`
` UNITED STATES PATENT AND TRADEMARK OFFICE
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`
`
`VMWARE, INC., INTERNATIONAL
`BUSINESS MACHINES
`CORPORATION, and ORACLE
`AMERICA, INC.
`
`
`
`Case No. IPR2014-00901
`
`Case No. IPR2014-00949
`
`Patent No. 6,978,346
`
`Petitioners
`
`v.
`
`ELECTRONICS AND
`TELECOMMUNICATIONS
`RESEARCH INSTITUTE
`
`Patent Owner
`
`
`
`
`
`DECLARATION OF DR. THOMAS M. CONTE
`IN SUPPORT OF PATENT OWNER’S
`RESPONSE TO PETITION FOR INTER PARTES REVIEW
`
`
`Page 1 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`I.
`
`1.
`
`ENGAGEMENT
`
`I have been retained in connection with inter partes review nos. IPR2014-
`
`00901 and IPR2014-00949 regarding U.S. Patent No. 6,978,346. I understand that
`
`the patent is owned by Electronics and Telecommunications Research Institute
`
`(ETRI) and exclusively licensed to Safe Storage, LLC.
`
`2.
`
`I am being compensated at my standard hourly rate for my work on this
`
`matter, including providing this declaration. My compensation is not dependent on
`
`the outcome of this review and in no way affects the substance of my testimony in
`
`this declaration.
`
`3.
`
`I have no financial interest in ETRI; Safe Storage, LLC; the '346 Patent; or in
`
`the outcome of any proceeding involving the '346 Patent.
`
`II. QUALIFICATIONS AND EXPERTISE
`
`4.
`
`I am a (Full) Professor of Computer Science and Electrical & Computer
`
`Engineering with permanent tenure at Georgia Institute of Technology ("Georgia
`
`Tech"). I have been in this position since mid-2008. Prior to that, I was a (Full)
`
`Professor of Electrical and Computer Engineering with permanent tenure at North
`
`Carolina State University ("NC State") from July 1995 to June 2008. I was an
`
`Assistant Professor of Electrical and Computer Engineering at the University of
`
`South Carolina from 1992 to 1995.
`
`Page 2 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`5.
`
`I received my Bachelor of Electrical Engineering degree from the University
`
`of Delaware in 1986; I received my Master of Science in Electrical Engineering
`
`from the University of Illinois at Urbana-Champaign in 1988; and I received my
`
`Doctor of Philosophy in Electrical Engineering from the University of Illinois at
`
`Urbana-Champaign in 1992.
`
`6.
`
`I am a Fellow of the Institute of Electrical and Electronic Engineers ("IEEE")
`
`and a member of the Association for Computing Machinery ("ACM"). I am the
`
`2015 President of the IEEE Computer Society. I have previously held many other
`
`leadership positions in the IEEE and ACM.
`
`7.
`
`I have been the Editor in Chief of the ACM Transactions on Architecture and
`
`Compiler Optimization, and Associate Editor of several journals, including the
`
`Journal on Instruction Level Parallelism, IEEE Embedded Systems Letters, the
`
`IEEE Micro magazine, the IEEE Computer magazine, IEEE Transactions on
`
`Computers (two terms, 1998-2004) and ACM Transactions on Embedded
`
`Computer Systems.
`
`8. From 2000 to 2002, while on leave from NC State, I served as the Chief
`
`Microarchitect and Manager of the Back-End Compiler team for digital signal
`
`microprocessor vendor BOPS, Inc.
`
`Page 3 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`9.
`
`I have in excess of 80 peer-reviewed technical publications, many of which
`
`are frequently cited and three of which have won Best Paper awards.
`
`10. In my 20+ years as a professor of electrical and computer engineering and
`
`computer science, I have taught mass storage (disk drive) interface technologies,
`
`computer network principles, embedded system design, and related computer
`
`architectures. This has given me a deep understanding and expertise in mass
`
`storage design and technology, including fault tolerance as it relates to mass
`
`storage devices.
`
`11. A copy of my curriculum vita is attached to this declaration as Exhibit A.
`
`III. MATERIALS CONSIDERED AND ASSUMPTIONS MADE
`
`12. In addition to any document specifically referenced in this declaration, I have
`
`reviewed and am familiar with the following documents in the record of this case:
`
`Document
`
`'346 Patent
`
`Prosecution History of the '346 Patent
`
`Exhibit/
`Paper No.
`
`1001
`
`2001
`
`Second Correction Petition for Inter Partes Review filed in this case
`on July 21, 2014 ("Petition")
`
`1
`
`Declaration of Dr. Robert Horst filed with the Petition
`1003
`Kevin J. Smith, "Storage Area Networks; Unclogging LANs and 1006
`
`Page 4 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`Improving Data Accessibility, " dated /29/98 ("Mylex")
`
`U.S. Patent No. 5,574,950 ("Hathorn")
`
`Patent Owner’s Preliminary Response filed on Sept. 29, 2014
`
`Decision Instituting Inter Partes Review on Dec. 11, 2014
`
`Webster’s Computer Dictionary (9th ed. 2001)
`
`Microsoft Computer Dictionary (5th ed. 2002)
`
`Final Written Decision in IPR2013-00635 on Feb. 27, 2015
`
`
`
`1005
`
`11
`
`14
`
`2004
`
`2005
`
`39
`
`13. I understand that the '346 Patent granted from U.S. Patent Appl. No.
`
`09/753,245 (Exhibit 2001, pages 1-25) ('245 Application), which was filed at the
`
`United States Patent and Trademark Office (PTO) on December 29, 2000.
`
`14. I understand that the '245 Application claimed priority to Korean Patent Appl.
`
`No. 2000-54807 (Exhibit 2001, pages 61-83) (Korean Priority Application), which
`
`was filed at the Korean Patent Office on September 19, 2000. I understand that as
`
`a result the effective filing date of the '346 Patent is September 19, 2000. I also
`
`assume that this is the invention date for the '346 Patent, although I understand that
`
`the inventors may have actually invented the invention before that date.
`
`15. On June 20, 2014, I provided a declaration in support of the patent owner in
`
`IPR2013-00635 regarding the '346 Patent and Hathorn. In that declaration I
`
`explained why in my opinion Hathorn does not anticipate the claims of the '346
`
`Page 5 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`Patent. For example, I explained in that declaration my opinion that Figure 3 of
`
`Hathorn does not include a RAID or RAID controllers, as claimed in the '346
`
`Patent.
`
`16. I understand that the Patent Trial and Appeal Board issued a final written
`
`decision in IPR2013-00635 on February 27, 2015, concluding that Hathorn does
`
`not anticipate the claims of the '346 Patent.
`
`17. I understand that the Petitioners in the present cases now assert that claims 1-
`
`9 of the '346 Patent are obvious over Mylex and Hathorn under 35 U.S.C. § 103
`
`and that trial has been instituted to answer that question.
`
`18. I have been instructed to give the claims of the '346 Patent at issue in this
`
`review their broadest reasonable interpretation in light of the '346 Patent’s
`
`specification and prosecution history. I understand that this interpretation is from
`
`the perspective of one of ordinary skill in the art at the time of the patent’s
`
`effective filing date.
`
`19. I understand that the following is a quotation of 35 U.S.C. § 103(a), which
`
`governs the obviousness of inventions:
`
`A patent may not be obtained though the invention is not
`identically disclosed or described as set forth in section
`102 of this title, if the differences between the subject
`
`Page 6 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`matter sought to be patented and the prior art are such
`that the subject matter as a whole would have been
`obvious at the time the invention was made to a person
`having ordinary skill in the art to which said subject
`matter pertains. Patentability shall not be negative by the
`manner in which the invention was made.
`
`20. I have been informed that the Supreme Court of the United States, in the 1966
`
`case of Graham v. John Deere, explained that an obviousness inquiry should
`
`consider the following non-exclusive factors: (1) The scope and content of the
`
`prior art; (2) the differences between the claimed invention and the prior art; (3)
`
`the level of ordinary skill in the pertinent art; and (4) so-called "secondary
`
`considerations" of non-obviousness, such as commercial success, long felt but
`
`unsolved need, failure of others, etc., if any.
`
`21. I have not considered any secondary considerations of non-obviousness in
`
`this case, although I recognize that the patent owner may present such secondary
`
`considerations to support its positions. I have concluded that the claims of the '346
`
`Patent are not obvious over the prior art at issue in the case, regardless of
`
`secondary considerations. Any secondary considerations would only further
`
`support the conclusion of non-obviousness in this case.
`
`Page 7 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`22. I further understand that a conclusion that a patent claim is obvious under 35
`
`U.S.C. § 103 requires that one or more prior art references collectively teach or
`
`suggest all of the limitations of the claim arranged together as stated in the claim.
`
`23. I have been informed that the Supreme Court of the United States, in the 2007
`
`case of KSR v. Teleflex, explained that a conclusion that a patent claim is obvious
`
`under 35 U.S.C. § 103 must also be supported by "some articulated reasoning with
`
`some rational underpinning to support the legal conclusion of obviousness."
`
`24. I have also been informed that the most common rationale for obviousness is
`
`the so-called teaching-suggestion-motivation test. I have been informed that a
`
`conclusion of obviousness under this rationale requires at least: (1) a finding that
`
`there was some teaching, suggestion, or motivation, either in the prior art
`
`references themselves or in the knowledge generally available to one of ordinary
`
`skill in the art, to modify the reference or to combine reference teachings; (2) a
`
`finding that there was reasonable expectation of success; and (3) whatever
`
`additional findings based on the Graham v. John Deere factors that may be
`
`necessary, in view of the facts of the case.
`
`25. I have also been informed that another obviousness rationale that might be
`
`applicable in this case is combining prior art elements according to known methods
`
`to yield predictable results. I have been informed that a conclusion of obviousness
`
`Page 8 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`under this rationale typically involves at least: (1) a finding that the prior art
`
`included each element claimed, although not necessarily in a single prior art
`
`reference, with the only difference between the claimed invention and the prior art
`
`being the lack of actual combination of the elements in a single prior art reference;
`
`(2) a finding that one of ordinary skill in the art could have combined the elements
`
`as claimed by known methods, and that in combination, each element merely
`
`performs the same function as it does separately; (3) a finding that one of ordinary
`
`skill in the art would have recognized that the results of the combination were
`
`predictable; and (4) whatever additional findings based on the Graham v. John
`
`Deere factors that may be necessary, in view of the facts of the case. I have been
`
`informed that under this rationale, it can also be important to identify a reason that
`
`would have prompted a person of ordinary skill in the art to combine the elements
`
`in the way the claimed invention does.
`
`26. I also understand that the obviousness inquiry must not be tainted by
`
`hindsight bias from knowledge of the invention under consideration. In other
`
`words, it is improper to use the invention as a blueprint for putting together
`
`teachings from the prior art. Instead, a conclusion of obviousness must be based
`
`solely on knowledge from the prior art just before the invention was made.
`
`Page 9 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`IV. TECHNOLOGY TUTORIAL
`
`27. The technology at issue in this case involves fault tolerance in mass storage
`
`devices. Generally a (non-fault-tolerant) mass storage device is an electro-
`
`magnetic mechanical disk drive. Modern disk drives have multiple disks
`
`(sometimes called "platters") where each disk is a spinning media with its own
`
`read/write head. The overall disk drive is comprised of a stack of each disk into a
`
`column that rotates together. Each disk's head is attached to an actuator. A "disk
`
`drive" includes all of these disks, heads, etc., within one enclosure, vis.:
`
`
`
`These disk drives are also called "DASDs" (for "Direct Access Storage Devices")
`
`by International Business Machines Corporation. Each disk drive has a certain
`
`amount of capacity to store binary data. Because it has moving parts, disk drives
`
`are known to fail. In addition, the circuitry that interfaces with the disk drive,
`
`called a "controller" in the art, also is prone to failure. A failure that results in the
`
`loss of data can be catastrophic. A standard engineering figure of merit in
`
`Page 10 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`measuring the potential of failure is the mean time to failure (MTTF). One way to
`
`extend MTTF is to construct the disk drive out of highly reliable, rugged
`
`components. These components increase the expense of the disk drive, hence this
`
`approach to fault tolerance is called a Single, Large (Expensive) Disk drive
`
`(SLED).
`
`28. Another method to extend MTTF to guard against failure is to periodically
`
`make a backup copy of the contents of a disk drive on another device. Periodic
`
`backups suffer from the drawback that any changes made after the last backup date
`
`are lost when a failure occurs.
`
`A. RAID
`
`29. Redundant operation of multiple disk drives is another way to extend MTTF
`
`via making backup copies of data. In one such technology, called Redundant
`
`Arrays of Inexpensive1 Disk drives ("RAID"), the mass storage device is built
`
`from multiple, physical disk drives. In a RAID, the RAID controller translates a
`
`write to the RAID into data being written in parallel to two or more disk drives. If
`
`
`1 Over time, the "I" in the RAID acronym has changed from "Inexpensive"
`to "Independent." There is no technical difference between "Redundant Array of
`Inexpensive Disks" and "Redundant Array of Independent Disks" in the art. To
`avoid ambiguity, I will use "RAID" throughout this declaration.
`
`Page 11 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`one of the disk drives fails, then the backup of the data is up to date on the
`
`redundant drives, thereby allowing for recovery from disk drive failure. RAID
`
`technology only guards against individual disk drive failure. It is not capable of
`
`recovering from RAID controller failure.
`
`30. The idea of providing multiple physical disk drives for redundancy is not the
`
`novel part of a RAID. Rather, what sets a RAID apart is that it is a "black box"
`
`that can be interchanged with a traditional SLED without needing to change the
`
`hardware or software interfaces. Webster’s Computer Dictionary defines "RAID
`
`x" (where x = 0, 1, and 2) as "[a] type of RAID storage device that combines two
`
`or more hard disk drives into a single logical drive …" Webster’s Computer
`
`Dictionary 9th ed. 2001, at 308. Similarly, the Microsoft Computer Dictionary
`
`defines "RAID" saying it is "[a] data storage method in which data is distributed
`
`across a group of computer disk drives that function as a single storage unit …"
`
`Microsoft Computer Dictionary 5th ed. 2002, at 437.
`
`31. The "black box" nature of a RAID (i.e., the dictionary definitions of it being a
`
`"single logical drive," or "single storage unit") is accomplished via an intelligent
`
`RAID controller. Thus a RAID must include this controller plus two or more
`
`physical disk drives (to support redundancy).
`
`Page 12 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`32. One significant advantage of a RAID is that the user of a RAID interacts with
`
`the RAID controller as if it were a controller for a single disk drive, and this in
`
`turn simplifies replacing traditional disk drives with RAIDs.
`
`B.
`
`The '346 Patent
`
`33. The '346 Patent describes an apparatus capable of recovering from RAID
`
`controller failure in a multiple-host environment. '346 Patent at Abstract. The '346
`
`Patent explains prior art approaches to such controller failure. One such prior art
`
`apparatus is illustrated in Fig. 1, vis.:
`
`'346 Patent at Fig. 1.
`
`
`
`Page 13 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`34. The approach in Fig. 1 does not provide fault tolerance, but it does provide
`
`high communications bandwidth with the RAID. '346 Patent at 1:35-38. In this
`
`approach, the two host computers, 100 and 101, each contain network interface
`
`controllers to communicate with the RAID, 110 and 111 (resp.). These network
`
`interface controllers inside the hosts have direct connections (120 and 121) with
`
`two RAID controllers (140 and 141) via each RAID controller’s own network
`
`interface controller (150 and 151). These two controllers allow the two host
`
`computers to generally transfer data to/from the RAID independently, and thus in
`
`parallel. '346 Patent at 1: 24-34.
`
`35. In sum, even though there are two redundant RAID controllers and network
`
`interface controllers in this prior art apparatus, a failure of a RAID controller or its
`
`network interface controller results in a partial system failure because it disables a
`
`host from accessing the RAID.
`
`36. In contrast, a second prior art apparatus that allows for fault tolerance is
`
`presented in the '346, vis.:
`
`Page 14 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`
`
`'346 Patent at Fig. 2.
`
`37. In this second prior art approach, each RAID controller is augmented with a
`
`communication controller, 221 and 222, in order for the two controllers to
`
`exchange information to facilitate failover in the case of a failure of one or the
`
`other. Additionally, in contrast to Fig. 1, the two hosts are now connected to the
`
`RAID via a HUB or SWITCH 210. If for example one of the RAID controllers 230
`
`itself fails or its connection to the HUB or SWITCH 210 fails, the HUB or
`
`Page 15 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`SWITCH 210 can still provide connectivity between both hosts and the RAID
`
`itself via re-routing accesses to the other RAID controller 231. A drawback is that
`
`the failure mode just described provides the hosts with only half of the access
`
`bandwidth to the RAID as was provided in the original, fully-operational system.
`
`'346 Patent at col. 1:58-59. Another drawback is that a failure of the HUB or
`
`SWITCH 210 results in a system failure because it disables both hosts from
`
`accessing the RAID.
`
`38. The '346 Patent also describes a third prior art apparatus in Fig. 3, vis.:
`
`
`
`Page 16 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`'346 Patent at Fig. 3.
`
`39. In the above Fig. 3 prior art, if one of the RAID controllers fails, both hosts
`
`maintain connectivity to the RAID 340 via switches 310, 311, 320 and 321. As
`
`with Fig. 2, when a failure occurs, host access bandwidth to the RAID is reduced.
`
`Also, a failure of any one of the ports 310 or 311 results in a system failure
`
`because it disables both hosts from being able to fully access the RAID.
`
`40. In contrast with the prior art apparati, the '346 Patent and claims present an
`
`apparatus that does not suffer from the drawbacks of Fig. 2 or Fig 3. This
`
`approach is illustrated below, vis.:
`
`
`
`Page 17 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`'346 Patent at Fig. 4.
`
`41. The above apparatus provides additional fault tolerance as well as
`
`preservation of pre-fault bandwidth after failure of either RAID controller 480 or
`
`481. In contrast to the prior art apparati, the above apparatus provides two HUB or
`
`SWITCH modules, and each RAID controller contains redundant network
`
`interfaces, 470 and 471 for RAID controller 460, and 480 and 481 for RAID
`
`controller 461. If there is a failure in one RAID controller, for example 461, then
`
`all of the host access requests can be forwarded to the non-failed RAID controller,
`
`e.g., 460, via the interconnection 450. This is shown below, vis:
`
`
`
`
`
`42. Unlike prior art with just one HUB or SWITCH (e.g., Fig. 2), having two
`
`HUB or SWITCH modules provides additional fault tolerance, as a failure of one
`
`HUB or SWITCH will only prevent some hosts from accessing the RAID. For
`
`Page 18 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`example, hosts 400-402 can continue to access the RAID even if the HUB or
`
`SWITCH 441 fails. Also, in the case of failure of one RAID controller in the
`
`system of Fig. 4, unlike prior art (Figs. 1-3), there is no longer a reduction in
`
`access bandwidth between the hosts. '346 Patent at 3:1-9.
`
`43. Alternative preferred embodiments of the invention are shown in Figs. 5 and
`
`6.
`
`44. The claims of the '346 Patent are reproduced below, vis.:
`
`1. An apparatus for a redundant interconnection between multiple hosts and a
`RAID, comprising:
`
`1(a): a first RAID controlling units and a second RAID controlling unit for
`processing a requirement of numerous host computers, the first RAID
`controlling unit including a first network controlling unit and a second
`network controlling unit, and the second RAID controlling unit including a
`third network controlling unit and a fourth network controlling unit; and
`
`1(b): a plurality of connection units for connecting the first RAID
`controlling units and the second RAID controlling unit to the numerous
`host computers, wherein the first RAID controlling unit and the second
`RAID controlling unit directly exchange information with the numerous
`host computers through the plurality of connecting units, and the first
`network controlling unit exchanges information with the fourth network
`controlling unit, and the second network controlling unit exchanges
`information with the third network controlling unit.
`
`2. The apparatus as recited in claim 1, wherein said respective RAID controlling
`units are connected to the plurality of individual connecting units.
`
`3. The apparatus as recited in claim 2, wherein the first network interface
`controlling unit is coupled to the connecting unit of one side and the second
`network interface controlling unit is coupled to the connecting unit of another
`
`Page 19 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`side.
`
`4. The apparatus as recited in claim 3, wherein
`
`4(a): the first network interface controlling unit and the third network
`interface controlling unit process the requirement of the numerous host
`computers; and
`
`4(b): the second network interface controlling unit and the fourth network
`controlling unit are used for communication between the first RAID
`controlling unit and the second RAID controlling unit when the first and
`second RAID controlling units are not faulty and the second network
`interface controlling unit and the fourth network controlling unit are used
`for executing a function of the first network interface controlling unit and
`the third network controlling unit when one of the first RAID controlling
`unit and the second RAID controlling unit is faulty.
`
`5. The apparatus as recited in claim 1, wherein said plurality of connecting units
`have at least three connection ports, two of the at least three connection ports is
`coupled to one of the first network interface controlling unit and the third
`network controlling unit and the rest of the connection ports being provided as a
`hub equipment connected with the numerous host computers.
`
`6. The apparatus as recited in claim 1, wherein said plurality of connecting units
`have at least three connection ports, two of the at least three connection port are
`coupled to one of the first network interface controlling unit and the third
`network controlling unit and the rest of the connection ports being provided as a
`network switch equipment connected with the numerous host computers.
`
`7. The apparatus as recited in claim 1, wherein said plurality of connecting units
`have at least five connection ports, four of the at least five connection ports is
`coupled to one of the first network interface controlling unit and the third
`network controlling unit and the rest of the connection ports being provided as a
`switch connected with the numerous host computers.
`
`8. The apparatus as recited in claim 1, wherein the first network interface
`controlling unit of the first RAID controlling unit being connected to a first
`connecting unit, the second network interface controlling unit of said first RAID
`controlling unit being connected to a second connecting unit, the third network
`interface controlling unit of the second RAID controlling unit being connected to
`the second connecting unit, and the fourth network interface controlling unit of
`
`Page 20 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`the second RAID controlling unit being connected to the first connecting unit.
`
`9. An apparatus for a redundant interconnection between multiple host computers
`and a RAID, the apparatus comprising:
`
`9(a): a plurality of connection units for connecting the host computers and
`the RAID;
`
`9(b): a first and a second RAID controllers, included in the RAID, each of
`which having a first network interface controller and a second network
`interface controller for processing requests from the plurality of the host
`computers connected through the plurality of the connection units,
`
`9(c): wherein the first network interface controller in the first RAID
`controller supplies data to the host computers connected through the
`plurality of connection units and processes information transmitted from
`the second network interface controller in the second RAID controller,
`
`9(d): wherein the first network interface controller in the second RAID
`controller supplies data to the host computers connected through the
`plurality of connection units and processes information transmitted from
`the second network interface controller in the first RAID controller,
`
`9(e): wherein the second network interface controller in the first RAID
`controller is used for fault tolerance by performing functions of the first
`network interface controller in the second RAID controller when the
`second RAID controller is faulty, and
`
`9(f): wherein the second network interface controller in the second RAID
`controller is used for fault tolerance by performing functions of the first
`network interface controller in the first RAID controller when the first
`RAID controller is faulty, and
`
`9(g): wherein the first network controlling unit in the first RAID
`the second network
`controlling unit exchanges
`information with
`controlling unit in the second RAID controlling unit, and the second
`network controlling unit in the first RAID controlling unit exchanges
`information with the first network controlling unit in the second RAID
`controlling unit.
`
`
`
`Page 21 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`45. I note that the inventors are non-native speakers of English. One consequence
`
`of this is that there is a typographical error in claim 1 wherein "first RAID
`
`controlling units" should instead have been "first RAID controlling unit", vis.
`
`[highlighting added]:
`
`1(a): a first RAID controlling units and a second RAID controlling unit for
`processing a requirement of numerous host computers, the first RAID controlling
`unit including a first network controlling unit and a second network controlling
`unit, and the second RAID controlling unit including a third network controlling
`unit and a fourth network controlling unit; and
`
`1(b): a plurality of connection units for connecting the first RAID controlling
`units and the second RAID controlling unit to the numerous host computers,
`wherein the first RAID controlling unit and the second RAID controlling unit
`directly exchange information with the numerous host computers through the
`plurality of connecting units, and the first network controlling unit exchanges
`information with the fourth network controlling unit, and the second network
`controlling unit exchanges information with the third network controlling unit.
`
`
`
`46. The remainder of claim 1 is consistent with this being a typographical error.
`
`For example, 1(a) states, "a first RAID controlling units and a second RAID
`
`controlling unit for processing a requirement of numerous host computers, the first
`
`RAID controlling unit including…". Similarly, 1(b) states, "first RAID controlling
`
`units and the second RAID controlling unit to the numerous host computers,
`
`wherein the first RAID controlling unit and the second RAID controlling unit….".
`
`Page 22 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`47. The case of this being a typographical error is further reinforced in, for
`
`example, dependent claim 8 that only refers to "first RAID controlling unit," rather
`
`than "first RAID controlling units", vis.:
`
`8. The apparatus as recited in claim 1, wherein the first network interface
`controlling unit of the first RAID controlling unit being connected to a first
`connecting unit, the second network interface controlling unit of said first RAID
`controlling unit being connected to a second connecting unit, the third network
`interface controlling unit of the second RAID controlling unit being connected to
`the second connecting unit, and the fourth network interface controlling unit of
`the second RAID controlling unit being connected to the first connecting unit.
`
`C.
`
`Fibre Channel
`
`48. Fibre Channel is a type of network that has become a standard for connecting
`
`storage devices to computers. The '346 Patent says at 3:25-30 that the invention
`
`may work with any manner of network including a fibre channel network. Mylex
`
`refers frequently to fibre channel as one of the types of networks that connect the
`
`host computers to the RAID controllers. For example, in Figs. 17-19, the host-side
`
`network represented by the oval above the RAID controllers is a fibre channel
`
`network.
`
`49. A fibre channel network typically utilizes fiber optic cables to transmit data at
`
`high rates, although it is possible to implement the fibre channel standard using
`
`electrical, rather than fiber optic, connections. Fibre channel networks are packet-
`
`Page 23 of 58
`
`Pat. Owner ETRI Ex. 2301
`IPR2014-00901, -00949
`VMware, IBM, & Oracle v. ETRI
`
`
`
`switched, meaning that all communications are broken up into packets, each of
`
`which is transported independently through the network to its destination. Each
`
`packet includes extra information in a header that includes the address of its
`
`destination. That address is used to route the packet through the network to its
`
`destination, where the packets are depacketized and the information reassembled
`
`from various packets in order to reconstruct the message.
`
`50. Fibre channel networks can be constructed in various forms, including fibre-
`
`channel arbitrated loops and fiber channel switched fabrics. A fibre channel
`
`arbitrated loop (sometimes denoted "FC-AL") connects all of the nodes of the
`
`network in a loop. Packets are passed around the loop (in one direction only) until
`
`they reach their destination. Access to the loop is arbitrated using a token passing
`
`scheme, where a token is passed around the loop, a node wanting to transmit holds
`
`that token, transmits, and then appends a new token to the end of its message when
`
`done. This permits other nodes to add their messages to the chain, etc. Once a
`
`packet reaches its destination, the node recognizes it is addressed to it and the
`
`packet is removed from the loop. The '346 Patent at 3:19-24 describes connection
`
`unit 440 in Fig. 4 as a FC-AL hub.
`
`51. In a fibre channel switched fabric (sometimes denoted "FC-SW"), a node or
`
`group of nodes is connected to a switch, which routes pack