`FOR THE EASTERN DISTRICT OF TEXAS
`TYLER DIVISION
`
`FatPipe, Inc.,
`
`Plaintiff/Counter-Defendant,
`
`v.
`
`Talari Networks, Inc.,
`
`Defendant/Counterclaimant.
`
`Civil Action No.: 6:15-cv-0458-RWS
`
`JURY TRIAL DEMANDED
`
`PLAINTIFF FATPIPE, INC.’S PATENT RULE 3-1 DISCLOSURE OF
`ASSERTED CLAIMS AND INFRINGEMENT CONTENTIONS
`
`In accordance with the Court’s Order and P.R. 3-1, Plaintiff FatPipe, Inc. (“FatPipe”) is
`
`serving on Defendant Talari Networks, Inc.. (“Talari") its 3-1 Disclosure of Asserted Claims and
`
`Infringement Contentions ("Infringement Contentions"). FatPipe’s Infringement Contentions are
`
`based upon information available to FatPipe as of the date hereof without the benefit of
`
`discovery in this litigation. Because FatPipe has not yet received documents, source code,
`
`deposition testimony or other discovery relevant to its Infringement Contentions from Defendant,
`
`and because FatPipe's investigation is ongoing, FatPipe reserves the right to supplement and/or
`
`modify its Infringement Contentions to the full extent permitted under Patent Rule 3-6 and this
`
`Court's Orders.
`
`I. Patent Rule 3-1(a) - Identification of Asserted Claims
`
`With respect to U.S. Patent No. 6,775,235 ("'235 Patent"), FatPipe asserts that Talari
`
`infringes the following claims of the asserted patent directly, contributorily and/or by
`
`inducement: Claims 4, 5, 7-15 and 19.
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`With respect to U.S. Patent No. 7,406,048 ("'048 Patent"), FatPipe asserts that Talari
`
`infringes the following claims of the asserted patent directly, contributorily and/or by
`
`inducement: Claims 1-24.
`
`II. Patent Rule 3-1(b) - Identification of Accused Instrumentalities
`
`For the purposes of Patent Rule 3-1(b), FatPipe asserts that each of the asserted claims for
`
`the ‘235 and ‘048 patents is infringed by the Talari Mercury family of appliances including,
`
`without limitation, at least the T510, T730, T750, T860, T3010, T5000 and T5200 appliances
`
`and the VT500 virtual appliance. FatPipe reserves the right to amend this identification to the
`
`full extent permitted under Patent Rule 3-6 and this Court's Orders.
`
`III. Patent Rule 3-1(c) - Claim Charts
`
`Appendix I, attached hereto and incorporated herein in its entirety, provides preliminary
`
`charts identifying specifically where each element of the asserted claims is found in the Talari’s
`
`accused instrumentalities. FatPipe’s investigation and analysis of the accused instrumentalities is
`
`based upon information made publicly available by Talari and FatPipe’s own investigations. As
`
`such, the charts appended and incorporated as Appendix I set forth FatPipe’s current contentions
`
`with respect to where each element of the asserted claims is found in the accused
`
`instrumentalities. All citations to evidence in FatPipe’s Infringement Contentions are illustrative
`
`only and shall in no way limit FatPipe’s reliance on additional evidence obtained in discovery in
`
`support of its Infringement Contentions. To the fullest extent permitted under Patent Rule 3-6
`
`and this Court's Orders, FatPipe reserves the right to amend and/or supplement its Infringement
`
`Contentions for any of the following reasons:
`
`-1-
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`(i) FatPipe’'s positions regarding infringement of specific claims will depend on how
`
`those claims are construed by the Court. Because claim construction in this case has not yet
`
`occurred, FatPipe cannot take a final position on the bases for infringement of the asserted claims.
`
`(ii) While FatPipe’s investigation and analysis of Talari’s infringement is based upon
`
`information made publicly available by Talari and FatPipe’s own investigations, additional
`
`discovery from Defendant is necessary before FatPipe can take final positions on the bases for
`
`infringement of the asserted claims.
`
`(iii) Many of the claim elements in the claims of the asserted patents are performed by
`
`Talari’s software. Accordingly, pursuant to Paragraph 3 of the Court’s Discovery Order, FatPipe
`
`reserves the right to amend and/or supplement its Infringement Contention after Talari produces
`
`its software code.
`
`IV. Patent Rule 3-1(d) - Literal Infringement/Doctrine of Equivalents
`
`For purposes of Patent Rule 3-1(d), each element of each asserted claim for the ‘235 and
`
`‘048 Patents is considered to be literally present and present under the Doctrine of Equivalents
`
`within each of the accused instrumentalities. To the fullest extent permitted under patent Rule 3-
`
`6 and this Court's Orders, FatPipe reserves the right to amend and/or supplement this disclosure
`
`for any of the following reasons:
`
`(i) FatPipe’s positions regarding infringement of specific claims will depend on how
`
`those claims are construed by the Court. Because claim construction in this case has not yet
`
`occurred, FatPipe cannot take a final position on the bases for infringement of the asserted claims
`
`under the Doctrine of Equivalents.
`
`(ii) While FatPipe’s investigation and analysis of Defendant's infringement is based upon
`
`information made publicly available by Defendant and FatPipe’s own investigations, additional
`
`-2-
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`discovery from Defendant is necessary before FatPipe can take final positions on the bases for
`
`infringement of the asserted claims under the Doctrine of Equivalents.
`
`(iii) Talari has not taken a position regarding whether any elements of the asserted claims
`
`are not literally present, as well as the basis for such a position. As such FatPipe reserves the
`
`right to amend and/or supplement this disclosure after Talari discloses the basis for its contention
`
`that it does not infringe the asserted patents.
`
`(iv) Many of the claim elements in the claims of the asserted patents are performed by
`
`Talari’s software. Accordingly, pursuant to Paragraph 3 of the Court’s Discovery Order, FatPipe
`
`reserves the right to amend and/or supplement its Infringement Contention after Talari produces
`
`its software code.
`
`V. Patent Rule 3-1(e) - Priority Dates
`
`For purposes of Patent Rule 3-1(e), claims 4 and 19 of the ‘235 Patent are believed to be
`
`entitled to a priority date of December 29, 2000. Claims 5 and 7-15 of the ‘235 patent and all the
`
`asserted claims of the ‘048 Patent are believed to be entitled to a priority date of February 8,
`
`2002. However, FatPipe’s positions regarding the priority date of specific claims will depend on
`
`how those claims are construed by the Court. Accordingly, FatPipe reserves the right to amend
`
`and/or supplement its positions regarding the priority date of the asserted claims to the fullest
`
`extent permitted under Patent Rule 3-6 and this Court's Orders
`
`VI. Patent Rule 3-1(f)
`
`For purposes of Patent Rule 3-1(f), the following FatPipe products are believed to
`
`practice the inventions claimed in the asserted claims: FatPipe WARP, FatPipe MPVPN, and
`
`FatPipe IPVPN.
`
`-3-
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`Dated: October 15, 2015
`
`By:
`
`/s/ Timothy J. Carroll
`Timothy J. Carroll
`
`Timothy J. Carroll (Lead Attorney)
`Illinois State Bar No.: 6269515
`Steven M. Lubezny
`Illinois State Bar No.: 6275394
`PERKINS COIE LLP
`131 S. Dearborn Street, Suite 1700
`Chicago, IL 60603-5559
`Phone: (312) 324-8400
`Fax: (312) 324-9400
`tcarroll@perkinscoie.com
`slubenzy@perkinscoie.com
`
`Kelly D. Hine
`Texas Bar No.: 24002290
`PERKINS COIE LLP
`500 N. Akard Street, Suite 3300
`Dallas, TX 75201
`Phone: (214) 965-7700
`Fax: (214) 965-7799
`khine@perkinscoie.com
`
`Attorneys for Plaintiff/Counter-Defendant FatPipe, Inc.
`
`-4-
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`No. Claim Language
`
`Accused Instrumentalities
`
`APPENDIX I
`
`U.S. Patent No. 6,775,235
`
`4
`
`(a) A controller which controls access
`to multiple networks in a parallel
`network configuration, suitable
`networks comprising Internet-based
`networks and private networks from
`at least one more provider, in
`combination, the controller
`comprising:
`
`As shown below, the accused Talari devices are controllers that control access to multiple
`networks in a parallel network configuration, suitable networks comprising Internet-based
`networks and private networks from at least one more provider, in combination.
`
`Talari APN | Theory of Operation – Technical Paper at p. 30
`
`(b) a site interface connecting the
`controller to a site;
`
`The accused Talari devices provide a site interface connecting the controller to a site. For
`example, each accused instrumentality includes at least one Ethernet port that connects the
`controller to a LAN. See Talari Appliances Datasheet.
`
`Below is one example configuration illustrated in Talari’s documentation. Although the
`image is taken from a guide for the T510 device, all of the accused devices can be configured
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`in a similar manner.
`
`(c) at least two network interfaces
`which send packets toward the
`networks; and
`
`T510 Getting Started Guide version 1.1 at p. 5
`
`The accused Talari devices provide at least two network interfaces which send packets toward
`the networks. For example, each accused devices includes multiple Ethernet ports that can be
`configured to send packets towards the Internet-based and private networks. See Talari
`Appliances Datasheet. One of these interfaces can be configured to be the WAN interface
`associated with a private network and another configured to be the WAN interface associated
`with the Internet.
`
`Below is an exemplary illustration showing the interfaces for the T510 Talari device.
`Although the other accused devices may have a different configuration of interfaces, each
`accused devices includes at least two network interfaces which send packets towards the
`
`- 2 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`networks.
`
`T510 APN Appliance Hardware Guide
`
`(d) a packet path selector which
`selects between network interfaces on
`a per-packet basis according to at
`least: a destination of the packet, an
`optional presence of alternate paths to
`that destination, and at least one
`specified criterion for selecting
`
`As described below, the accused Talari devices contain a packet path selector that selects
`network interfaces on a per-packet basis according to at least a destination of the packet, an
`optional presence of alternate paths to that destination, and at least one specified criterion for
`selecting between alternate paths when such alternate paths are present.
`
`Talari’s “theory of operation” documentation explains that a packet received from the LAN
`(site) may be directed to a Conduit Service, Internet Network Service, Intranet Network
`
`- 3 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`between alternate paths when such
`alternate paths are present;
`
`Service, or Passthrough Network service:
`
`“Within the APN, routes are the binding of IP networks to a particular network
`service, such as for example, a Conduit between two sites. When an APN Appliances
`Ethernet interface receives a packet, the packet is evaluated against the set of routes
`available and an appropriate route is selected. The route selected directs the packet to
`a specified service. The service then directs the traffic to its destination whether it is
`the local network, Conduit Service, Intranet Service, Internet Service or Passthrough.”
`Talari APN – Theory of Operation” at p. 44.
`
`Talari defines the Conduit service as follows:
`
`“The Conduit service is a logical combination of one or more paths, and is the typical
`mode for enterprise site-to-site Intranet traffic, utilizing the full value of the Talari’s
`Adaptive Private Networking. In this mode, depending on configuration, the traffic is
`actively managed across multiple WAN links to create an end-to-end Conduit.” See,
`e.g., T750 Hardware Guide at Glossary.
`
`Packets are directed between network interfaces, on a per-packet basis, based on a destination
`of the packet, an optional presence of alternate paths to that destination, and at least one
`specified criterion:
`
`-
`
`“Talari APN Appliances (APNA) have the ability to determine which path to send
`traffic over on a per--
`packet basis. •
`
`This is not necessarily the most expensive link or the one with the most bandwidth.
`
`There are a variety of factors that a Talari APNA considers when determining which
`path to put a packet on as it adapts to the conditions of the WAN
`
`When a Talari APNA receives a packet from the Conduit (a WAN Egress packet), it
`calculates the quality of the path from which it was received, based on one-way trip
`latency, loss, bandwidth, and jitter. When the Talari APNA has a packet to send on the
`Conduit (a WAN Ingress packet), the Talari APNA checks the quality of all paths and
`
`- 4 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`sends the packet over the best available path.
`
`All paths within the Talari APNA are classified with a use state of GOOD, BAD, or
`DEAD. A path is declared DEAD if there are no Conduit packets received on the path
`for 1.5 seconds and the Talari APNA will not attempt to use a DEAD path.
`
`To declare a path BAD is a bit more complicated and the Talari APNA uses a
`sophisticated algorithm to calculate the loss, latency, and jitter before making this
`determination…
`
`For all GOOD paths, the Talari APNA uses this same method of calculation to
`determine which of the GOOD paths is declared the best path. This is done on a per-
`packet basis, not measured by amount of time. One packet may be sent using an
`Internet path and the next may be sent via MPLS, unless Persistent Path is configured
`in the Talari APNA (this configuration option will be explained below).
`
`Once the Talari APNA determines the best path, traffic may be influenced further
`depending on specific options enabled within the Talari APNA config. The four
`options for mode of transmission are, Duplicate Path, Load Balancing, Persistent Path,
`and Override Service.
`
`… W
`
`ith the Load Balancing option, traffic is put on the best path until that path runs out
`of available bandwidth. The additional traffic will then spill over to the next best path.
`For example with two WAN Links, one containing 1.5M of bandwidth 10ms of
`latency, and the other with 5M of bandwidth but 20ms latency detected. Path 1 will
`be considered the best path by the Talari APNA and used initially when trying to push
`4M of traffic. Talari APNA will fill this pipe with 1.5M, but after 5 ms Talari APNA
`will spill the remaining traffic over onto path 2. Once path 1 has the bandwidth
`necessary, if it is still considered the better path due to the lower latency value, traffic
`will spill back from path 2 to path 1 seamlessly. Load Balancing is the default option
`when creating rules in the Talari APNA config.
`
`- 5 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`With Persistent Path, the latency is allowed to build for 50ms on the best path before
`traffic jumps onto the next best path. Traffic will begin and continue on the best path
`until the latency has built up to the point where the second path is now considered
`better than the first. Traffic will continue on path 2 until the latency on path 1 scales
`back down to a value below that of path 2. Once this level has been reached, the
`Talari APNA will again consider path 1 to be more reliable than path 2. Persistent
`path should be configured for real-time and interactive traffic only, not for bulk data
`unless the exact amount of bandwidth per application is known and available.”
`“Talari APN -Best Path Determination” at pp. 1-2
`
`Talari APN – Theory of Operation” at p. 47.
`
`“The Talari APN Appliance continuously manages the network application traffic
`load based on custom and traffic monitoring. Utilizing a rules engine, the APN
`Appliance is able to identify differing application traffic types that have configurable
`requirements, such as those for applications. The rules engine associates network
`applications with per-rule properties such as QoS with 16 separate classes of services,
`queue depth allocations, and tunable proper/requirements that determine the intended
`
`- 6 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`behavior of session traffic as it travels across the WAN. “Talari APN – Theory of
`Operation at p. 25
`
`Talari Appliances Datasheet
`
`The Internet and Intranet Network Services also perform load balancing and link bandwidth
`management:
`
`“The Internet Service is for traffic between an enterprise user and sites on the public
`Internet. The Talari APN does not encapsulate traffic of this type. During times of
`congestion, the Talari APN actively manages bandwidth by shaping and grooming
`Internet traffic relative to other services use of the WAN Link bandwidth as per the
`configuration established by the administrator. In addition, the APN is able to provide
`session load balancing between multiple WAN Links. The load balancing is only at
`the source IP and destination IP level, so all sessions to a particular Internet server
`from a particular IP host at the premise will utilize only one WAN Link. The APN
`service also supports an active/standby WAN Link fail- over function for increased
`high-availability. The APN does not provide any security for the Internet service, so
`there should be an external firewall. It is not possible to configure the Internet Service
`on an untrusted Ethernet segment. If attempted, the configuration facility will reject
`
`- 7 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`the configuration as invalid.
`
`… T
`
`he Intranet Service is for any portion of enterprise site-to-site traffic that has not
`been defined for transmission across the APNs Conduit Service. As with Internet
`Service traffic, it remains un-encapsulated and, the APN manages bandwidth by rate
`limiting this traffic to other service times during times of contention and congestion.
`The APN is able to support multiple Intranet Services and is capable of failing over
`from other network services as a backup for higher network availability. An example
`may be using an Intranet Service to access an MPLS network to transmit traffic to
`remote locations that do not have an APN Appliance.” Talari APN – Theory of
`Operation at p. 32.
`
`As described below, the accused Talari devices receive a packet through the site (LAN)
`interface and send the packet through the network (WAN) interface selected by the packet
`path selector.
`
`“The user has the ability to configure specific ports on a per WAN link basis. When
`the user configures this option the Talari appliance will use this port as the source port
`for conduit WAN “ingress” traffic. From a Talari perspective, conduit WAN ingress
`traffic is traffic that is from the LAN to WAN or traffic that is from the local
`Talari appliance to the conduit for a destination Talari appliance.” UDP Port
`Configuration Options App Note at p. 4
`
`(e) wherein the controller receives a
`packet through the site interface and
`sends the packet through the network
`interface that was selected by the
`packet path selector.
`
`- 8 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`Talari APN Glossary at p. 8
`
`5
`
`(a) A method for combining
`connections for access to multiple
`parallel disparate networks, the
`method comprising the steps of:
`
`As shown below, the accused Talari devices provide a method for combining connections for
`access to multiple parallel disparate networks
`
`Talari APN | Theory of Operation – Technical Paper at p. 30
`
`(b) obtaining at least two known
`location address ranges which have
`associated networks;
`
`As described below, the accused Talari devices obtain at least two know address ranges which
`have associated networks.
`
`- 9 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`(c) obtaining topology information
`which specifies associated networks
`that provide, when working,
`connectivity between a current
`location and at least one destination
`location;
`
`Adaptive Private Networking Configuration Editor User’s Guide at p. 20
`
`The accused Talari devices obtain topology information which specifies associated networks
`that provide, when working, connectivity between a current location and at least one
`destination location.
`
`The topology information is obtained during configuration:
`
`Talari APN | Implementation Guide, User Manual at p. 11
`
`- 10 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`The site defines each LAN that contains specific destination hosts. The Connections define
`the interconnections between sites. Together these define the topology information. For
`instance, Talari’s documentation discloses an example of a site in New York and a site in
`London
`
`Talari APN | Implementation Guide, User Manual at p. 12
`
`Talaris’ documentation describes creating connectivity between the sites:
`
`- 11 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`Talari APN | Implementation Guide, User Manual at p. 18
`
`(d) receiving at the current location a
`packet which identifies a particular
`destination location by specifying a
`destination address for the destination
`
`As described below, the accused Talari devices receive at the current location a packet which
`identifies a particular destination location by specifying a destination address for the
`destination location;
`
`“Within the APN, routes are the binding of IP networks to a particular network
`
`- 12 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`location;
`
`service, such as for example, a Conduit between two sites. When an APN Appliances
`Ethernet interface receives a packet, the packet is evaluated against the set of routes
`available and an appropriate route is selected. The route selected directs the packet to
`a specified service. The service then directs the traffic to its destination whether it is
`the local network, Conduit Service, Intranet Service, Internet Service or Passthrough.”
`Talari APN – Theory of Operation” at p. 44. See also discussion in 4(d) above.
`
`(e) determining whether the
`destination address lies within a
`known location address range;
`
`The accused Talari devices determine whether the destination address lies within a known
`location address range. As discussed in 5b above, location address ranges are known. When
`a packet is received on a port, it is typically routed to an outgoing port. This routing
`necessarily makes a determination if the destination address of the IP packet lies within the
`known location address range(s).
`
`As described below, the accused Talari devices select a network path from among paths to
`disparate associated networks, said networks being in parallel at the current location, each of
`said networks specified in the topology information as capable of providing connectivity
`between the current location and the destination location;
`
`(f) selecting a network path from
`among paths to disparate associated
`networks, said networks being in
`parallel at the current location, each
`of said networks specified in the
`topology information as capable of
`providing connectivity between the
`current location and the destination
`location;
`
`- 13 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`Talari APN | Theory of Operation – Technical Paper at p. 30
`
`“When a Talari APNA receives a packet from the Conduit (a WAN Egress packet), it
`calculates the quality of the path from which it was received, based on one-way trip
`latency, loss, bandwidth, and jitter. When the Talari APNA has a packet to send on the
`Conduit (a WAN Ingress packet), the Talari APNA checks the quality of all paths and
`sends the packet over the best available path.
`
`All paths within the Talari APNA are classified with a use state of GOOD, BAD, or
`DEAD. A path is declared DEAD if there are no Conduit packets received on the path
`for 1.5 seconds and the Talari APNA will not attempt to use a DEAD path.
`
`To declare a path BAD is a bit more complicated and the Talari APNA uses a
`sophisticated algorithm to calculate the loss, latency, and jitter before making this
`determination…
`
`For all GOOD paths, the Talari APNA uses this same method of calculation to
`determine which of the GOOD paths is declared the best path. This is done on a per-
`
`- 14 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`packet basis, not measured by amount of time. One packet may be sent using an
`Internet path and the next may be sent via MPLS, unless Persistent Path is configured
`in the Talari APNA (this configuration option will be explained below).
`
`Once the Talari APNA determines the best path, traffic may be influenced further
`depending on specific options enabled within the Talari APNA config. The four
`options for mode of transmission are, Duplicate Path, Load Balancing, Persistent Path,
`and Override Service.
`
`… W
`
`ith the Load Balancing option, traffic is put on the best path until that path runs out
`of available bandwidth. The additional traffic will then spill over to the next best path.
`For example with two WAN Links, one containing 1.5M of bandwidth 10ms of
`latency, and the other with 5M of bandwidth but 20ms latency detected. Path 1 will
`be considered the best path by the Talari APNA and used initially when trying to push
`4M of traffic. Talari APNA will fill this pipe with 1.5M, but after 5 ms Talari APNA
`will spill the remaining traffic over onto path 2. Once path 1 has the bandwidth
`necessary, if it is still considered the better path due to the lower latency value, traffic
`will spill back from path 2 to path 1 seamlessly. Load Balancing is the default option
`when creating rules in the Talari APNA config.
`
`With Persistent Path, the latency is allowed to build for 50ms on the best path before
`traffic jumps onto the next best path. Traffic will begin and continue on the best path
`until the latency has built up to the point where the second path is now considered
`better than the first. Traffic will continue on path 2 until the latency on path 1 scales
`back down to a value below that of path 2. Once this level has been reached, the
`Talari APNA will again consider path 1 to be more reliable than path 2. Persistent
`path should be configured for real-time and interactive traffic only, not for bulk data
`unless the exact amount of bandwidth per application is known and available.” Talari
`APN -Best Path Determination at pp. 1-2
`
`(g) forwarding the packet on the
`
`As described below, the accused Talari devices forward the packet on the selected path:
`
`- 15 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`selected network path.
`
`“The user has the ability to configure specific ports on a per WAN link basis.
`When the user configures this option the Talari appliance will use this port as the
`source port for conduit WAN “ingress” traffic. From a Talari perspective, conduit
`WAN ingress traffic is traffic that is from the LAN to WAN or traffic that is
`from the local Talari appliance to the conduit for a destination Talari appliance.”
`UDP Port Configuration Options App Note at p. 4
`
`APN Glossary at p. 8
`
`7
`
`The method of claim 5, wherein the
`forwarding step forwards the packet
`toward the Internet when the packet’s
`destination address does not lie
`within any known location address
`range.
`
`As described below, the accused Talari devices forwards the packet toward the Internet when
`the packet’s destination address does not lie within any known location address range.
`
`The Talari controller can be configured to have a “default route” such that when the packet’s
`destination address does not lie within any known location address range, it is forwarded over
`the default route:
`
`PN | Theory of Operation at p. 51
`
`- 16 -
`
`Talari A
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`Talari APN | Implementation Guide at p. 19.
`
`As shown below, the packets can be forwarded toward the Internet:
`
`Talari APN | Implementation Guide at p. 7.
`
`8
`
`The method of claim 5, wherein the
`destination address identifies a
`destination location to which only a
`
`As described below, in the accused Talari devices, the destination address identifies a
`destination location to which only a single associated network provides connectivity from the
`current location, and the forwarding step forwards the packet to that single associated
`
`- 17 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`single associated network provides
`connectivity from the current
`location, and the forwarding step
`forwards the packet to that single
`associated network.
`
`network.
`
`The accused Talari devices include the ability select a route by reference to routing tables.
`Selecting a route for a packet identifies an associated network interface, and hence a
`particular network, over which a packet should be forwarded:
`
`9
`
`The method of claim 5, wherein
`repeated instances of the selecting
`step make network path selections on
`a packet by-packet basis.
`
`Talari APN | Theory of Operation at p. 47
`
`As described below, the accused Talari devices are configured such that repeated instances of
`the selecting step make network path selections on a packet by-packet basis.
`
`For all GOOD paths, the Talari APNA uses this same method of calculation to
`determine which of the GOOD paths is declared the best path. This is done on a per-
`packet basis, not measured by amount of time. One packet may be sent using an
`Internet path and the next may be sent via MPLS, unless Persistent Path is configured
`in the Talari APNA (this configuration option will be explained below).
`
`“Talari APN -Best Path Determination”at pp. 1-2
`
`- 18 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`Talari Appliances - Giving Enterprises a Biffer, Better, More Reliable WAN that
`Dramatically Reduces Costs at p. 3.
`
`10
`
`The method of claim 5, wherein
`repeated instances of the selecting
`step make network path selections on
`a per session basis.
`
`Talari Appliances Datasheet
`
`As described below, the accused Talari devices are configured such that repeated instances of
`the selecting step make network path selections on a per session basis.
`
`- 19 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`Talari APN | Theory of Operation at p 52
`
`Talari APN | Theory of Operation at p 51
`
`Talari APN | Theory of Operation at p 54
`
`- 20 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`11
`
`The method of claim 5, wherein the
`selecting step selects the network
`path at least in part on the basis of a
`dynamic load-balancing criterion.
`
`As described below, the accused Talari devices select the network path at least in part on the
`basis of a dynamic load-balancing criterion.
`
`Talari APN | Theory of Operation at p. 54
`
`With the Load Balancing option, traffic is put on the best path until that path runs out
`of available bandwidth. The additional traffic will then spill over to the next best path.
`For example with two WAN Links, one containing 1.5M of bandwidth 10ms of
`latency, and the other with 5M of bandwidth but 20ms latency detected. Path 1 will
`be considered the best path by the Talari APNA and used initially when trying to push
`4M of traffic. Talari APNA will fill this pipe with 1.5M, but after 5 ms Talari APNA
`will spill the remaining traffic over onto path 2. Once path 1 has the bandwidth
`necessary, if it is still considered the better path due to the lower latency value, traffic
`will spill back from path 2 to path 1 seamlessly. Load Balancing is the default option
`when creating rules in the Talari APNA config.
`
`Talari APN -Best Path Determination” at pp. 1-2
`
`- 21 -
`
`Talari Networks Inc. - Exhibit 1010
`
`
`
`12
`
`The method of claim 11, wherein
`repeated instances of the selecting
`step select between network paths at
`least in part on the basis of a dynamic
`load-balancing criterion which tends
`to balance line loads by distributing
`packets between lines.
`
`As described below, in the accused Talari devices, repeated instances of the selecting step
`select between network paths at least in part on the basis of a dynamic load-balancing
`criterion which tends to balance line loads by distributing packets between lines.
`
`Talari APN | Theory of Operation at p. 54
`
`With the Load Balancing option, traffic is put on the best path until that path runs out
`of available bandwidth. The additional traffic will then spill over to the next best path.
`For example with two WAN Links, one containing 1.5M of bandwidth 10ms of
`latency, and the other with 5M of bandwidth but 20ms latency detected. Path 1 will
`be considered the best path by the Talari APNA and used initially when trying to push
`4M of traffic. Talari APNA will fill this pipe with 1.5M, but after 5 ms Talari APNA
`will spill the remaining traffic over onto path 2. Once path 1 has the bandwidth
`necessary, if it is still considered the better path due to the lower latency value, traffic
`will spill back from path 2 to path 1 seamlessly. Load Balancing is the default option
`when creating rules in the Talari APNA config.
`
`Talari APN