..
6
Benchmarking Past
RFC 1242, Benchmarking Terminology
for Network Interconnection Devices
LAN Switches, Firewalls, ATM, F.Relay
IP Multicast
Single Router OSPF Control Plane
Convergence
BGP Control Plane Converge. (Terms)
Resource Reservation (Terms)
7
Benchmarking Present
IGP Dataplane Convergence Time
– More than the usual number of DISCUSSs
Accelerated Stress Benchmarking
Protection Switching Technologies
– Methodology for MPLS-FRR
IPv6 at Publication Requested
IPsec devices
8
Work Proposal Summary Matrix
Work Area
>
Criteria \/
Proposal
In Scope of
Charter?
(acm)
Draft(s)
SIP
Mcast VPN
Scalability
LDP
Converg
MPLS
BFD
WLAN
switch
RSVPTE
Y
Y
Y
Y
Y
Y
Y
Y
Y
No
Overlap
w/802.11
T
Y
1+
Draft
gives
motivat.
1+1
Draft
gives
motivat.
Y
1+1
Sig. Support
at meetings
Sig. Support
on List
Testing to
support Dev?
Y
1+
Y
1+1
Y
Interest
when first
proposed
9
How can You help the
good old BMWG?
Review Drafts!
Early Review >> Late Review
Establish a more formal process?
10