Napatech
recently announced the availability of unique features that provide
visibility into tunnels for network monitoring and analysis of
value-added services running over IP, such as VOIP, VPN and mobile
data. Napatech's network adapters have the unique capability to
intelligently distribute 10Gbps flows in real-time to up to 32 CPU
cores based on Ethernet, IP, MPLS, layer 4 and now GTP, GRE and SCTP
tunnel information.
As more services transition to 1 Gbps and 10 Gbps IP networks, the
importance of network monitoring and analysis is increasing - but so
are the demands on intelligence and processing capabilities. By using
Napatech's family of network adapters for real-time network analysis,
Napatech's OEM customers can build high performance network appliances
providing new levels of intelligence and throughput performance.
"Tunneling is often used in networking to transparently and
securely transport data over a Wide Area Network. This is particularly
true for value-added services that rely on IP. However, other network
adapters used for network monitoring and analysis can only recognize
the Ethernet frame and IP header and not the tunneling protocol or
encapsulation information. Napatech is the first network adapter vendor
to provide visibility into these tunnels for network monitoring and
analysis purposes. This combined with our unique ability to
intelligently distribute 10 Gbps flows to up to 32 CPU cores in
real-time makes it possible to fully monitor and analyze VOIP, VPNs and
Mobile Data Networks", says Erik Norup, President, Napatech, Inc.
Napatech's network adapters provide sustained full line-rate 1Gbps
and 10Gbps throughput per port both on reception and transmission of
packets. The adapter recognizes layer 2 through 4 protocols including
SCTP, GTP and GRE tunneling protocols, as well as a number of
encapsulation methods, such as VLANs, multiple MPLS labels and Cisco
ISL. Napatech network adapters can also compute 17 different types of
hash-keys, which can be dynamically selected based on the recognized
protocol information. Using the computed hash-key or an equal
distribution method, flows can be distributed across 32 CPU cores,
allowing parallel processing. This allows the application to fully
utilize available CPU cores.