Software AG USA, Inc.

Patent Trials and Appeals Board

Dec 8, 2020

2019005354 (P.T.A.B. Dec. 8, 2020)

UNITED STATES PATENT AND TRADEMARK OFFICE
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O. Box 1450
Alexandria, Virginia 22313-1450
www.uspto.gov
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
14/449,517 08/01/2014 Gagan MEHRA AC-5135-130 6384
23117 7590 12/08/2020
NIXON & VANDERHYE, PC
901 NORTH GLEBE ROAD, 11TH FLOOR
ARLINGTON, VA 22203
EXAMINER
DAVANLOU, SOHEILA
ART UNIT PAPER NUMBER
2153
NOTIFICATION DATE DELIVERY MODE
12/08/2020 ELECTRONIC
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the
following e-mail address(es):
PTOMAIL@nixonvan.com
pair_nixon@firsttofile.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte GAGAN MEHRA and MANISH DEVGAN

Appeal 2019-005354
Application 14/449,517
Technology Center 2100
Before ERIC S. FRAHM, JOHNNY A. KUMAR, and
MICHAEL T. CYGAN, Administrative Patent Judges.

KUMAR, Administrative Patent Judge.
DECISION ON APPEAL
STATEMENT OF THE CASE
Claims 1–25 are pending, stand rejected, appealed by Appellant,1 and
the subject of our decision under 35 U.S.C. § 134(a). See Final Act. 1. We
have jurisdiction under 35 U.S.C. § 6(b).
We REVERSE.
Illustrative Claim
Independent claim 1 illustrates the invention as follows, with disputed
elements highlighted in italics:
1. A method of performing a map reduce sequence in
connection with a plurality of computer nodes in a distributed

1 We use the word Appellant to refer to “applicant” as defined in 37 C.F.R.
§ 1.42(a). Appellant identifies the real party in interest as Software AG.
App. Br. 3.
Appeal 2019-005354
Application 14/449,517

2
network system, each of the plurality of computer nodes
including at least one hardware processor, the distributed
network system including volatile storage, the method
comprising:
executing, on multiple ones of the plurality of computer
nodes, a map job, as part of the map reduce sequence, by using
input data to generate, from each of the respectively executed
map jobs, a plurality of intermediate output elements, each said
intermediate output element including a key-value pair, where
each of the respectively executed map jobs is identified by a
corresponding map task identifier and each intermediate output
element includes substantive data generated as a result of
executing the map job using the input data;
writing, for each one of the generated plurality of
intermediate output elements, the corresponding intermediate
output element as the value of a corresponding in memory key-
value pair that is stored in the volatile storage of the distributed
network system, where the key for the corresponding in-memory
key-value pair is based on a combination of a map task identifier
for the map job that generated the corresponding intermediate
output element and a reduce task identifier for a reduce job;
wherein a plurality of reduce jobs of the map reduce
sequence are configured to execute on multiple ones of the
plurality of computer nodes, the execution of each of the plurality
of reduce jobs configured to, upon receipt of electronic
instructions:
for each reduce job, retrieve, from the volatile
storage of the distributed network system, values of multiple
corresponding in-memory key-value pairs that have keys that are
based at least in part on a reduce task identifier assigned to the
corresponding reduce job,
execute the corresponding reduce job, using at least one
processor of the respective computer node, using the intermediate
output elements that are included in the respective values of the
multiple corresponding in-memory key-value pairs, and output a
result from the reduce job for the map reduce sequence.
App. Br. 27, 28, Claims Appendix.
Appeal 2019-005354
Application 14/449,517

3
Rejections on Appeal
1. Claims 1, 2, 6, 7, 9–18, 21, and 23–25 stand rejected under 35 U.S.C.
§ 102 as allegedly being anticipated by Liu (U.S. Publication No.
2012/0317579 A1; Dec. 13, 2012).
2. Claims 3–5, 8, 19, 20, and 22 stand rejected under 35 U.S.C. § 103 as
allegedly being obvious over Liu and Harris et al. (U.S. Publication
No. 2012/0222005 A1; Aug. 30, 2012) (“Harris”).
ANALYSIS
Appellant asserts, among other things, the claimed elements in
independent claim 1 “where the key for the corresponding in-memory key-
value pair is based on a combination of a map task identifier for the map job
that generated the corresponding intermediate output element and a reduce
task identifier for a reduce job,” (hereinafter “the disputed limitation”) are
not taught by Liu. See App. Br. 14–19.2
The Examiner determines:
Liu [0027] - [0028] teaches the argued limitation “where the key
for the corresponding in-memory key-value pair is based on a
combination of a map task identifier for the map job that
generated the corresponding intermediate output element and a
reduce task identifier for a reduce job”, because the hash function
takes both (e.g. a combination) the node ID and map ID pair (e.g.
a map task identifier) and “processing node id, sub-task ID” (e.g.
a reduce task identifier) and generates a value (e.g. a hashing
key-value).
As shown above Liu teaches:
• A hash function taking the combination data (e.g., output)
and/or metadata (e.g., first key (processing node id, key value

2 Although Appellant makes other arguments in the Briefs, we do not
address them because we find this argument is dispositive of the appeal.
Appeal 2019-005354
Application 14/449,517

4
pair input) and second key (reduce sub-task ID), and generating
a value (e.g. “key”) corresponding to an identification of the
reduce queue storing the intermediate output.
[0029] In one embodiment, shuffling occurs as soon as the results
are available and aggregated into a single output consistent with
and corresponding to the original input (map) task. According to
such implementations, data shuffling may be performed at least
partially overlapping with map processing, thereby further
reducing total processing time required. When a reduce node
103a, 103b completes a reduce task, it writes two pieces of
information to the persistent storage service 113: the node ID and
reduce ID pair, and the number of output key-value pairs the
node generated while processing reduce queue.
As shown above Liu teaches generating a “key” that the “key” is
based on both the a map task identifier and the a reduce task
identifier.
Final Act. 20. (emphases added, original emphases omitted).

Appellant contends “Liu does not expressly or inherently disclose
such elements and in particular does not disclose constructing a key for a
key value pair based on both map and reduce task identifiers.” App. Br. 15.
We agree with Appellant (App. Br. 16–18) that the Examiner-cited
portions of Liu (see ¶¶ 27, 28, and 29) do not explicitly or inherently
disclose the disputed limitation as required by Appellant’s claim 1. In other
words, the Examiner has not shown that the key is based on a combination
of a map task identifier and a reduce task identifier because Liu discloses
writing two pieces of information that include the identifiers, but does not
disclose a key-value couple where the key is based on both identifiers that
are paired with a particular value. In particular, we find that the Examiner
has not explained how the key is based on the map task identifier. At best
the Examiner has shown that the value is based on the map task identifier,
Appeal 2019-005354
Application 14/449,517

5
but the key is only stated to be an identification of the reduce queue storing
the intermediate output.
To anticipate a prior art reference must “disclose all elements of the
claim within the four corners of the document, and it must disclose those
elements arranged as in the claim.” Microsoft Corp. v. Biscotti, Inc., 878
F.3d 1052, 1068 (Fed. Cir. 2017) (internal quotations and citations omitted).
See Richardson v. Suzuki Motor Co., 868 F.2d 1226, 1236 (Fed. Cir. 1989);
Verdegaal Bros. v. Union Oil Co. of California, 814 F.2d 628, 631 (Fed. Cir.
1987); MPEP § 2131.
Consequently, we are constrained by the record before us to find that
the Examiner erred in finding Liu anticipates Appellant’s claim 1.
Independent claims 13, 15, and 16 include limitations of commensurate
scope. Since we reverse the rejections of independent claims 1, 13, 15, and
16 on appeal, we also reverse the rejection of the corresponding dependent
claims.

CONCLUSION
The Examiner erred in rejecting claims 1, 2, 6, 7, 9–18, 21, and 23–25
under 35 U.S.C. § 102(a) and claims 3–5, 8, 19, 20, and 22 under 35 U.S.C.
§ 103(a).

Appeal 2019-005354
Application 14/449,517

6

DECISION SUMMARY
In summary:

REVERSED

Claim(s)
Rejected
35 U.S.C.
§
Reference(s) Affirmed Reversed
1, 2, 6, 7,
9–18, 21,
23–25
102(a) Liu 1, 2, 6, 7,
9–18, 21,
23–25
3–5, 8, 19,
20, 22
103 Liu, Harris 3–5, 8, 19,
20, 22
Overall
Outcome
1–25