Ex Parte Branson et al

Patent Trial and Appeal Board

Sep 26, 2016

13755146 (P.T.A.B. Sep. 26, 2016)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR
131755,146 0113112013 Michael J. Branson
46296 7590 09/28/2016
MARTIN & ASSOCIATES, LLC
P.O. BOX548
CARTHAGE, MO 64836-0548
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
ATTORNEY DOCKET NO. CONFIRMATION NO.
ROC920100043US2 1919
EXAMINER
DAO,TUANC.
ART UNIT PAPER NUMBER
2194
NOTIFICATION DATE DELIVERY MODE
09/28/2016 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):
derekm@ideaprotect.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte MICHAEL J. BRANSON and JOHN M. SANTOSUOSSO
Appeal2015-003295
Application 13/755,146
Technology Center 2100
Before HUNG H. BUI, AMBER L. HAGY, and JOHN R. KENNY,
Administrative Patent Judges.
BUI, Administrative Patent Judge.
DECISION ON APPEAL
Appellants seek our review under 35 U.S.C. § 134(a) of the
Examiner's Final Office Action rejecting claims 1-5, which are all of the
claims pending on appeal. 1 We have jurisdiction under 35 U.S.C. § 6(b).
We REVERSE.2 Pursuant to our authority under 37 C.F.R.
§ 41.50(b), we enter A NEW GROUND OF REJECTION for independent
claim 1.
1 According to Appellants, the real party in interest is International Business
Machines Corporation. App. Br. 1.
2 Our Decision refers to Appellants' Appeal Brief (filed July 17, 2014,
"App. Br."); Reply Brief (filed December 17, 2014, "Reply Br.");
Examiner's Answer (mailed October 17, 2014, "Ans."); Final Office Action
(mailed March 7, 2014, "Final Act."); and original Specification (filed
January 31, 2013, "Spec.").
Appeal2015-003295
Application 13/755,146
STATEMENT OF THE CASE
Appellants' invention relates to "distributed computing" and
"dynamically allocating a job or a processing unit (part of a job) on a multi-
nodal, parallel computer system." Spec. i-fi-12, 5. According to Appellants,
[a Jn application or sometimes a part of an application is often
referred to as a "job." In distributed computing, a job may be
broken up into separate run time units (referred herein as
processing units) and executed on different nodes of the system.
The processing units are assigned to a node in the distributed
system by a job scheduler or job optimizer.
Id. at 5.
Independent claim 1 is illustrative of Appellants' invention, as
reproduced with disputed limitations emphasized below:
1. A computer implemented method for dynamically
changing allocation of processing units on a multi-nodal
computer system comprising the steps of:
executing an application having a plurality jobs, each with
at least one processing unit on a plurality of nodes, \'l1here each
node has at least one central processing unit
(CPU) and a memory;
collecting metrics associated with the multi-nodal
computer system, the application, the jobs, the plurality of nodes
and the processing units;
analyzing the metrics;
when there is an over utilized resource performing the
steps of:
identifying jobs affecting the over utilized resource;
assessing potential job and processing unit
permutations; and
dynamically changing the allocation of the
processing units on the compute nodes based on the
collected metrics by splitting a job running on a first node
into multiple jobs allocated to different nodes of the
plurality of nodes,
2
Appeal2015-003295
Application 13/755,146
wherein interprocess communication between processing
units of the job is within a node before splitting the job, and after
splitting the job the interprocess communication is handled by
the multi-nodal computer system over a communication link
between the different nodes.
App. Br. 20 (Claims App. (emphasis added)).
Examiner's Rejections and References
(1) Claim 1 stands rejected under 35 U.S.C. § 103(a) as being
unpatentable over Barsness et al., (US 2009/0158276 Al, published June 18,
2009, "Barsness") and Abu-Ghazaleh et al. (US 2009/0083390 Al,
published Mar. 26, 2009, "Abu"). Final Act. 3-5.
(2) Claim 2 stands rejected under 35 U.S.C. § 103(a) as being
unpatentable over Barsness, Abu, and Barsness et al., (US 2009/031636 Al;
published Dec. 17, 2009, "Barsness '636"). Final Act. 5---6.
(3) Claims 3-5 stand rejected under 35 U.S.C. § 103(a) as being
unpatentable over Barsness, Abu, and Votta et al., (US 2010/0011254 Al;
published Jan. 14, 2010, "Votta"). Final Act. 6-9.
ANALYSIS
§ 103(a) Rejection of Claim 1 based on Barsness and Abu
With respect to claim 1, the Examiner finds Barsness teaches a
computer implemented method including the disputed limitations:
(1) "dynamically changing []allocation of []processing units on the compute
nodes based on the collected metrics by splitting a job running on a first
node into multiple jobs allocated to different nodes of the plurality of
nodes;" (2) "wherein interprocess communication between processing units
3
Appeal2015-003295
Application 13/755,146
of the job is within a node before splitting the job[,] and after splitting the
job the interprocess communication is handled by the multi-nodal computer
system over a communication link between the different nodes." Final Act.
3, 4 (citing Barsness i-fi-19, 23, 28, and 43, Figs. 2, and 6-8) (emphasis
added).
The Examiner acknowledges Barsness does not expressly teach, but
relies on Abu for teaching "an application having a plurality of jobs" in
order to support a conclusion of obviousness, i.e., "provide for the purpose
of discovering and monitoring efficiency resources." Id. at 4--5 (citing Abu
i-f 18, Figs. 2 and 6-7).
Appellants do not dispute the Examiner's factual findings regarding
Abu. Nor do Appellants challenge the Examiner's rationale for combining
Barsness and Abu. Instead, Appellants contend the Examiner erred in
finding that Barsness teaches or suggests the disputed limitations:
(1) "dynamically changing allocation of processing units on the compute
nodes based on the collected metrics by splitting a job running on a first
node into multiple jobs allocated to different nodes of the plurality of
nodes;" and (2) its related wherein clause, i.e., "wherein interprocess
communication between processing units of the job is within a node before
splitting the job, and after splitting the job the interprocess communication is
handled by the multi-nodal computer system over a communication link
between the different nodes" as recited in claim 1. App. Br. 5-9; see also
Reply Br. 3-7.
In particular, Appellants acknowledge Barsness teaches dynamic
distribution [allocation] of I/O nodes on a multi-node computer system, but
argue (1) Barsness' disclosure relates to "the allocation of [I/OJ nodes, not
4
Appeal2015-003295
Application 13/755,146
jobs" and (2) "allocating an additional l/O node to a block of nodes does not
split a job [into multiple jobs] that is executing" in the manner recited by
Appellants' claim 1. App. Br. 6-7 (citing Fig. 6-7); see also Reply Br. 3.
According to Appellants, "Claim 1 is concerned with the allocation of jobs
and Barsness is concerned with the allocation of nodes." Reply Br. 3--4.
The Examiner does not address the distinction between: ( 1) the
dynamic allocation of I/O nodes to a job as disclosed by Barsness; and
(2) the dynamic allocation of a job's processing units into multiple jobs to be
executed at different nodes in the manner recited in Appellants' claim 1.
Instead, the Examiner responds that: (1) Barsness' Figure 6 states "Job with
Extensive I/O is Dynamically Allocated An Additional I/O Node;"
(2) Barsness' Figure 7 states "Job Executing On the 16 nodes Now with 2
I/O Nodes;" and (3) paragraphs 28 and 43 of Barsness describe the use of a
job scheduler 142 to allocate and schedule work, including "load[ing] a job
record 144 from data storage 138 for placement on the compute nodes" and
using "[t]he information in the job record ... for dynamically allocating the
jobs." Ans. 5---6 (citing Barsness i-fi-128, 43, and Figs. 6-7) (emphasis added).
Based on the teachings of Barsness, the Examiner reasons that: (1) "the job
[of Barsness must be] executed by two or more nodes" and "in the context of
multi-nodal parallel scalable Blue Gene system [of Barsness], the job must
be divided/portioned into parts to be distributed/scheduled and executed into
two allocated I/O nodes including their associated computing nodes." Id. at
6.
We do not agree with the Examiner's reasoning. As correctly
recognized by Appellants, Barsness teaches dynamic allocation of I/O nodes
to a job, and not allocation of jobs. App. Br. 6 (citing Barsness i-fi-137-38,
5
Appeal2015-003295
Application 13/755,146
Figs. 6-7). Contrary to the Examiner's characterization, Figures 6 and 7 of
Barsness only illustrate the dynamic allocation of I/O nodes in a parallel
computer system. Similarly, paragraphs 28 and 43 of Barsness refer to
(1) information as part of job record 144 stored at compute node 140, shown
in Figure 1 of Barsness, and (2) that information can be "used for
dynamically allocating the jobs." Barsness i-fi-128, 43. At best, Barsness
suggests that ajob(s) can be allocated to different nodes. However,
Barsness does not teach or suggest the manner in which a job is
divided/partitioned to different nodes. Certainly, Barsness does not teach or
suggest the disputed limitations of Appellant's claim 1, including:
( 1) dynamically changing allocation of processing units on the compute
nodes based on the collected metrics by splitting a job running on a first
node into multiple jobs allocated to different nodes of the plurality of
nodes;" and (2) "wherein interprocess communication between processing
units of the job is within a node before splitting the job, and after splitting
the job the interprocess communication is handled by the multi-nodal
computer system over a communication link between the different nodes."
Because Barsness fails to teach or suggest the disputed limitations of
claim 1, we do not sustain the Examiner's obviousness rejection of claim 1
and its dependent claims 2-5.
6
Appeal2015-003295
Application 13/755,146
NEW GROUND OF REJECTION:
35 U.S.C. § 103(a)- CLAIM 1
In addition to the Examiner's factual findings regarding Barsness as
outlined in the record (Final Act. 3-5; see also Ans. 5---6), we also find
Barsness teaches: (1) applications and jobs are interchangeable, and an
application can have multiple jobs (Barsness i-f 8); and (2) in addition to
allocating I/O nodes to a job, the jobs can also be dynamically allocated
based on "the job description, job execution history, and application control
parameters" (Barsness i-f 43).
We further find, moreover, the concept of dividing a job into multiple
parts (i.e., jobs or processing units) at an originating node and then
distributing those parts across different (destination) nodes to be executed in
parallel is well known in the distributed computing art. For example, Votta
teaches, during distributed parallel computation:
[A] large job is divided into N chunks, which are then distributed
across N nodes to run in parallel. Because the chunks complete,
the overall job is completed in approximately I/Nth the amount
of time than when the job is run on one node.
Votta ,-r 80.
Based on the teachings of Barsness, and Votta, we find an ordinarily
skilled artisan would have sufficient technical knowledge and found it
obvious to dynamically change allocation of a job's processing units based
on collected metrics by splitting a job running on a first [originating] node
into multiple jobs allocated to different [destination] nodes in the manner
recited in Appellants' claim 1.
Accordingly, using our authority under 37 C.F.R. § 41.50(b), we enter
a new ground of rejection for claim 1 as being obvious over Barsness and
7
Appeal2015-003295
Application 13/755,146
Votta. With respect to remaining dependent claims 2-5, we leave these
claims to the Examiner to evaluate whether they are independently and
patentably distinct from Appellants' independent claim 1.
DECISION
We REVERSE the Examiner's Final rejection of claims 1-5 under
35 U.S.C. § 103(a) based on Barsness, Abu, Barsness '636, and Votta.
However, we enter a NEW GROUND OF REJECTION for claim 1 as being
unpatentable under 35 U.S.C. § 103(a) over Barsness and Votta.
This decision contains a new ground of rejection pursuant to 37
C.F.R. § 41.50(b) (2010). 37 C.F.R. § 41.50(b) provides "[a] new ground of
rejection pursuant to this paragraph shall not be considered final for judicial
review."
Rule 37 C.F.R. § 41.50(b) also provides that Appellant, WITHIN
TWO MONTHS FROM THE DATE OF THE DECISION; must exercise
one of the following two options with respect to the new grounds of
rejection to avoid termination of the appeal as to the rejected claims:
(1) Reopen prosecution. Submit an appropriate
amendment of the claims so rejected or new evidence
relating to the claims so rejected, or both, and have the
matter reconsidered by the examiner, in which event the
prosecution will be remanded to the examiner ....
(2) Request rehearing. Request that the proceeding
be reheard under § 41.52 by the Board upon the same
record ....
8
Appeal2015-003295
Application 13/755,146
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a)(l )(iv).
REVERSED
37 C.F.R. § 41.50(b)
9