Transcript Shenick
Use of Traffic Emulation in
Interoperability Testing
Mark Lambe
Product Marketing Manager
Shenick Network Systems
Shenick Network Systems
1
www.shenick.com
Overview
Telepresence is the delivery of high performance voice and video which plays
a critical role in service differentiation.
A clear understanding of quality and performance limitations under a variety
of operational conditions, including interoperability, is imperative to success.
Today’s discussion covers the benefits of adopting real-time, per end point
user emulation and analysis to deliver a highly credible and pragmatic quality
test program.
Shenick Network Systems
2
www.shenick.com
Define Interoperability
Interoperability means different things to different people at each stage
The Telepresence deployment will function correctly i.e. each component in the
system can communicate and operate with each other.
The Telepresence deployment operates correctly on the proposed infrastructure with
minimum disruption.
The Telepresence deployment will operate alongside other services (video, voice,
data) without impacting QoE/QoS.
The use of different Vendor end-point systems/components.
Telepresence has unified the way we communicate
Therefore a unified test solution covering all aspects of Interoperability is required.
Vendor
Interoperability
Shenick Network Systems
3
www.shenick.com
System Functionality
Emulate the most realistic components, therefore giving the most accurate results
A Complete test process - examine performance from end-point boot-up to call ended
Performance test configuration/management server responses
Ability of end-points to connect to management servers
Time to download end-point firmware settings, upgrades, etc
Time to register through to time to join calls
Participate in real/live Telepresence calls
Measure actual video/audio quality
Security Testing
Attempt registration with un-assigned MAC addresses
Shenick Network Systems
4
www.shenick.com
Operational Functionality
Preliminary testing
Network Performance
Connectability – End-points can communicate with management servers
–
What impact will firewalls have on Telepresence traffic flows coming from unsecure to secure networks
Test Telepresence management servers under varying network and load conditions
–
If a DDoS attack occurs at the firewall, what impact has this on the Telepresence conference
Hardware Performance
Determine latency measurements
CPU Processing performance limitations
Quality
Test end-to-end video/audio quality using perceptive analysis tools
Test bandwidth performance use various video grades (low, medium, high resolutions)
Classification Testing
QoS Testing
Are Telepresence flows identified and handled correctly
False negative tests – test with flows of similar nature ( e.g. P2P)
False positive tests – a Telepresence flow is identified and handled incorrectly
Shenick Network Systems
5
www.shenick.com
Meeting Functionality
Flow Control Testing
Video Switching – test video switching policies
who has the floor testing
Who is shouting the loudest to win control of the screen
Sporadic Interruption through dynamic behaviour or un-muting of a participant
Audio – test the capability to deliver every end-points audio flows
a heated discussion
Security feature Testing
Assess muted room appearance functionality
Attempt to join conference with unknown/un-invited
numbers
Shenick Network Systems
6
www.shenick.com
Emulated Telepresence Activity
Per Flow Emulation & Performance Analysis
Emulate real stateful end-points, with real
user activity
Register with Telepresence Management
Servers
Participate in real calls
Accurate representations of real world
devices
Per end-point unique MAC address assignment
Static and Dynamic IP address allocation
IPv4 and IPv6 capabilities
diversifEye’s flexibility enables users configure
device properties easily, matching those in the
Telepresence device management database
Shenick Network Systems
7
www.shenick.com
Real User Activity
Participation with as close to real activity is key in performance testing
Telepresence systems and deployments, dynamic behavior adds a level of
un-predictability to the likely results outcome
Unique video and audio per emulated Telepresence
end-point
Ability to vary codecs for video and audio
Auxiliary inputs e.g. Videos, Presentations
Dynamic behaviour to emulated end-points
Loudness (peak level, duration, etc)
Mute / Un-mute calls in live tests
Shenick’s diversifEye enables users set
various levels of loudness per emulated
Te l e prese nce e nd -poi nts
Mute / Un-mute calls during live tests
Shenick Network Systems
8
www.shenick.com
Sample Performance Analysis
Measure performance on each and every individual emulated Telepresence end-point
In multi-point calls determine how varying network architectures/settings impact call
performance
Sample Performance Metrics per unique Telepresence end-point
Video Quality – Use perceptive analysis tools to deliver MOS scores
Audio Quality – Use perceptive analysis tools to deliver MOS scores
RTP Bandwidth – bits or packets per second
SSRC rate – Sent / received
RTP packet metrics – Dropped/Duplicate
RTP Video frame count / Video frame jitter
Out Calls Attempted / Established / Rejected
Shenick’s diversifEye delivers real-time
results analysis, on a per emulated
Telepresence end-point basis. Examine
during live tests the impact of different
traffic volumes and application types
Shenick Network Systems
9
www.shenick.com
Open Test Strategy
Telepresence will run in parallel to other services such as voice, video and data.
It’s essential to determine the impact Telepresence deployments have on existing services
It’s also important to understand how Telepresence is impacted by adding other traffic sources
and varying these flows loads.
Test with both legitimate and illegal traffic flows, spam, virus, worms, DDOS, etc
In the following open test scenario
What’s a relevant / realistic MOS score for Telepresence in live networks?
Who would you perceive as the happiest customer?
Telepresence MOS
(video / audio)
HTTP Downloads
(seconds)
P2P traffic flows
(present)
VoIP MoS
Email send time
(ms)
Enterprise A
4.5 / 4.5
25
3
950
Enterprise B
3.9 / 4.0
1
4.6
5,000
Enterprise C
4.2 / 4.4
10
4.5
2,000
Shenick Network Systems
10
www.shenick.com
diversifEye
diversifEye emulates stateful clients with real world functionality
and features for Telepresence
diversifEye provides stateful traffic emulation of many common
applications and traffic flow types:
IGMP / MLD, VoD (RTSP), VoIP, Dual Hosted VoIP, RTP, PPPoE, HTTP, P2P,
VLAN, DHCP, P2P TCP Playback, SMTP, POP3, FTP, TWAMP, SSL/TLS
Mix legal and illegal flows: include DDoS type attacks, emails
with vrius, spam attacks, etc
diversifEye may support a mix of client and/or server applications
using IPv4 and/or IPv6 on a single test interface port.
Support for IPv6 transition protocols: Dual-Stack Lite, 6rd / 6to4
diversifEye is used in ‘Per flow’ application emulation and
performance analysis of multiple environments including : xDSL,
xPON, IMS, CABLE, WiMAX & LTE
Capture and replay TCP, UDP (files >1Gb)
Fully compliant TCP stack
Shenick Network Systems
11
www.shenick.com
servicEye Spider
Remote Network Test with centralized management and control
Tap into the power of a sophisticated lab tester from multiple
remote network sites
Shenick Network Systems
12
www.shenick.com
Thank You!
Questions & Answers
email : [email protected]
Shenick Network Systems
13
www.shenick.com