`571-272-7822
`
`
`
`
`
`Paper 19
`Entered: October 1, 2019
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`AVI NETWORKS, INC.,
`Petitioner,
`
`v.
`
`CITRIX SYSTEMS, INC.,
`Patent Owner.
`_______________
`
`Case IPR2019-00844
`Patent 8,631,120 B2
`_______________
`
`
`
`Before JUSTIN T. ARBES, PATRICK M. BOUCHER, and
`FREDERICK C. LANEY, Administrative Patent Judges.
`
`LANEY, Administrative Patent Judge.
`
`
`
`DECISION
`Granting Institution of Inter Partes Review
`35 U.S.C. § 314
`
`
`
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`
`INTRODUCTION
`I.
`Avi Networks, Inc. (“Avi”) filed a Petition (Paper 6; “Pet.”) to
`institute an inter partes review of claims 1, 2, 4–10, and 12–17 of U.S.
`Patent No. 8,631,120 B2 (Ex. 1001; “the ’120 patent”). Citrix Systems, Inc.
`(“Citrix”) filed a Preliminary Response (Paper 12; “Prelim. Resp.”). We
`have statutory authority over this dispute pursuant to 35 U.S.C. § 314, which
`provides that an inter partes review may not be instituted “unless . . . there is
`a reasonable likelihood that the petitioner would prevail with respect to at
`least 1 of the claims challenged in the petition.”
`If an inter partes review is instituted, a final written decision under
`35 U.S.C. § 318(a) must decide the patentability of all claims challenged in
`the petition. SAS Inst., Inc. v. Iancu, 138 S. Ct. 1348, 1359–60 (2018).
`Upon consideration of the Petition, the Preliminary Response, and the
`evidence cited by the parties, we determine that there is a reasonable
`likelihood that Avi will prevail with respect to at least 1 of the claims
`challenged in the Petition. Accordingly, we institute an inter partes review
`of all claims and grounds challenged in the Petition. This is not a final
`patentability determination of any claim under 35 U.S.C. § 318(a), but only
`an initial determination about whether to institute review.
`
`BACKGROUND
`II.
`Related Proceedings
`A.
`Avi indicates that the ’120 patent is the subject of a lawsuit between
`the parties in the U.S. District Court for the District of Delaware (Pet. 1;
`Ex. 1003, Citrix Systems, Inc. v. Avi Networks, Inc., No. 1:17-cv-1843).
`Additionally, Avi indicates U.S. Patent No. 9,148,493 B2, which is a
`
` 1
`
`
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`continuation of the ’120 patent, is the subject of the above lawsuit and
`IPR2019-00845. Pet. 2. Citrix identifies the same matters. Paper 9, 1
`(“Patent Owner’s Mandatory Notices”).
`The ’120 Patent
`B.
`The ’120 patent “relates generally to Internet client-server
`applications, and more specifically to determining when a client is finished,
`either temporarily (e.g., in ‘think time’) or permanently, with a connection in
`order to more efficiently utilize the pooling of connections between clients
`and servers over the Internet.” Ex. 1001, 1:17–22. The ’120 patent
`represents that the disclosed apparatus, method, and computer program are
`“for efficiently pooling network client-server connections” and may be
`“implemented within an interface unit connecting a plurality of servers to the
`Internet, which is in turn connected to a plurality of clients.” Id. 2:53–58.
`To accomplish this efficient pooling of network client-server connections,
`the ’120 patent describes the following steps:
`opening a connection between a first client and the interface unit;
`determining whether a connection between the interface unit and
`a server is finished being utilized by the first client; opening a
`connection between a second client and the interface unit; if no
`free connection is open between the interface unit and the server,
`then allowing the second client to access information on the
`server via the same connection utilized by the first client without
`waiting for the first client to initiate closing the connection; and
`delinking the connections between the first and second clients
`and the interface unit while keeping open the connection between
`the interface unit and the server.
`Id. 2:60–3:4. Illustratively, the above steps are shown in the ’120 patent in
`Figure 7, reproduced below.
`
`
` 2
`
`
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`
`
`Figure 7 above depicts a message flow diagram illustrating the steps that an
`Internet client-server application causes a device to follow to enable
`connection pooling. Id. 3:28–30.
`
`Illustrative Claim
`C.
`Four of the challenged claims are independent––claims 1, 9, 16, and
`17. Claim 1 is illustrative of the claimed subject matter and is reproduced
`below.
`
` 1. A method of pooling a transport layer connection to a
`server, the method comprising:
`
`(a) receiving, by a device intermediary to a plurality of
`clients and a server, a first request from a first client over a
`first transport layer connection between the first client and
`the device and transmitting the first request over a second
`
` 3
`
`
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`transport layer connection between the device and the
`server;
`
`(b) receiving, by the device, a second request from a
`second client over a third transport layer connection
`between the second client and the device;
`
`(c) determining, by the device from monitoring
`responses to requests from the server, via a content length
`parameter identified in application data of a first response
`that a last byte of the application data has been transmitted
`from
`the
`server and
`the
`server has completed
`communicating the first response to the first request over
`the second transport layer connection;
`the
`
`(d)
`identifying, by
`the device based on
`determination and prior to receiving a command to close the
`first client’s connection to the server, that the second
`transport layer connection is available for use by another
`client of the plurality of clients; and
`
`(e) transmitting, by the device, the second request over
`the second transport layer connection to the server.
`
`
`References and Materials Relied Upon
`D.
`In addition to Declarations by Kevin Jeffay, Ph.D. (Ex. 1007), Duane
`Wessels (Ex. 1008), and Alex Rousskov, Ph.D. (Ex. 1009), Avi relies on the
`following references and materials in support of the asserted grounds of
`unpatentability:
`
`References and Materials
`Source Code for Squid Cache version 2.0 (“Squid”)
`(Oct. 2, 1998)
`WO 00/28433 (pub. May 18, 2000) (“Susai”)
`
`Exhibit No.
`1004
`
`1005
`
` 4
`
`
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`Request for Comments No. 2068, Hypertext Transfer
`Protocol – HTTP/1.1 (Jan. 1997)
`(“RFC 2068”)
`
`1006
`
`
`
`E.
`Challenged Claims
`1, 2, 4–8, 16
`1, 2, 4–8, 16
`9, 10, 12–15, 17
`1, 2, 4–8, 16
`1, 2, 4–10, 12–17
`
`Asserted Grounds of Unpatentability
`Statutory Basis1
`Reference(s)
`35 U.S.C. § 102(b) Squid
`35 U.S.C. § 103
`Squid
`35 U.S.C. § 103
`Squid and RFC 2068
`35 U.S.C. § 103
`Susai and Squid
`35 U.S.C. § 103
`Susai and RFC 2068
`
`III. ANALYSIS
`Citrix raises a preliminary issue that we address before reaching Avi’s
`alleged grounds challenging claims of the ’120 patent. Citrix contends that
`we should exercise our discretion under 35 U.S.C. § 314(a) to deny
`institution in view of the related litigation. Prelim. Resp. 16–22. Citrix
`argues that, because Avi waited until “the last possible day” allowable under
`35 U.S.C. § 315(b) to file the Petition, instituting an inter partes review
`(IPR) of “the ’120 Patent would be a highly duplicative effort that would not
`complete until long after the district court trial’s completion” and would
`create inequities because Avi would be “able to adjust its IPR strategy based
`upon [Citrix’s] district court defense against [Avi’s] invalidity contentions.”
`Id. at 20–21. Citrix contends, “the trial in co-pending district court litigation
`
`
`1 Because the patent application resulting in the ’120 patent was filed before
`the effective date of the Leahy-Smith America Invents Act (“AIA”), we
`refer to the pre-AIA versions of 35 U.S.C. §§ 102 and 103.
`
` 5
`
`
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`will analyze the same prior art references and substantially similar invalidity
`arguments as the Grounds raised in the instant petition.” Id. at 19. The trial
`in the related litigation is currently scheduled to end about two months
`before the statutory deadline for a final decision in this case. Id. at 18 (citing
`Ex. 2002, 17). According to Citrix, “instituting an IPR of the ’120 Patent
`would not provide an effective and efficient alternative to district court
`litigation.” Id. at 20.
`Under the particular factual circumstances of this case, we are not
`persuaded that the timing of Avi’s filing of the instant Petition and the
`overlap of efforts in the concurrent litigation warrant our exercise of
`discretion under 35 U.S.C. § 314(a) to deny institution. As an initial matter,
`because 35 U.S.C. § 315(b) sets the deadline to file a Petition “1 year after
`the date on which the petitioner . . . is served with a complaint alleging
`infringement of the patent,” we are mindful that Congress recognized that an
`instituted IPR proceeding and the associated litigation may involve some
`amount of duplicative efforts. Citrix has not persuasively shown, however,
`that any of the potentially duplicative efforts are more than those that would
`ordinarily occur as a result of filing a Petition within the time window
`Congress set. Nor has Citrix explained sufficiently the alleged tactical
`advantages Avi would receive by proceeding forward with this case.
`Although Citrix relies on the precedential decision NHK Spring Co. v.
`Intri-Plex Technologies, Inc., IPR2018-00752, Paper 8, 19–20 (PTAB 2018)
`(“NHK Spring”), to support its argument (see Prelim. Resp. 17–18), we find
`it distinguishable from this one. In particular, in NHK Spring, the related
`district court case was “nearing its final stages,” with fact discovery
`completed, expert discovery nearly completed, and a trial scheduled for
`
` 6
`
`
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`approximately six months before the final written decision deadline. NHK
`Spring at 20. In the related litigation here, however, fact and expert
`discovery are both ongoing and the trial currently is scheduled to end about
`two months before the final written decision deadline. See Ex. 2002, 5, 7,
`17; Ex. 3001. Furthermore, the schedule in the related litigation appears to
`be more dynamic than the one in NHK Spring and indicators exist that
`suggest the litigation may not be at its final stages. For example, although
`the Scheduling Order entered in the litigation set dates for the close of
`discovery (fact, expert, etc.), the parties have recently stipulated to extending
`those deadlines (which the Court accepted) and they are seeking to extend
`them again. See Ex. 2002, 5, 7; Ex. 3001 (proposing extending the fact
`discovery deadline to October 3, 2019); Ex. 3002 (proposing extending the
`expert discovery deadline to January 17, 2020). Given the willingness of
`both parties in this case to modify and extend the discovery dates that were
`set forth in the Scheduling Order, we are not persuaded on this record that
`the district court proceeding is necessarily nearing its final stages.
`Moreover, given the timing of the two proceedings, it is not certain that a
`trial will even begin before we have entered the final written decision in this
`case. Therefore, we decline to exercise our discretion under 35 U.S.C. §
`314(a) to deny institution.
`We now turn to Avi’s alleged grounds of unpatentability of the
`challenged claims of the ’120 patent. Citrix contends that these challenges
`are deficient because they do not establish the second transport layer
`connection is available for use by another client of the plurality of clients
`“prior to receiving a command to close the first client’s connection to the
`server,” as recited in independent claims 1 and 16, or “while the first client’s
`
` 7
`
`
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`connection with the server is open,” as recited in independent claims 9 and
`17. Prelim. Resp. 22–40. As we explain below, on this record and at this
`stage of the proceeding, we find Avi has presented evidence that is sufficient
`to establish Squid discloses the disputed limitations. To be clear, Avi’s
`evidence is sufficient to establish a reasonable likelihood of prevailing as to
`at least one claim and, therefore, we proceed with an inter partes review, but
`whether any of Avi’s alleged grounds of unpatentability are proven by a
`preponderance of the evidence will be determined once a complete record
`has been developed during trial.
`Asserted Ground of Anticipation by Squid
`A.
`Avi argues that Squid anticipates challenged claims 1, 2, 4–8, and 16.
`
`Pet. 22–63. To anticipate a patent claim under 35 U.S.C. § 102, a single
`prior art reference must “describe every element of the claimed invention,
`either expressly or inherently,” to one of ordinary skill in the art. Advanced
`Display Sys., Inc. v. Kent State Univ., 212 F.3d 1272, 1282 (Fed. Cir. 2000).
`“A reference may anticipate inherently if a claim limitation that is not
`expressly disclosed ‘is necessarily present, or inherent, in the single
`anticipating reference.’” In re Montgomery, 677 F.3d 1375, 1379–80 (Fed.
`Cir. 2012) (quoting Verizon Servs. Corp. v. Cox Fibernet Va., Inc., 602 F.3d
`1325, 1337 (Fed. Cir. 2010), and Schering Corp. v. Geneva Pharms., Inc.,
`339 F.3d 1373, 1377 (Fed. Cir. 2003)). “[I]nherency . . . may not be
`established by probabilities or possibilities.” Id. at 1380 (citing Bettcher
`Indus., Inc. v. Bunzl USA, Inc., 661 F.3d 629, 639 (Fed. Cir. 2011), and In re
`Oelrich, 666 F.2d 578, 581 (CCPA 1981)).
`
` 8
`
`
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`
`1.
`Level of Ordinary Skill in the Art
`Avi asserts that the class of people having ordinary skill in the
`relevant field of art would either have “a Bachelor’s degree in Computer
`Science, Computer Engineering, Electrical Engineering, or an equivalent
`discipline, and at least one year’s worth of experience developing
`client/server systems and/or application layer network communication
`protocols” or, alternatively, “two or more years of work experience in
`computer networking and in the development of client-server systems.” Pet.
`9 (citing Ex. 1007 ¶ 24). Citrix neither sets forth a specific formulation
`regarding the level of ordinary skill in the art nor objects to Avi’s
`contentions regarding who would qualify as one of ordinary skill in the art.
`Furthermore, Citrix does not suggest Avi’s proposal for the level of ordinary
`skill leads to a defective understanding of how a skilled artisan would
`understand either the ’120 patent or the prior art.
`On this record, the cited references are representative of the level of
`ordinary skill in the art, which we find to be supportive of the skill level
`suggested by Avi. See Okajima v. Bourdeau, 261 F.3d 1350, 1355 (Fed. Cir.
`2001) (the level of ordinary skill in the art may be evidenced by the cited
`references themselves). Therefore, we adopt Avi’s proposed formulation of
`the level of ordinary skill in the art for our analysis of the issues below.
`2.
`Claim Construction
`Avi only proposes construing the phrase “transport layer connection”
`because it contends there is no other claim language that has an impact on
`the unpatentability analyses set forth in the Petition. Pet. 7. Avi argues that
`the intrinsic evidence supports construing this phrase to mean a “connection
`at the transport layer between two devices such that there is no application
`
` 9
`
`
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`layer connection between those two devices.” Id. at 18. However, Avi
`“recognizes that its proposed construction is not the only possible
`construction for the term ‘transport layer connection’” and states that “[t]he
`term could alternatively be construed more broadly to include any
`connection that includes a transport layer connection, even if there are
`higher layer connections between the devices.” Id. at 21. Specifically, Avi
`contends that only the grounds relying on Susai as a primary reference are
`applicable if the Board adopts Avi’s narrower construction, whereas all
`grounds are applicable if the Board adopts the broader construction. Id. at
`22. Citrix disagrees with Avi’s proposed narrow construction, but does not
`propose its own construction in the Preliminary Response. Prelim. Resp. 51
`n.10. Citrix also argues that, because none of the alleged grounds set forth
`evidence that satisfies Avi’s proposed construction, Avi failed to satisfy its
`burden under 37 C.F.R. § 42.104(b)(4). Id. at 50–53.
`Where, as here, a petition is filed on or after November 13, 2018, the
`Board applies the same claim construction standard that would be used to
`construe the claim in a civil action under 35 U.S.C. § 282(b). See Changes
`to the Claim Construction Standard for Interpreting Claims in Trial
`Proceedings Before the Patent Trial and Appeal Board, 83 Fed. Reg. 51,340
`(Oct. 11, 2018) (amending 37 C.F.R. § 42.100(b) effective November 13,
`2018 (now codified at 37 C.F.R. 42.100(b) (2019)). “In determining the
`meaning of the disputed claim limitation, we look principally to the intrinsic
`evidence of record, examining the claim language itself, the written
`description, and the prosecution history, if in evidence.” DePuy Spine, Inc.
`v. Medtronic Sofamor Danek, Inc., 469 F.3d 1005, 1014 (Fed. Cir. 2006)
`(citing Phillips v. AWH Corp., 415 F.3d 1303, 1312–17 (Fed. Cir. 2005) (en
`
`10
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`banc)). Extrinsic evidence is “less significant than the intrinsic record in
`determining ‘the legally operative meaning of claim language.’” Phillips,
`415 F.3d at 1317 (citation omitted). Under the above standard, claim terms
`“are generally given their ordinary and customary meaning” as understood
`by a person of ordinary skill in the art at the time of the invention. Id. at
`1312–13.
`An applicant may rebut that presumption by providing a definition of
`the term in the specification with reasonable clarity, deliberateness, and
`precision. In re Paulsen, 30 F.3d 1475, 1480 (Fed. Cir. 1994). An applicant
`also may narrow the meaning of a claim term by disclaiming or disavowing
`claim scope; however, such a “disclaimer or disavowal of claim scope must
`be clear and unmistakable, requiring ‘words or expressions of manifest
`exclusion or restriction’ in the intrinsic record.” Unwired Planet, LLC v.
`Apple Inc., 829 F.3d 1353, 1358 (Fed. Cir. 2016) (quoting Teleflex, Inc. v.
`Ficosa N. Am. Corp., 299 F.3d 1313, 1327 (Fed. Cir. 2002)). In the absence
`of such a definition or disclaimer, limitations are not to be read from the
`specification into the claims. In re Van Geuns, 988 F.2d 1181, 1184 (Fed.
`Cir. 1993).
`Avi proposes a construction for the “transport layer connection”
`phrase that relies on the plain meaning of the terms in the phrase and adds a
`limitation to narrow the scope of their meaning. Specifically, in addition to
`asserting that the phrase refers to a “connection at the transport layer
`between two devices” (i.e., the plain meaning), Avi contends that a skilled
`artisan would also have understood the phrase to exclude any “application
`layer connection between those two devices.” Pet. 18. Avi argues that its
`position finds support in the ’120 patent’s characterization of a translation
`
`11
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`technique used with the “present invention” and described by “U.S. patent
`application Ser. No 09/188,709 . . . entitled, ‘Internet Client-Server
`Multiplexer’” as having “[a] significant advantage” because “no application
`layer interaction is required.” Id. at 19 (citing Ex. 1001, 6:23–32, 9:12–
`12:36). In addition, Avi argues that the prosecution history reinforces its
`position because the claims were distinguished from the prior art because
`they reuse a connection with the server without having to wait for the client
`to send a close command, which allegedly is a “distinction over the prior art
`[that] only arguably makes sense if the ‘transport layer connection’ does not
`include an application layer connection.” Id. at 20–21 (citing Ex. 1002, 470;
`Ex. 1007 ¶ 144).
`Having considered the evidence Avi identifies, we find that it does not
`provide the clear and unmistakable representations necessary to narrow the
`plain meaning of “transport layer connection.” We, accordingly, adopt the
`plain meaning and reject Avi’s proposed narrowing construction for
`“transport layer connection.” Based on the current record, we interpret
`“transport layer connection” to mean any connection at the transport layer
`between two devices, even if there are higher layer connections. See Pet. 18,
`21–22. In view of our decision, we find also that Citrix’s argument that Avi
`failed to satisfy its burden under 37 C.F.R. § 42.104(b)(4) is not persuasive.
`We determine, for purposes of this Decision, that no other claim term
`requires express construction. See Vivid Techs., Inc. v. Am. Sci. & Eng’g,
`Inc., 200 F.3d 795, 803 (Fed. Cir. 1999) (“[O]nly those terms need be
`construed that are in controversy, and only to the extent necessary to resolve
`the controversy.”); Nidec Motor Corp. v. Zhongshan Broad Ocean Motor
`
`
`12
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`Co., 868 F.3d 1013, 1017 (Fed. Cir. 2017) (applying Vivid Techs. in the
`context of an inter partes review).
`3.
`Overview of Squid
`The reference Avi relies on in its anticipation ground, Squid, is a set
`of four source code files titled “client_side.c,” “forward.c,” “http.c,” and
`“pconn.c.” Ex. 1004. Avi cites a book describing “Squid” as an open
`source web proxy that can be used to cache and forward HTTP requests
`between a client and server. Ex. 1013, 20. Squid is code that enables a
`device to “to be both a proxy and a cache.” Id. “As a proxy, Squid is an
`intermediary in a web transaction” and “accepts a request from a client,
`processes that request, and then forwards the request to the origin server.”
`Id. “As a cache, Squid stores recently retrieved web content for possible
`reuse later.” Id. Avi provides arguments and evidence to show that the
`Squid source code files were “distributed together in a single compressed
`file” and constitute a single prior art printed publication under 35 U.S.C.
`§ 102(b). See Pet. 26–30; Ex. 1008 ¶¶ 28–30; Ex. 1009 ¶¶ 22–25;
`Exs. 1011, 1014, 1017–23.
`4.
`Applying Squid to Independent Claims 1 and 16
`The preamble of claim 1 recites a “method of pooling a transport layer
`connection to a server” (Ex. 1001, 17:26–27), which Avi contends is shown
`by Squid’s disclosure of a “web proxy (device) that sets up a pool of
`persistent connections or ‘pconns’ and stores them in a table for later use
`(pool of one or more transport layer connections between the device and a
`server).” Pet. 38 (citing Ex. 1004, 37:2264–67 (httpAccept()). Avi
`identifies source code that it contends, “[d]uring the device setup, Squid
`discloses establishing [a] table of persistent connections (‘pconns’), which is
`
`13
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`a pool of transport layer connections.” Id. at 41 (citing Ex. 1004, 66:161–
`75). “Each of these pconns contains a transport layer connection as they are
`each established using a comm_open() to create a TCP connection.” Id.
`(citing Ex. 1004, 45:194–208); Ex. 1007 ¶ 152.
`Citrix does not dispute that Squid sets up a pool of persistent
`connections to a server in its Preliminary Response. See generally Prelim.
`Resp. 22–35. On this record, having reviewed Avi’s contentions and the
`evidence cited in support thereof, we are persuaded that Avi has shown
`sufficiently at this stage of the proceeding that Squid discloses a “method of
`pooling a transport layer connection to a server,” as claim 1 recites.
`Step (a) of claim 1 recites,
`(a) receiving, by a device intermediary to a plurality of clients
`and a server, a first request from a first client over a first transport
`layer connection between the first client and the device and
`transmitting the first request over a second transport layer
`connection between the device and the server.
`Ex. 1001, 17:28–33. Avi contends that step (a) above is shown by Squid
`because “[w]hen Squid receives a request from a client (a first request of a
`first client over a first connection), it will forward that request over a pconn
`between the Squid and the server (second transport layer connection).” Pet.
`38–39 (citing Ex. 1004, 37:2264–38:304 (httpAccept()), 33:1988–36:2211
`(clientReadRequest()), 2:109–3:130 (clientAccessCheck()), 3:178–4:215
`(clientAccessCheckDone()), 4:217–5:247 (clientRedirectDone()), 27:1640–
`28:1704 (clientProcessRequest()), 28:1706–29:1750 (clientProcessMiss())).
`Avi asserts the Squid code performs the following operations: “Squid
`receives requests from a first client over a first transport layer connection in
`the function httpAccept() and calls clientReadRequest() when necessary.”
`
`
`14
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`Id. at 41–42 (citing Ex. 1004, 37:2264–38:2304). “When there is a cache
`miss, Squid will transmit that request over a second transport layer
`connection from the pool of one or more transport layer connections with the
`server by calling fwdStart(),” which “transmits the first request over the
`second transport layer connection from the pool of available connections
`that it finds using pconnPop().” Id. at 42 (citing Ex. 1004, 28:1706–
`29:1750, 44:187–45:193); Ex. 1007 ¶ 153.
`In its Preliminary Response, Citrix does not dispute the source code of
`Squid is for an intermediary device that receives requests from a plurality of
`clients over a transport layer connection between a client and the device and
`transmits the request over a transport layer connection between the device
`and a server. See generally Prelim. Resp. 22–35. On this record, having
`reviewed Avi’s contentions and the evidence cited in support thereof, we are
`persuaded that Avi has shown sufficiently at this stage of the proceeding that
`Squid discloses step (a) of claim 1.
`Step (b) of claim 1 recites,
`(b) receiving, by the device, a second request from a second
`client over a third transport layer connection between the
`second client and the device.
`Ex. 1001, 17:34–37. Avi contends step (b) above is performed because
`“Squid receives a second request from a second client over a third transport
`layer connection in the function httpAccept() and calls clientReadRequest()
`when necessary.” Pet. 43 (citing Ex. 1004, 37:2264–38:2304 (httpAccept());
`Ex. 1007 ¶ 154.
`In its Preliminary Response, Citrix does not dispute Squid enables a
`device to receive multiple requests from multiple clients over respective
`
`
`15
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`transport layer connections between them. See generally Prelim. Resp. 22–
`35. On this record, having reviewed Avi’s contentions and the evidence
`cited in support thereof, we are persuaded that Avi has shown sufficiently at
`this stage of the proceeding that Squid discloses step (b) of claim 1.
`Steps (c) and (d) of claim 1 recite,
`(c) determining, by the device from monitoring responses
`to requests from the server, via a content length parameter
`identified in application data of a first response that a last byte of
`the application data has been transmitted from the server and the
`server has completed communicating the first response to the
`first request over the second transport layer connection;
`(d) identifying, by the device based on the determination
`and prior to receiving a command to close the first client’s
`connection to the server, that the second transport layer
`connection is available for use by another client of the plurality
`of clients.
`Ex. 1001, 17:38–49. Avi contends Squid performs steps (c) and (d) because
`“Squid discloses determining that the second transport layer connection is
`available based upon the server completing communicating its response to
`the first request by checking the content_length parame[]ter in the response
`from the server in httpPconnTransferDone().” Pet. 43 (citing Ex. 1004,
`55:385–95). “If the server is done sending its message, Squid discloses
`maintaining the second transport layer connection open by calling
`pconnPush(),” which “puts the persistent connection back into the pool so
`that the next client may reuse it.” Id. at 44 (citing Ex. 1004, 56:501–57:517,
`67:195–211). Avi asserts that the source code of Squid cause a device to
`operate as follows:
`Squid determines when the transfer over the pconn is completed
`by checking the content length parameter of the server’s response
`
`
`16
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`(determining that the server has completed communicating the
`last byte of application data via the content length parameter).
`. . . If Squid determines that the response is done based upon the
`content length parameter (identifying that the second transport
`device is available for use by another client), then it will add the
`connection to the table of available pconns without waiting on a
`command from the client to close the connection (prior to
`receiving a close command from the first client).
`Id. at 40 (citing Ex. 1004, 55:397–57:519 (httpReadReply()), 54:341–55:395
`(httpPconnTransferDone()), 66:177–67:211 (pconnPush())); Ex. 1007 ¶ 155;
`Ex. 1008 ¶¶ 21–24; Ex. 1009 ¶¶ 14–17.
`Regarding steps (c) and (d) of claim 1, we first address Citrix’s
`arguments by noting that, except for the “prior to receiving a command to
`close the first client’s connection to the server” limitation, Citrix does not
`dispute at this stage that Squid discloses the elements that steps (c) and (d)
`recite. See Prelim. Resp. 24–32. Citrix argues that the Petition lacks
`sufficient evidence to show Squid discloses the “requirement for timing
`when the second transport layer connection must be identified as available.”
`Id. at 24. In other words, in the Preliminary Response, Citrix argues that the
`evidence in the Petition is insufficient to show Squid discloses the timing
`element of step (d), but does not dispute that Squid’s code discloses (1)
`determining the server response is done based upon the content length
`parameter; (2) identifying, based on determining that the server response is
`done, that the second transport layer connection is available for use by
`another client; and (3) adding the connection to the table of available
`pconns. See id. at 24–32. Citrix argues that the evidence in the Petition
`regarding the timing element of step (d) is defective because Avi does not
`show the functions cited in the code excerpt relied upon in the Petition are
`
`
`17
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`executed before a close command for the first client’s connection to the
`server is received. Id.
`To support its argument, Citrix identifies the fact that the code excerpt
`cited by Avi includes six different functions between when Squid would call
`the httpPconnTransferDone() and pconnPush() functions, which are the
`functions that the Petition relies upon to show steps (c) and (d) of claim 1.
`Id. at 29–30. Citrix provides the annotated figure below with the six
`functions highlighted in red. Id.
`
`
`
`The above figure shows code lines 501–520 from Squid that have been
`annotated by Citrix to highlight the six different functions executed after
`Squid calls the httpPconnTransferDone() function and before it calls the
`pconnPush() function. Id. at 30.
`
`Specifically, the above figure highlights the commSetDefer()
`function, the commSetTimeout() function, the commSetSelect() function,
`the comm_remove_close_handler() function, the storeComplete() function,
`
`18
`
`
`
`
`IPR2019-00844
`Patent 8,631,120 B2
`
`
`and the fwdUnregister() function. Id. Citrix asserts that there are no
`function definitions for at least two of these functions and a skilled artisan
`could not have determined the amount of time that would be necessary to
`execute the httpPconnTransferDone() and pconnPush() functions from this
`excerpt of code. Id. at 30–31. Citrix argues, “even the Petition’s allegation
`that Squid ‘will add the connection to the table of available pconns without
`waiting on a command from the client to close the connection’ is impossible
`to verify from the source code files alone provided in Exhibit 1004.” Id. at
`31 (citing Pet.