Ex Parte Gunda et al

Board of Patent Appeals and Interferences

Mar 27, 2009

10321120 (B.P.A.I. Mar. 27, 2009)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________

Ex parte LAXMIKANT GUNDA, BALAJI NARASIMHAN,
SARA ABRAHAM, SHIE-REI HUANG, and NAGARAJ SHYAM
____________

Appeal 2008-2285
Application 10/321,120
Technology Center 2100
____________

Decided:1 March 30, 2009
____________

Before JAMES D. THOMAS, JEAN R. HOMERE, and
STEPHEN C. SIU, Administrative Patent Judges.
THOMAS, Administrative Patent Judge.

DECISION ON APPEAL

1 The two-month time period for filing an appeal or commencing a civil
action, as recited in 37 C.F.R. § 1.304, begins to run from the decided date
shown on this page of the decision. The time period does not run from the
Mail Date (paper delivery) or Notification Date (electronic delivery).
Appeal 2008-2285
Application 10/321,120

2
STATEMENT OF THE CASE
This is an appeal under 35 U.S.C. § 134(a) from the Examiner’s final
rejection of claims 1-25. We have jurisdiction under 35 U.S.C. § 6(b).
We Affirm.

INVENTION
Appellants’ invention relates to distributed file system I/O recovery in
storage networks. Detection of loss of access (server and/or network
failures) leads to the I/O recovery mechanism 220 first introduced in
disclosed Figure 3 to store failed and new I/O access requests.
(Abstract 30).

REPRESENTATIVE CLAIM
Claim 1 further illustrates the invention, and is reproduced below:
1. A system, comprising:

one or more storage devices;

a server node coupled to the one or more storage devices via a storage
network; and

a client node comprising an application and configured to access the
server node via a network, wherein the application is configured to access
data on the one or more storage devices by sending I/0 requests to the server
node over the network;

wherein the client node is configured to:
Appeal 2008-2285
Application 10/321,120

3
detect loss of access to the server node via the network, wherein said
loss generates one or more failed I/O requests from the application to the
server node;

store failed I/0 requests and I/0 requests issued from the application to
the server node after detection of the loss of access;

detect recovery of access to the server node via the network; and

send the stored I/0 requests to the server node after detection of the
recovery.

PRIOR ART AND EXAMINER’S REJECTIONS
The Examiner relies upon the following references as evidence of
unpatentability:
Murphy US 6,185,695 B1 Feb. 06, 2001
Purcell US 6,246,666 B1 Jun. 12, 2001
Gatto US 2002/0174160 A1 Nov. 21, 2002
(filed May 21, 2001)

Frank Naude, Oracle Real Application Server/Parallel Server FAQ,
available at, http://www.orafag.com/faqrac.htm, (2001.)

All claims on appeal, claims 1-25, stand rejected under 35 U.S.C.
§ 103. As evidence of obviousness in the first stated rejection of claims 1, 3,
5-7, 9, 11-14, 16, 18-20, 22, 24, and 25, the Examiner relies upon Murphy in
view of Purcell. In a second stated rejection of dependent claims 4, 10, 17,
and 23, the Examiner adds Oracle to the first combination of references.
Appeal 2008-2285
Application 10/321,120

4
Lastly, in a third stated rejection of dependent claims 2, 8, 15, and 21, the
Examiner relies upon Murphy in view Purcell, further in view of Gatto.

Claim Groupings
According to Appellants’ arguments at pages 13-16 of the principal
Brief, Appellants consider independent claim 1 as representative of the
subject matter of independent claims 1, 7, 13, 14, and 20. We do so as well.
According to the positions set forth at the bottom of page 16 of this Brief,
Appellants consider claim 5 as representative of the subject matter of
independent claims 5, 11, 18, and 24. Page 16 of the Brief considers
dependent claim 3 as representative of the subject matter of claims 3, 9, 16,
and 22. Next, according to the arguments at page 17 of the Brief, Appellants
consider the subject matter of dependent claim 6 as representative of the
subject matter of claims 6, 12, 19, and 25. As to the Examiner’s second
stated rejection, pages 19 and 20 of the Brief consider dependent claim 4 as
representative of the subject matter of claims 4, 10, 17, and 23. Lastly,
according to the third stated rejection of the Examiner, the arguments at page
18 with respect to claim 2 consider this claim as representative of the subject
matter of claims 2, 8, 15, and 21.

ISSUE
Have Appellants shown that the Examiner erred in finding that the
combination of teachings of the respectively applied prior art in the
Appeal 2008-2285
Application 10/321,120

5
Examiner’s three stated rejections would have rendered obvious to one of
ordinary skill in the art the subject matter of the respectively rejected,
representative claims within 35 U.S.C. § 103.

FINDINGS OF FACT
1. Appellants’ prior art Figures 1 and 2 are discussed at Specification
pages 1 and 2 in paragraphs [0003] –[0006]. Exemplary storage networks
include a so-called Storage Area Network (SAN). Input output failures
include file server failures and network connection failures. As discussed in
paragraph [0006] at page 2 of the Specification as filed, this so-called SAN
model network places storage on its own dedicated network separate from
the main user control network. This permits direct accessibility on what is
commonly known in the art as Fibre Channel technology. Corresponding to
prior art Figure 1, Appellants’ disclosed Figure 3 introduces the I/O recovery
mechanism 220 to provide the claimed recovery features; a corresponding
improvement is shown in disclosed Figure 4 over that which is in prior art
Figure 2.

2. Murphy’s Abstract teaches transparent failover from a primary
copy of an object on a first sever to a secondary copy of the object on a
second server when the first server fails, or otherwise become unresponsive.
Figures 1, 2A, 2B, 3, 5, and 9 are the most pertinent in Murphy. Figure 1
shows such a transparent failover mechanism 104, 108, 112, and 116 in
Appeal 2008-2285
Application 10/321,120

6
respective distributed computing nodes 102, 106, 110, and 114, where any of
these nodes may function both as clients and/or server nodes. The so-called
highly available server 211 in Figure 2B comprises a primary server 212 and
a secondary server 213 with identical object storages. Upon a reply failure
the client 200 may invoke a request retry to the secondary server 213 to
receive a reply for the desired object. This results in a transparent retrying
or repeating of a desired access for a data object. Figure 3 shows a replica
handler at a particular system software layer and Figure 5 shows a replica
manager 500 to detect server failures. Figure 9 depicts a failover flowchart
from the primary server to a secondary server, such as illustrated in Figure
2B in a clustered server environment. The transparent failover mechanism
104 in Figure 1 of Murphy appears to compare directly with the Appellants’
I/O recovery mechanism 220 noted in finding of fact 2 and Appellants’
disclosed Figures 3 and 4.

3. Murphy’s column 1 reveals a background of the prior art that
includes the switching over to a backup copy of another sever, which
functionality is referred to as “failover.” Lines 35-40 state “[e]xisting client-
server systems typically rely on the client application program to explicitly
detect and recover from server failures. For example, a client application
program typically includes code that explicitly specifies timeout and retry
procedures.” To permit such a retry capability, the prior art implicitly must
store an accessibility to the desired object and permit a timeout to include a
Appeal 2008-2285
Application 10/321,120

7
retry of instances until the timeout occurs or a successful accessibility occurs
within a given time allotted. Murphy’s column 3, lines 16-24 teach:
Replica Manager--a process or mechanism that keep track of the
various primary and secondary copies of an object on various servers.
Retrying Transparently--retrying an invocation to an object without
explicit retrying by the client application program.
Transparent Failover--A failover that occurs automatically, without
explicit failure detection and retry commands from a client application
program.

The paragraph bridging columns 3 and 4 of Murphy also indicate that a so-
called Solaris MC operating system is utilized in the embodiment in Murphy
which is stated to be a UNIX-based operating system.

4. Murphy’s Figure 9 appears to show details of the transparent
failover mechanism 104, 108, 112, 116 illustrated in Figure 1 of this
reference which has a corresponding discussion beginning at column 10, line
19 through the end of column 11, line 11. As properly relied upon by the
Examiner, the showing in Figure 9 and the discussion there indicates that,
upon detection of a server failure, so-called in-progress invocations to
retrieve data are stated to “wait” to be processed, thus clearly indicating that
they are stored during the time in which the retry of them is indicated in
Figure 9, block 906. Block 908 also “blocks” new invocation for storage
access which also correspondingly indicates that such new accesses are
stored until the failover is complete. Once the switchover to the respective
Appeal 2008-2285
Application 10/321,120

8
new server is complete, the respective in progress invocations and the new
invocations for data access are unblocked to permit their accessibility in the
secondary server.

5. Purcell’s cluster approach also permits failover processes as
initially illustrated in Figure 1. This Figure shows general operating systems
to include real time operating systems. With respect to Figure 1, column 5,
lines 10-16 state:
The communication link 320 can utilize any communication medium
and protocol including a Fibre Channel and a Server Net Fail-over
link both of which are commonly known in the industry. I/O
subsystem 150 and I/O subsystem 151 are capable of Peer-to-Peer
communication across the communication link 320 in a manner
commonly known in the industry.

6. In addition to the Examiner’s relied upon teachings of various
benefits of Oracle’s parallel sever option (PSO/OPS) relied upon by the
Examiner at the top of page 2 of this publication, the discussion at the
bottom of page 1 indicates that the environment in which this operates
includes UNIX’s operating systems operating in clusters, the indication of
which is repeated at the middle of page 2 of this reference.

7. As relied upon by the Examiner, Gatto’s network controller
operates in part as discussed at page 7 [para. 0077] which is reproduced
here:
Appeal 2008-2285
Application 10/321,120

9
When the cached transaction packet (3) fails to reach the remote
server 534, a first retry packet (7) is sent through the network after a
predetermined and/or selectable period of time. If no
acknowledgement is received from the remote server 534 after a
predetermined and/or selectable period of time, a second retry packet
(8) is sent to the network. The same scenario may be repeated forever
or (preferably) for a selectable number or retries or period of time or
until an acknowledgement is received from the remote server 534.

PRINCIPLES OF LAW
"[T]he PTO gives claims their 'broadest reasonable interpretation.'"
In re Bigio, 381 F.3d 1320, 1324 (Fed. Cir. 2004) (quoting In re Hyatt, 211
F.3d 1367, 1372 (Fed. Cir. 2000)). "Moreover, limitations are not to be read
into the claims from the specification." In re Van Geuns, 988 F.2d 1181,
1184 (Fed. Cir. 1993) (citing In re Zletz, 893 F.2d 319, 321 (Fed. Cir.
1989)). Our reviewing court has repeatedly warned against confining the
claims to specific embodiments described in the specification. Phillips v.
AWH Corp., 415 F.3d 1303, 1323 (Fed. Cir. 2005) (en banc).
One cannot show nonobviousness by attacking references individually
where the rejections are based on combinations of references. In re Merck
& Co., Inc., 800 F.2d 1091, 1097 (Fed. Cir. 1986).
Section 103 forbids issuance of a patent when “the differences
between the subject 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.”

KSR Int'l Co. v. Teleflex Inc., 127 S. Ct. 1727, 1734 (2007).
Appeal 2008-2285
Application 10/321,120

10
The Supreme Court emphasized “the need for caution in granting a
patent based on the combination of elements found in the prior art,” and
discussed circumstances in which a patent might be determined to be
obvious. KSR, 127 S. Ct. at 1739 (citing Graham v. John Deere Co., 383
U.S. 1, 12 (1966)). The Court reaffirmed principles based on its precedent
that “[t]he combination of familiar elements according to known methods is
likely to be obvious when it does no more than yield predictable results.” Id.
The operative question in this “functional approach” is thus “whether the
improvement is more than the predictable use of prior art elements according
to their established functions.” Id. at 1740.
We must determine whether or not the differences between the subject
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.
See KSR, 127 S. Ct. at 1734. Obviousness determination is not the result of
a rigid formula, and we will consider the facts of a case and the common
sense of those skilled in the art. Leapfrog Enters., Inc. v. Fisher-Price, Inc.,
485 F.3d 1157, 1161 (Fed. Cir. 2007). That is, the test for obviousness is
rather what the combined teachings of the references would have suggested
to those of ordinary skill in the art. See In re Keller, 642 F.2d 413, 425
(CCPA 1981).

Appeal 2008-2285
Application 10/321,120

11
ANALYSIS
With respect to Appellants’ arguments at pages 13-16 of the principal
Brief as applied to representative independent claims 1 and 5 on appeal, the
same arguments are presented that focus upon the claimed feature of a client
node configured to “store failed I/O requests and I/O requests issued from
the application to the server node after detection of the loss of access.” This
feature relates to past or present requests and to newly received requests
after the detection of the loss of access occurs. Notwithstanding Appellants’
arguments otherwise, the artisan would plainly understand from the
Examiner’s reliance upon Figure 9 and the discussion at columns 10 and 11
of Murphy noted in finding of fact 4 that Murphy clearly would indicate to
the artisan that these requests are stored, whose desired accessibility is
resumed once the failover to the non-failing server occurs. Moreover, we
believe the artisan would clearly understand that this was well known in the
art according to our discussion at finding of fact 3 with respect to Murphy’s
prior art teachings at column 1. To have concluded otherwise, would have
required the loss of access of the desired accessibility, which the artisan
clearly would have not permitted to happen in a dynamic network
environment as disclosed in Murphy. Based upon these teachings, we
conclude that the artisan would have necessarily determined from Murphy’s
own teachings as well as those buttressed by his own prior art that a storage
function of both types of I/O requests would have been inherent within by
the operation of Murphy contrary to the urging at the bottom of page 14 of
Appeal 2008-2285
Application 10/321,120

12
the principal Brief on appeal. As Figure 1 of Murphy shows, and as our
discussion of transparency, and as our findings of fact 2 and 4 indicate,
transparency of retry of invocations of data accessibility in a storage network
system would require this to permit the automatic recovery of failed servers
as taught.
We find misplaced Appellants’ urging at page 15 of the principal
Brief that Murphy fails to teach or suggest that the client node is configured
to detect recovery of access to the “same” server node. Representative
independent claim 1 does not require accessibility to the same server node.
Moreover, Murphy’s Figure 2B also illustrates the highly available server
211, which the artisan may well view as a single or the same server to the
extent argued since it includes two redundant servers. As essentially
admitted by Appellants at the bottom of page 15 of this Brief, it “is common
parlance in the art of highly available server systems to refer to a cluster of
server nodes as a highly available server since the cluster logically appears
as a single server to clients even though the highly available server includes
multiple server nodes.” The virtual nature of this transparency is
unpersuasively repeated at page 5 of the Reply Brief.
As our overall discussion of Murphy in findings of fact 2 through 4
would indicate to the artisan, the feature of dependent claim 6 as a
representative claim argued at page 17 of the principal Brief on appeal, is
met since the claimed repeated accessibility is clearly noted as a concept in
the art, particularly in the discussion noted in finding of fact 3 relative to the
Appeal 2008-2285
Application 10/321,120

13
prior art at column 1 of Murphy. The artisan would understand the repeated
accessibility would be tried until the time out period would expire.
In this respect then the discussion of the Examiner’s third stated
rejection regarding a corresponding feature recited in dependent claim 2 and
argued at page 18 of the principal Brief, in not disputing the teachings of
Gatto (see finding of fact 5), the allegation is made that Murphy teaches
away from a client node repeatedly sending one of the failed I/O request to a
server node until a failed I/O request succeeds. The combinability of
Murphy, Purcell, and Gatto is not explicitly contested other than the
unsubstantiated urging that Murphy is alleged to teach away from the
concept. Our earlier findings in this opinion indicate otherwise. “A
reference may be said to teach away when a person of ordinary skill, upon
[examining] the reference, would be discouraged from following the path set
out in the reference, or would be lead in a direction of divergent from the
path that was taken by applicant.” Para-Ordnance Mfg., Inc. v. SGS
Importers Int’l, Inc., 73 F.3d 1085, 1090 (Fed. Cir. 1995) (quoting In re
Gurley, 27 F.3d 551, 553 (Fed. Cir. 1994)).
Within the first stated rejection the arguments presented as to
dependent claim 3 at page 16 of principal Brief are also misplaced.
Although we tend to agree with the Appellants’ position that Murphy does
not explicitly teach direct access file system, Appellants own disclosure, as
set forth in finding of fact 1, indicates that the so-called SAN system, upon
which Appellants disclosed invention is based, was well known in the art to
Appeal 2008-2285
Application 10/321,120

14
provide such a feature. This so-called SAN fabric mentioned here is
acknowledged to be a part of the Purcell system as indicated in finding of
fact 5. The Appellants’ arguments do not contest the proper combinability
of Purcell and Murphy within 35 U.S.C. § 103.
Next, we address the features of the Examiner’s second stated
rejection of dependent claim 4 as representative of the claims rejected here.
This claim merely requires that “the application is an Oracle Parallel Server
(OPS) application.” This is precisely what Oracle teaches as noted in
finding of fact 6. Appellants’ arguments at pages 19 and 20 of the principal
Brief appear to us to be mere speculation. The end of finding of fact 4
clearly indicates that the Murphy’s operating system is a UNIX- based
system for which the desirability and advantages outlined by the Examiner
in the Answer are clearly compatible with the Oracle UNIX clusters. The
corresponding arguments at pages 7 and 8 of the Reply Brief are
unpersuasive because it is Oracle that teaches its applicability to the UNIX-
based system of Murphy. Appellants have merely utilized what is well
known to be useable within the prior art.

CONCLUSION OF LAW
Appellants have not shown that the Examiner erred in finding that the
combination of teachings of the applied prior art teaches the combination of
features of the respectively rejected, representative claims within 35 U.S.C.
§ 103.
Appeal 2008-2285
Application 10/321,120

15
DECISION
The Examiner’s three separately stated rejections under 35 U.S.C.
§ 103 of claims 1-25 on appeal are affirmed. No arguments are presented to
us other than with respect to the claims discussed in this opinion. Claims 1-
25 are unpatentable.
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a).

AFFIRMED

rwk

MEYERTONS, HOOD, KIVLIN, KOWERT & GOETZEL, P.C.
P.O. BOX 398
AUSTIN, TX 78767-0398