Ex Parte Gordon

Board of Patent Appeals and Interferences

Mar 8, 2010

10955559 (B.P.A.I. Mar. 8, 2010)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________________

Ex parte MARK R. GORDON
____________________

Appeal 2009-004333
Application 10/955,559
Technology Center 2100
____________________

Decided: March 8, 2010
____________________

Before JAY P. LUCAS, THU A. DANG, and DEBRA K. STEPHENS,
Administrative Patent Judges.

DANG, Administrative Patent Judge.

DECISION ON APPEAL

Appeal 2009-004333
Application 10/955,559

I. STATEMENT OF CASE
Appellant appeals the Examiner’s final rejection of claims 1, 3-10,
and 12-28 under 35 U.S.C. § 134(a). Claims 2, 11, and 29 have been
canceled. We have jurisdiction under 35 U.S.C. § 6(b).
We affirm.

A. INVENTION
According to Appellant, the invention relates to “a method, system,
and computer program product for adaptive database buffer memory
management using dynamic Structured Query Language (SQL) statement
cache statistics” (Spec. 29).

B. ILLUSTRATIVE CLAIM
Claim 1 is exemplary and reproduced below:

1. A method for database buffer memory management,
comprising:

obtaining Structured Query Language (SQL) statement
cache statistics;

inferring re-use of a page in a buffer pool memory using
the SQL statement cache statistics; and

selectively freeing a memory location of the page in the
buffer pool memory based on the inferred re-use of the page.
2
Appeal 2009-004333
Application 10/955,559

C. REJECTIONS
The prior art relied upon by the Examiner in rejecting the claims on
appeal is:
Ting US 2002/0013887 A1 Jan. 31, 2002
Hayes US 2002/0046204 A1 Apr. 18, 2002
Ngai US 2005/0086263 A1 Apr. 21, 2005

Claims 1, 10, 19, 20, and 28 stand rejected under 35 U.S.C. § 103(a)
as being unpatentable over Ting in view of Ngai.
Claims 3-9, 12-18, and 21-27 stand rejected under 35 U.S.C. § 103(a)
as being unpatentable over Ting and Ngai in view of Hayes.1

II. ISSUE
Has Appellant shown the Examiner erred in concluding that the
combination of Ting and Ngai would have suggested “inferring re-use of a
page in a buffer pool memory using the SQL statement cache statistics”
(claim 1)?

III. FINDINGS OF FACT
The following Findings of Fact (FF) are shown by a preponderance of
the evidence.
Ting
1) Ting discloses a data buffer memory management system for
increasing the effectiveness and efficiency of buffer replacement selection (¶

1 A rejection of claims 9-20 under 35 U.S.C. § 101 has been withdrawn by
the Examiner (Ans. 17).

3
Appeal 2009-004333
Application 10/955,559

[0001]).
2) The system employs a page buffer hierarchy for prioritizing the
victimization of page buffers, which includes: determining a measure of the
favorability for victimizing each page buffer; and victimizing the page
buffers based on their measured favorability for victimization (¶ [0012]).
3) The measured favorability for victimizing a page buffer is
based on the page’s likelihood of being referenced in the future, e.g., the
likelihood that a page will be accessed more often than other pages (¶
[0025]).
4) To assess the measured favorability for victimizing a page
buffer, the system “predict[s] the likelihood of early re-use of the page”
based on reference counts, i.e., based on how many times the page has been
fixed since being placed in the buffer pool (¶¶ [0018] and [0025-26]).
5) The identity and reference count of a victimized page is copied
to a “history table” such that, if the page is placed back in the buffer pool,
the page’s last reference count can be used to determine the measured
favorability for victimizing the corresponding page buffer (¶ [0035]).
Ngai
6) Ngai discloses a workload repository (WR) that collects and
manages “important system performance statistics” for a database server
(Abstract; ¶ [0019]).
7) The WR includes “historical statistics tables” for storing
performance statistics (¶ [0036]), most of which are tracked via cumulative
totals (¶ [0044]).
8) According to Ngai, “SQL statistics is one of the most important
kind of statistics used for performance analysis” (¶ [0087]).
4
Appeal 2009-004333
Application 10/955,559

Appellant’s Disclosure
9) Appellant’s disclosure acknowledges that a skilled artisan
would have recognized SQL statistics as including page reference count
information, stating, “Database systems using dynamic SQL maintain a
variety of performance-related statistics . . . . [that] vary from system to
system, but generally include … page references (e.g., getpages in DB2)”
(Spec. 5, ¶ [0019]).
10) Appellant’s disclosure also acknowledges that “[p]reviously, in
systems using static SQL, there was not a straightforward way to determine
the system-wide performance statistics for an individual SQL statement”
(Spec. 6, ¶ [0022]); but “[w]ith the advent of systems based completely on
dynamic SQL, however, statistics for all executions of an individual SQL
statement are aggregated in the statement cache statistics” (id.).

IV. PRINCIPLES OF LAW
35 U.S.C. § 103(a)
“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., 550 U.S. 398, 406
(2007).
In KSR, 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. Id. at 415 (citing Graham v. John Deere Co., 383 U.S. 1, 12
5
Appeal 2009-004333
Application 10/955,559

(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. at 416.
In affirming a determination of obviousness, the Federal Circuit has
relied, in part, on an applicant’s failure to present evidence that the proposed
combination of teachings was “uniquely challenging or difficult for one of
ordinary skill in the art” or “represented an unobvious step over the prior
art.” Leapfrog Enters., Inc. v. Fisher-Price, Inc., 485 F.3d 1157, 1162 (Fed.
Cir. 2007) (citing KSR, 550 U.S. at 418-19).
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).

V. ANALYSIS
In this Decision, we have considered only those arguments actually
made by Appellant. Arguments which Appellant could have made but did
not make in the Brief have not been considered and are deemed to be
waived. See 37 C.F.R. § 41.37(c)(1)(vii).
As to independent claim 1, Appellant argues that “neither Ting nor
Ngai teach or suggest using SQL statement cache statistics to infer re-use of
a page as claimed in claim 1” (App. Br. 8); and that “[a]lthough SQL
statement cache statistics have been used for performance analysis (e.g.,
locating inefficient SQL statements, finding SQL statements having I/O
delays or serialization constraints, etc.), such SQL statement cache statistics
are not used by Ngai to infer re-use of a page” (id.). Appellant further
argues that “even if Ngai and Ting were combined” as proposed, the
6
Appeal 2009-004333
Application 10/955,559

combination “would still fail to teach or suggest using SQL statement cache
statistics to infer re-use of a page as claimed in claim 1, since this feature is
not disclosed or suggested in either of the references” (id.).
However, the Examiner finds that “Ting teaches selectively freeing a
memory location of the page in the buffer pool memory based on the
inferred re-use of the page” (Ans. 18; emphasis omitted). The Examiner
acknowledges that “Ting does not explicitly disclose obtaining Structured
Query Language (SOL) statement cache statistics,” but finds that “Ngai, on
the other hand, teaches obtaining Structured Ouery Language (SOL)
statement cache statistics” (id.; emphasis omitted).
Thus, the issue we address on appeal is whether Appellant has shown
the Examiner erred in concluding that the combination of Ting and Ngai
would have suggested “inferring re-use of a page in a buffer pool memory
using the SQL statement cache statistics” (claim 1).
By contending that “neither Ting nor Ngai teach or suggest using SQL
statement cache statistics to infer re-use of a page as claimed in claim 1”
(App. Br. 8) and that this limitation “is not disclosed or suggested in either
of the references” (id.), Appellant appears to be arguing that, individually,
Ting and Ngai do not disclose the features of claim 1. However, the
Examiner rejects claim 1 over what the combined teachings of Ting and
Ngai would have suggested to a skilled artisan. One cannot show
nonobviousness by attacking references individually where the rejections are
based on a combination of references. See Merck, 800 F.2d at 1097.
Ting discloses a system of buffer replacement selection (FF 1). The
system measures the favorability for victimizing page buffers (FF 2), which
is based on the likelihood of their pages being referenced in the future (FF
7
Appeal 2009-004333
Application 10/955,559

3). To measure the favorability for victimizing a page buffer, the system
“predict[s] the likelihood of early re-use of the page” based on its reference
counts, i.e., based on how many times the page has been fixed since being
placed in the buffer pool (FF 4). The identity and reference count of a
victimized page is recorded to a “history table” such that, if the page is
placed back in the buffer pool, the page’s last reference count can be used to
determine the measured favorability for victimizing the page buffer (FF 5).
Thus, a skilled artisan would have understood Ting’s system as
recording the identity and reference count of a page to a history table, and as
inferring the reuse of that page in a buffer pool memory from the page
reference count.
Ngai discloses a workload repository (WR) that collects “important
system performance statistics” for a database server (FF 6). The WR uses
“historical statistics tables” to store the statistics, most of which are tracked
via cumulative totals (FF 7). According to Ngai, “SQL statistics is one of
the most important kind of statistics used for performance analysis” (FF 8).
A skilled artisan would have understood Ngai as suggesting the
importance of SQL statistics. Thus, in view of Ngai, the skilled artisan
would have modified Ting’s system to include the use of SQL statistics; e.g.,
to track cumulative SQL statistics. The skilled artisan would have further
modified Ting’s system to infer the reuse of a page from SQL statistics, just
as Ting’s system infers the reuse of a page from page reference count
information.
Though Ngai does not explicitly state that SQL statistics can be used
in that manner, Appellant acknowledges that SQL statistics were known to
include page reference count information, stating, “Database systems using
8
Appeal 2009-004333
Application 10/955,559

dynamic SQL maintain a variety of performance-related statistics . . . . [that]
vary from system to system, but generally include … page references (e.g.,
getpages in DB2)” (FF 9; emphasis added). Appellant further acknowledges
that the advent of dynamic SQL made “statistics for all executions” of SQL
statements readily available (FF 10). Thus, the skilled artisan would have
understood that SQL statistics could be used to determine page reference
count information and, as suggested by Ting, used to infer the reuse of a
page from that information.
For the above reasons, the Examiner did not err in concluding that the
combination of Ting and Ngai would have suggested “inferring re-use of a
page in a buffer pool memory using the SQL statement cache statistics”
(claim 1).
Appellant further argues that “the Examiner’s rationale for combining
Ngai and Ting is faulty because SQL statement cache statistics are not used
‘in order to manage its use of storage space so as to avoid ending up with a
repository of performance data that is larger than the database that it is
capturing data about’ as alleged by the Examiner” (App. Br. 8). However,
we conclude that the substitution of one known element (SQL statistics as in
Ngai) for another (page reference counts as in Ting) would have yielded
predictable results to one of ordinary skill in the art at the time of the
invention. That is, we conclude that inferring the reuse of a page from SQL
statistics, and more particularly from their page reference count information,
is no more than a simple arrangement of old elements, with each performing
the same function it had been known to perform, yielding no more than one
would expect from such an arrangement. See KSR, 550 U.S. at 417.
Appellant has presented no evidence that inferring the reuse of a page in this
9
Appeal 2009-004333
Application 10/955,559

manner would have been “uniquely challenging or difficult for one of
ordinary skill in the art” or “represented an unobvious step over the prior
art.” Leapfrog, 485 F.3d at 1162 (citing KSR, 550 U.S. at 418-19).
Accordingly, for the above reasons, we agree with the Examiner’s
conclusion that the combination of Ting and Ngai would have suggested the
invention of claim 1. We therefore affirm the rejection of claim 1 under 35
U.S.C. § 103(a) as being unpatentable over Ting and Ngai. As Appellant
does not provide separate arguments for independent claims 10 and 19, those
claims and dependent claim 20 also fall with claim 1. We therefore also
affirm the rejection of claims 10, 19, and 20 under 35 U.S.C. § 103(a) as
being unpatentable over Ting and Ngai.
Appellant does not present further arguments with respect to the
combination of Ting, Ngai, and Hayes, but rather state that the combination
fails to address the asserted deficiencies of Ting and Ngai (App. Br. 9). We
therefore affirm the rejection of claims 3-9, 12-18, and 21-27 under 35
U.S.C. § 103(a) as being unpatentable over Ting and Ngai in view of Hayes.

VI. CONCLUSIONS
Appellant has not shown the Examiner erred in concluding that claims
1, 10, 19, 20, and 28 are unpatentable over the teachings of Ting and Ngai.
Appellant has not shown the Examiner erred in concluding that claims
3-9, 12-18, and 21-27 are unpatentable over the teachings of Ting, Ngai, and
Hayes.
10
Appeal 2009-004333
Application 10/955,559

VII. DECISION
The Examiner’s decision rejecting claims 1, 3-10, and 12-28 under
35 U.S.C. § 103(a) is affirmed.
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a)(1)(iv).

AFFIRMED

peb

HOFFMAN WARNICK LLC
75 STATE ST
14 FL
ALBANY, NY 12207
11