#212127
September 2012
Commissioned by
Ubiquiti Networks, Inc.
Ubiquiti Networks EdgeRouter Lite
Gigabit Ethernet Routing Performance Evaluation Versus
Cisco Systems 3925 Integrated Services Router and Juniper Networks J6350
EXECUTIVE SUMMARY
THE BOTTOM LINE
Ubiquiti Networks' EdgeRouter Lite offers unprecedented price/performance
value, providing more than 100X higher performance per dollar than the Cisco
and Juniper products compared in this report. With the combination of its
broad feature set, advanced hardware platform, and disruptive price the
EdgeRouter Lite is positioned to bring enterprise-class performance to a much
wider audience.
The Ubiquiti EdgeRouter Lite:
1 Forwards 1 million packets per second of 64-byte
packets
2 Forwards at line rate (3Gbps) across three ports
with 512-byte packets or higher
3 Maintains consistent, high performance even with
firewall functionality
Ubiquiti Networks commissioned Tolly to evaluate the packet forwarding
performance of its new EdgeRouter Lite product and compare that to
enterprise-edge products from Cisco Systems and Juniper Networks. Tests
showed that the EdgeRouter Lite, priced at $99 exceeded the performance of
the competing devices that, for a basic configuration, cost in the range of
$6,500 to $7,500. See Figure 1. ...<continued on next page>
4 Provides 145X more Kpps per USD than Cisco and
205X more than Juniper
Layer 3 Bidirectional Gigabit Throughput Without Firewall
in Mbps and Kpps
As reported by Spirent TestCenter
(Higher values are better)
800
792
705 705
600
371
400
292
244 244
136
200
0
3,000 3,000
3000
64
512
Throughput (Mbps)
Throughput (Kpps)
1000
Throughput in Mbps without Firewall
Throughput in Kpps without Firewall
1000
2400
1,671
1800
1,242
1200
672
600
0
1518
Packet Size (bytes)
Ubiquiti Networks
3,000 3,000
532
249
64
512
1518
Packet Size (bytes)
Cisco Systems
Juniper Networks
Note: All products tested using three GbE ports on each DUT. Spirent throughput metrics include 12-byte Ethernet inter-frame gap (IFG). Juniper has
a fourth GbE port that was not used in this test.
Source: Tolly, June 2012
Figure 1
2012 Tolly Enterprises, LLC
Tolly.com
Page 1 of 7
Ubiquiti Networks EdgeRouter Lite vs. Cisco and Juniper
Though not part of this evaluation, the
EdgeRouter Lite supports many important
features such as VLANs, firewall, NAT, OSPF,
RIP, BGP, IPv4/v6 static/DHCP addressing,
IPsec, L2TP, PPPoE, VRRP, and more. While
the Cisco 3925 and the Juniper J6350 also
provide expansion slots for legacy WAN
interfaces such as T1/E1 and HDLC and
support additional routing functions such
as IS-IS and PIM-SIM, for customers not
requiring such less commonly used
features, the Ubiquiti Networks EdgeRouter
Lite provides high-performance, lowlatency IP routing at an extremely low
price.
#212127
Juniper J6350. The EdgeRouter Lite results
are 145X more than Cisco and 205X more
than Juniper.
Ubiquiti
Networks, Inc.
Test Results
EdgeRouter
Lite
Performance
GbE Routing
Performance
Tolly engineers tested the performance of
each solution under test with and without
a firewall functionality at three packet sizes,
64-byte, 512-byte and 1518-bytes.
Throughput was measured in Mbps and
Kpps.
Lite delivers consistently higher
throughput in both Mbps and Kpps than
the Cisco 3925 or Juniper J6350. See Figure
1.
Throughput Without Firewall in
Kpps
When results are analyzed in terms of how
many thousand packets per second (Kpps)
of throughput are delivered per dollar of
cost, Ubiquiti Networks delivers 10.10
Kpps/USD compared to between 0.12 and
0.02 for the Cisco 3925 and 0.05 for the
Tested
June 2012
The EdgeRouter Lite performed the best
across all packet sizes, forwarding over 1
million packets per second (1000.4 Kpps) of
64-byte packets, 704.9 Kpps of 512-byte
Tolly engineers tested the performance of
each solution with three packet sizes.
Engineers found the Ubiquiti EdgeRouter
Layer 3 Bidirectional Gigabit Throughput With Firewall in Mbps and Kpps
As reported by Spirent TestCenter
(Higher values are better)
Throughput in Kpps with Firewall
1,000
Throughput in Mbps with Firewall
1,000
2,400
670
700
Throughput (Mbps)
Throughput (Kpps)
850
550
371
400
289
250
100
3,000
2,850
3,000
120
140
64
512
244
1,748 1,650
1,800
1,228
1,200
672
595
600
142 134
80
0
249
64
1518
Packet Size
512
1518
Packet Size
Ubiquiti Networks
Cisco
Juniper
Note: All products tested using three GbE ports. Spirent throughput metrics include 12-byte Ethernet IFG. Juniper has a fourth GbE port that was not
used in this test.
Source: Tolly, June 2012
Figure 2
2012 Tolly Enterprises, LLC
Tolly.com
Page 2 of 7
Ubiquiti Networks EdgeRouter Lite vs. Cisco and Juniper
#212127
packets and 243.8 Kpps of 1518-byte
packets.
Juniper forwarded 532 Mbps and 249
Mbps, respectively. See Figure 1.
For the 512-byte and 1518-byte packet
sizes, the Cisco 3925 provided the same
performance as EdgeRouter Lite,
forwarding 704.9 and 243.8 Kpps,
respectively. The Cisco, however, fell short
forwarding 64-byte packets with 791.7
Kpps. See Figure 1.
For the 512-byte and 1518-byte packet size,
both Ubiquiti and Cisco demonstrated
100% line rate forwarding at 3000 Mbps.
more throughput in Kpps with a firewall
than Cisco, and 1.4X more Kpps than
Juniper. See Figure 2.
The EdgeRouter Lite forwarded over 1
million packets per second (1,000.4 Kpps)
for 64-byte packets, while Cisco and
Junipers performance suffered with the
addition of a firewall, each delivering
significantly less throughput at 120 Kpps
and 371.1 Kpps, respectively for 64-byte
packets. See Figure 2.
Juniper demonstrated significantly lower
throughput for the 512-byte and 1518-byte
packet sizes, at 1,242.2 Mbps and 1,641.1
Mbps, respectively.
The Juniper J6350 provided the lowest
performance across all three packet sizes.
Forwarding 371.1 Kpps of 64-byte packets,
291.9 Kpps of 512-byte packets and 135.8
Kpps of 1518-byte packets.
Firewall Throughput in Kpps
For the 512-byte packet size, Ubiquiti was
able to forward 669.6 Kpps, over 4.5X more
than Cisco at 139.9 Kpps and over 2X more
than Juniper at 288.6 Kpps.
To assess performance in a real-world
scenario, Tolly engineers evaluated each
solutions throughput in Kpps and Mbps
with a firewall enabled.
Throughput Without Firewall in
Mbps
For the 1518-byte packet size, the
EdgeRouter Lite again delivered higher
throughput than Cisco and Juniper.
Ubiquiti delivered 243.8 Kpps while Cisco
and Juniper forwarded 142.1 Kpps and
Ubiquitis performance was unaffected by
the addition of a firewall, still delivering the
highest throughput across all packet sizes
tested. On average, Ubiquiti delivered 3.7X
The EdgeRouter Lite forwarded 672.3 Mbps
for the 64-byte packet size, while Cisco and
Layer 3 Average Latency (Microseconds)
As reported by Spirent TestCenter
(Lower values are better)
Average Latency without Firewall
283.2
340.7
384.2
155
128.0
105.8
110
87.4
88.0
65
46.7
28.8
20
281.2
200
64
512
1518
Average Latency (Microseconds)
Average Latency (Microseconds)
200
Average Latency with Firewall
415.3
332.6
155
110
70.0
65
50.6 49.5
46.3
28.7 29.9
20
64
Packet Size
512
1518
Packet Size
Ubiquiti Networks
Cisco
Juniper
Note: All products tested using three GbE ports. Spirent throughput metrics include 12-byte Ethernet IFG. Juniper has a fourth GbE port that was not
used in this test.
Source: Tolly, June 2012
Figure 3
2012 Tolly Enterprises, LLC
Tolly.com
Page 3 of 7
Ubiquiti Networks EdgeRouter Lite vs. Cisco and Juniper
#212127
Bidirectional Gigabit Ethernet LAN Routing Price/Performance Comparison:
Kilopackets per US Dollar (USD)
64-Byte Packet Size With and Without Firewall Enabled
1
10
Note: not graphed to scale
Kilopackets per USD
0
01
0
.10
0
00
Ubiquiti Networks
Cisco
Without Firewall Enabled
Juniper
With Firewall Enabled
Bidireectional Gigabit Ethernet LAN Routing (3 Ports) Price/Perfoormance Calculations:
Kilopackets per US Dollar for 64-Byte Packet Size
Ubiquiti EdgeRouter Lite
Kilopackets
Juniperr J6350
(using 3 of 4 GbE ports)
No Firewall
Firewall Enabled
No Firewall
Firewall Enabled
No Firewall
Firewall Enabled
1000.446
1000.446
791.7
119.97
371.093
371.093
$999
CDW Price (USD)
Kilopackets per
USD
Cisco 3925 Inteegrated Services
Rou
uter
10.10
$6,4116.99
10.10
0.12
$7,5334.99
0.018
0.049
0.049
Notes: Pricing for Cisco and Juniper was obtained by Tolly engineers from CDW.com in June 2012. The Cisco and Juniper products are modular and
the prices quoted are for the most basic models found on CDW and are without WAN interfaces. The Juniper product is equipped with a fourth GbE
port that was not included in the testing. As EdgeRouter Lite was in pre-release at the time of this evaluation, its pricing information was provided by
Ubiquiti. 1 Kilopacket = 1,000 packets. Service contracts not included.
Source: Tolly, June 2012
2012 Tolly Enterprises, LLC
Figure 4
Tolly.com
Page 4 of 7
Ubiquiti Networks EdgeRouter Lite vs. Cisco and Juniper
134.1 Kpps, respectively.
With Firewall Enabled
Firewall Throughput in Mbps
Tolly engineers enabled a firewall and
measured average latency across all three
packet sizes. Ubiquitis performance was
unaffected, demonstrating low and in
some cases lower latency than without the
firewall, across all packet sizes. See Figure 3.
On average, across all packet sizes, the
Ubiquiti EdgeRouter Lite forwarded 4.3X
more Mbps than Cisco, and 3.7X more
Mbps than Juniper. See Figure 2.
The EdgeRouter Lite forwarded 672.3 Mbps
64-byte packets, while Cisco and Juniper
forwarded 80 Mbps and 249.4 Mbps,
respectively.
EdgeRouter Lite forwarded 2,850 Mbps of
512-byte packets, Cisco forwarded 595.3
Mbps and Juniper forwarded 1,228.1 Mbps.
For the 1518-byte packet size, EdgeRouter
Lite demonstrated 100% line rate at 3,000
Mbps, while Cisco and Juniper forwarded
1,748.4 and 1,650, respectively.
Latency
Tolly engineers measured the system
latency with and without a firewall
enabled. For all packet sizes, Ubiquiti
EdgeRouter Lite consistently provided the
lowest latency when compared to the Cisco
3925 and Juniper J6350.
Without Firewall
Without a firewall enabled, Ubiquitis
average latency for forwarding 64-byte
packets was 28.8 microseconds (s),
compared to 87.4 s for Cisco and 283.2 s
for Juniper. See Figure 3.
For 512-byte packet sizes, EdgeRouter Lites
average latency was 88 s. Cisco was the
next lowest at 105.8 s, and Juniper again
had the highest latency at 340.7 s.
For 1518-byte packets, EdgeRouter Lite
demonstrated the lowest latency by far at
46.7 s of latency, compared to Cisco at 128
s and Juniper at 384.2 s.
2012 Tolly Enterprises, LLC
For 64-byte packet sizes, EdgeRouter Lite
performed slightly better than its latency
without a firewall enabled, demonstrating
28.7 s of latency, compared to 29.9 s for
Cisco and 281.2 s with Juniper.
Fo r 5 1 2 - by te p a c k e t s, U b i q u i t i
demonstrated slightly higher latency, at
50.6 s while Cisco came in lowest at 49.5
s. Juniper demonstrated the highest
latency at 415.3 s.
For 1518-byte packets, EdgeRouter Lite
provided the lowest latency at 46.7 s.
Cisco was the next lowest at 70 s, and
Juniper was the highest at 332.6 s.
Price/Performance
Comparison
Kilopacket per USD
To demonstrate the price/performance
value of the Ubiquiti EdgeRouter Lite
compared to solutions from Cisco and
Juniper, Tolly engineers calculated how
many thousand packets per second (Kpps)
of throughput are delivered per dollar of
cost. The Kpps per dollar calculation uses
the highest Kpps throughput data for each
solution.
At over 10 Kilopackets per USD (both with
and without a firewall) Ubiqitis EdgeRouter
Lite represents the greatest value for
customers. Cisco and Juniper retail for
significantly more than Ubiquitis $99 USD
Tolly.com
#212127
at $6,416.99 for Cisco and $7,534.99 for
Juniper. See Figure 4 .
The Cisco 3925 ISR delivers .12 Kilopackets
per USD without a firewall, and .018
Kilopackets per USD when a firewall is
enabled.
The Juniper J6350 provides a slightly better
value than Cisco at .049 Kilopackets per
USD with and without a firewall enabled.
Test Setup &
Methodology
Test Bed Setup
The test bed consisted of the devices under
test (DUTs), connected directly to a Spirent
TestCenter SPT-2000 traffic generator
equipped with one CM-1G-D12 line card
equipped with 12 10/100/1000 Dual Media
GbE ports.
The devices under test were equipped as
detailed in Table 1. Each DUT was
connected to the Spirent traffic generator
using three GbE ports.
RFC 2544 Performance
Baseline Performance without
Firewall
To test the baseline performance of each
DUT, engineers reset the devices to their
factory default configuration. IP forwarding
was enabled, but firewall and connection
tracking features were disabled.
Three GbE ports on each DUT were
connected to the Spirent TestCenter, and
configured in a full-mesh topology - i.e.
each port sends and receives traffic from
every other port.
Page 5 of 7
Ubiquiti Networks EdgeRouter Lite vs. Cisco and Juniper
The Spirent TestCenter application running
on a Windows PC was used to configure the
parameters of the test traffic following the
methodology specified by RFC 2544. Tests
used binary search algorithm to determine
the maximum zero-loss throughput for the
packet sizes of 64, 512 and 1518 bytes, with
protocol set to UDP. Throughput was
measured in terms of megabits per second
(Mbps) and kilopackets per second (Kpps).
Last In First Out (LIFO) algorithm was used
to measure the average latency, measured
in microseconds (s).
Each test iteration was run with a 60
second duration, and each test repeated
three times to ensure repeatability of
results.
#212127
each port sends and receives traffic from
every other port.
The Spirent TestCenter application running
on a Windows PC was used to configure the
parameters of the test traffic following the
methodology specified by RFC 2544. Tests
used binary search algorithm to determine
the maximum zero-loss throughput for the
packet sizes of 64, 512 and 1518 bytes and
UDP port number 1024. Throughput was
measured in terms of Megabits per second
(Mbps) and Packets per second (pps). Last
In First Out (LIFO) algorithm was used to
measure the average latency, measured in
microseconds (s).
On each DUT, 25 stateless firewall rules
were configured in the form of Access
Control Lists (ACLs) to allow traffic
matching a particular UDP port number.
The first 24 rules do not match the test
traffic, while the 25th rule is configured to
allow traffic with the UDP port number
1024, matching that of the traffic. Each
packet of the test traffic gets processed by
each of the 25 ACLs defined in the firewall
component of the DUT.
Each test iteration was run with a 60
second duration, and each test was
repeated three times to ensure result
consistency.
Performance with Firewall
To test the performance of each DUT with
firewall turned on, engineers reset the
devices to their factor y default
configuration. IP forwarding and firewall
features were enabled, but connection
tracking features were disabled.
Three GbE ports on each DUT were
connected to the Spirent TestCenter, and
configured in a full-mesh topology - i.e.
Test Equipment Summary
The Tolly Group gratefully acknowledges the providers
of test equipment/software used in this project.
Vendor
Product
Web
Spirent
TestCenter SPT-2000
www.spirent.com
Device Under Test - Version Information
Device Under Test
Details
Software Version
Ubiquiti EdgeRouter Lite
Three 10/100/1000 Mbps Ethernet ports
Dual-core 500 MHz MIPS64 processor with
hardware acceleration for packet processing
Version 0.9.5
Cisco 3925 Integrated Services
Router
Model C3900-SPE100/K9
Three 10/100/1000 Mbps Ethernet ports
Cisco IOS Software,
C3900 Software (C3900-UNIVERSALK9-M)
Version 15.1(4)M2, RELEASE SOFTWARE (fc1)
Juniper J6350 Services Router
Four 10/100/1000 Mbps Ethernet ports
JUNOS Software Release [9.2-20090320.0]
Table 1
Source: Tolly, June 2012
2012 Tolly Enterprises, LLC
Tolly.com
Page 6 of 7
Ubiquiti Networks EdgeRouter Lite vs. Cisco and Juniper
About Tolly
The Tolly Group companies have been
delivering worldclass IT services for
more than 20 years. Tolly is a leading
global provider of thirdparty validation
services for vendors of IT products,
components and services.
You can reach the company by email at
[email protected], or by telephone at +1
561.391.5610.
#212127
Interaction with Competitors
In accordance with Tollys Fair Testing Charter, Tolly
personnel invited representatives from the
competing companies to review the testing.
Juniper declined to participate and Cisco did not
respond.
For more information on the Tolly Fair Testing Charter, visit:
http://www.tolly.com/FTC.aspx
Visit Tolly on the Internet at:
http://www.tolly.com
Terms of Usage
This document is provided, freeofcharge, to help you understand whether a given product, technology or service merits additional investigation
for your particular needs. Any decision to purchase a product must be based on your own assessment of suitability based on your needs. The
document should never be used as a substitute for advice from a qualified IT or business professional. This evaluation was focused on
illustrating specific features and/or performance of the product(s) and was conducted under controlled, laboratory conditions. Certain tests may
have been tailored to reflect performance under ideal conditions; performance may vary under realworld conditions. Users should run tests
based on their own realworld scenarios to validate performance for their own networks.
Reasonable efforts were made to ensure the accuracy of the data contained herein but errors and/or oversights can occur. The test/audit
documented herein may also rely on various test tools the accuracy of which is beyond our control. Furthermore, the document relies on certain
representations by the sponsor that are beyond our control to verify. Among these is that the software/hardware tested is production or
production track and is, or will be, available in equivalent or better form to commercial customers. Accordingly, this document is provided "as
is", and Tolly Enterprises, LLC (Tolly) gives no warranty, representation or undertaking, whether express or implied, and accepts no legal
responsibility, whether direct or indirect, for the accuracy, completeness, usefulness or suitability of any information contained herein. By
reviewing this document, you agree that your use of any information contained herein is at your own risk, and you accept all risks and
responsibility for losses, damages, costs and other consequences resulting directly or indirectly from any information or material available on it.
Tolly is not responsible for, and you agree to hold Tolly and its related affiliates harmless from any loss, harm, injury or damage resulting from or
arising out of your use of or reliance on any of the information provided herein.
Tolly makes no claim as to whether any product or company described herein is suitable for investment. You should obtain your own
independent professional advice, whether legal, accounting or otherwise, before proceeding with any investment or project related to any
information, products or companies described herein. When foreign translations exist, the English document is considered authoritative. To
assure accuracy, only use documents downloaded directly from Tolly.com.
No part of any document may be reproduced, in whole or in part, without the specific written permission of Tolly. All trademarks used in the
document are owned by their respective owners. You agree not to use any trademark in or as the whole or part of your own trademarks in
connection with any activities, products or services which are not ours, or in a manner which may be confusing, misleading or deceptive or in a
manner that disparages us or our information, projects or developments.
212127ul1mptwt20120828VerM
2012 Tolly Enterprises, LLC
Tolly.com
Page 7 of 7
