Ex Parte Karuppiah

Patent Trial and Appeal Board

Mar 31, 2014

11758860 (P.T.A.B. Mar. 31, 2014)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

Ex parte RAJKUMAR KARUPPIAH
____________

Appeal 2011-011116
Application 11/758,860
Technology Center 2100
____________

Before JENNIFER D. BAHR, JOHN C. KERINS and
ANNETTE R. REIMERS, Administrative Patent Judges.

REIMERS, Administrative Patent Judge.

DECISION ON APPEAL

Appeal 2011-011116
Application 11/758,860

2
STATEMENT OF THE CASE1,2
Rajkumar Karuppiah (Appellant) appeals under 35 U.S.C. § 134(a)
from the Examiner’s decision to reject under 35 U.S.C. § 103(a): (1) claims
5-8 as unpatentable over Alur (US 2005/0015641 A1; pub. Jan. 20, 2005),
Allen (US 5,381,145; iss. Jan. 10, 1995) and Folkert (US 7,734,602 B2; iss.
Jun. 8, 2010); and (2) claims 9-12 as unpatentable over Alur, Iwatsu (US
2006/0198518 A1; pub. Sep. 7, 2006) and Strasser (US 2003/0126247 A1;
pub. Jul. 3, 2003). Claims 1-4 have been cancelled. We have jurisdiction
under 35 U.S.C. § 6(b).
We AFFIRM-IN-PART.
CLAIMED SUBJECT MATTER
The claimed subject matter “relates generally to distributed computer
systems, and more specifically to backing up data from a production server
to a backup server.” Spec.1, ll. 2-3; figs. 3A, 3B. Claims 5, 7, 9 and 11 are

1 In the Appeal Brief, Appellant states that “Appellant[] filed a (first) Appeal
Brief on September 23, 2010. The Examiner responded with a Final
Rejection mailed on December 13, 2010, citing three new references . . . in
another 35 USC 103(a) rejection. The Examiner asserted that the Finality of
the Rejection was proper because ‘Applicant’s amendment necessitated the
new ground(s) of rejection’. There was no such Amendment. If the
Examiner is referring to the ‘minor’ Rule 116 Amendment in which
Applicant changed ‘computer readable storage media’ to ‘computer readable
tangible storage device’ of September 1, 2010, this was insufficient grounds
to necessitate the Final Rejection.” App. Br. 1. Appellant’s argument is
directed to a petitionable matter and not an appealable matter and is not
within the jurisdiction of the Board. See In re Mindick, 371 F.2d 892, 894
(CCPA 1967).
2 Throughout this opinion, we shall refer to the Appeal Brief filed April 20,
2011, as “App. Br.”
Appeal 2011-011116
Application 11/758,860

3
independent. Claims 5 and 9 are illustrative of the claimed subject matter
and recite:
5. A method for backing up one or more data files at
a backup server that correspond to one or more data files
at a primary server, the backup server being coupled to
the primary server via a network, the method comprising
the steps of:
the backup server receiving from the primary
server via the network a multiplicity of compressed files
which include data updates to the one or more data files
at the primary server for update to the corresponding one
or more data files at the backup server;
the backup server determining if the backup server
has more than a predetermined level of available
processor power to decompress two or more of the
compressed files at least partially in parallel, and
if so, the backup server decompressing two or
more of the compressed files at least partially in parallel,
and
if not, the backup server decompressing the two or
more compressed files sequentially; and
the backup server applying data updates from the
decompressed data files to the corresponding one or more
data files in the backup server.

9. A method for processing two or more data update
files at a primary server for transmission to a backup
server, the two or more data update files including data
updates to one or more data files stored at both the
primary server and the backup server, the backup server
being coupled to the primary server via a network, the
method comprising the steps of:
the primary server determining that the two or
more data update files at the primary server are waiting
to be compressed;
the primary server determining if the primary
server has more than a predetermined level of available
Appeal 2011-011116
Application 11/758,860

4
processor power and if the network has more than a
predetermined level of available bandwidth, and
if so, the primary server compressing the two or
more data update files at least partially in parallel, and
if not, the primary server compressing the two or
more data update files sequentially; and
the primary server sending to the backup server via
the network the two or more data update files which have
been compressed so the backup server can update the
corresponding one or more data files at the backup
server.

ANALYSIS
Obviousness over Alur, Allen and Folkert
Claims 5 and 7
Appellant does not present arguments for independent claim 7
separate from those presented for independent claim 5. App. Br. 7-17.
Accordingly, Appellant has argued claims 5 and 7 as a group for purposes of
the rejection of those claims under § 103(a). Claim 5 is representative of the
group and is selected for review, with claim 7 standing or falling with claim
5. See 37 C.F.R. § 41.37(c)(1)(vii) (2011).
Independent claim 5 calls for a method for backing up one or more
data files at a backup server including the step of “the backup server
determining if the backup server has more than a predetermined level of
available processor power to decompress two or more of the compressed
files at least partially in parallel.” App. Br. 30, Clms. App’x.
The Examiner found that Alur discloses the limitations of claim 5,
including the step of “the backup server determining if the backup server has
more than a predetermined level of available processor power to decompress
two or more of the compressed files at least partially in parallel” (citing
Appeal 2011-011116
Application 11/758,860

5
Alur, para. [0077]) except “Alur does not explicitly teach decompressing the
two or more compressed files.” Ans. 4. The Examiner relied on Allen for
disclosing “decompressing the two or more compressed files.” Id. (citing
Allen, Abstract). The Examiner concluded that
[i]t would have been obvious to one of ordinary
skill in the art at the time of invention to modify
the teachings of Alur to include file decompression
as disclosed by Allen. Since Alur already
disclos[es] compression of the backup files,
therefor[e] using decompression during restore
procedures would allow original data to be
reconstructed.
Id. at 4-5 (citing Allen, col. 20, ll. 6-9). The Examiner further found that
if Alur does not explicitly teach the backup server
determining if the backup server has more than a
predetermined level of available processor power
to decompress two or more of the compressed files
at least partially in parallel . . . Folkert discloses
the same in Col.21 lines 33-67 - Col.22 lines 33-
50.
Id. at 5 (emphasis omitted). The Examiner concluded that
[i]t would have been obvious to one of ordinary
skill in the art at the time of invention to modify
the teachings of Alur to include resource based
processing as disclosed by Folkert. Doing so
would allow optimal technologies and optimal
schedules in . . . light of changing business
objectives to deliver desired QoS [quality of
service] (Alur par.[0075]) and provide optimal
allocation of resources[.]
Id. (citing Folkert, Abstract).
Appellant contends that “[n]one of the cited references teaches or
suggests that the decision whether to decompress two or more files in
Appeal 2011-011116
Application 11/758,860

6
parallel or sequentially is based on the level of available processor power.”
App. Br. 8. Specifically, Appellant contends that
Alur et al. pertains to dynamic allocation of
backup and recovery resources based on
importance of the data; the most important system
gets the highest level of service and is recovered
more quickly. Policies specify the importance of
the data and therefore, the amount of resources
dedicated to the backup and recovery operations.
Alur et al. also teaches that fast recovery time
might use backup parallel, restore parallel, and roll
forward parallel, although does not sufficiently
explain what this means. Alur et al. does not teach
or suggest that the decision whether to decompress
two or more files in parallel or sequentially is
based on the level of available processor power.
Id. at 10.
Allen et al. pertains to decoding data in parallel,
and optimizing the process by utilizing two or
more decoders at a potential bottleneck in the
process, i.e. a feedback loop from a context model.
Different decoders are dedicated to different
context bins, and therefore, the context model is
not required to associate coded data with a
particular context bin. This avoids the need/wait
for feedback from the context model. Without the
need/wait for the feedback from the context model,
the decoding process is expedited. But, the manner
of decoding in Allen et al. is not based on the level
of available processing power as recited in present
claim 5. In fact, all files in Allen et al. are decoded
in the same manner, i.e. with the two or more
decoders in the feedback loop. The two or more
decoders in the feedback loop of Allen et al. are
“hardwired” and not selectively utilized based on
processor availability. Therefore, Allen et al. does
not fill the foregoing gap of Alur et al., i.e. does
not teach or even suggest that the decision whether
Appeal 2011-011116
Application 11/758,860

7
to decompress two or more files in parallel or
sequentially is based on the level of available
processor power, as recited in claim 5.
Id. at 13
Folkert et al. pertains to extracting a predetermined
set of data from a relational database at
predetermined times, and storing the data as a
materialized view (“MV”) so the data is readily
available for a query. There also may be some
processing of the data from the relational database
to generate the MV. The MV needs to be
refreshed as data in the relational database is
updated. Folkert et al. discloses parallel refreshing
of different MV entries. More specifically, Folkert
et al. also discloses parallel refreshing of different
entries in a single MV where the MV is large, and
parallel refreshing of different MVs where the
different MVs are small. However, Folkert et al.
does not teach or suggest parallel compression or
decompression. Rather, the parallel “refreshing” of
Folkert et al. is the fetching of data from the
relational database and incorporation into the MV.
While there may be some processing of the data
from the relational database to generate the MV,
there is no teaching or suggestion of compression
or decompression of this data in the process.
Folkert et al. does not teach or suggest that the
decision whether to decompress two or more files
in parallel or sequentially is based on the level of
available processor power, as recited in claim 5.
Id. at 17.
Appellant’s arguments are not persuasive. Appellant’s arguments are
individual attacks on the references rather than a challenge to the Examiner’s
combination of the teachings of the references. In re Merck & Co., 800 F.2d
1091 (Fed. Cir. 1986); In re Keller, 642 F.2d 413 (CCPA 1981) (one cannot
show nonobviousness by attacking references individually where the
Appeal 2011-011116
Application 11/758,860

8
rejections are based on combinations of references).
Folkert discloses that when parallel execution is unavailable the MVs
are executed one at time in the order in which they come. See Folkert, col.
21, ll. 35-37. In addition, Folkert discloses that new MVs (additional work)
are allocated resources as resources become available. The resource
constraints that regulate concurrent and parallel execution of MVs are the
number of concurrent processes allowed and the number of resources
available for parallel execution. See Folkert, col. 21, l. 65-col. 22, l. 2. As
such, we agree with the Examiner that Folkert “discloses dynamic
determination of parallel versus sequential processing based on available
processor power.” Ans. 20. Moreover, the Examiner relied on Allen for
disclosure of “decompressing two or more compressed files.” See Ans. 4,
20. The Examiner concluded that it would have been obvious to the skilled
artisan to modify Alur to include file decompression, as taught by Allen,
“[s]ince Alur already disclos[es] compression of the backup files, therefor[e]
using decompression during restore procedures would allow original data to
be reconstructed.” Id. at 5. The Examiner further concluded that it would
have been obvious to the skilled artisan to further modify Alur to include
resource based processing, as taught by Folkert, “[since] [d]oing so would
allow optimal technologies and optimal schedules in . . . light of changing
business objectives to deliver desired QoS [quality of service] . . . and
provide optimal allocation of resources[.]” Id. Appellant has not apprised
us of error regarding the Examiner’s findings or conclusions of obviousness.
Accordingly, for the foregoing reasons, the Examiner’s rejection of
independent claim 5 and claim 7, which falls with claim 5, as unpatentable
over Alur, Allen and Folkert is sustained.
Appeal 2011-011116
Application 11/758,860

9
Claim 6
Claim 6 calls for “the backup server beginning to decompress one of
the two or more compressed files, and before completion of the
decompression of the one compressed file.” App. Br. 31, Clms. App’x.
Appellant contends that
[t]his technique dynamically renders an accurate
assessment of the processor availability at a very
meaningful time, and is not taught or suggested by
Alur et al., Allen et al. and Folkert et al. . . . only
Allen et al. teaches parallel decoding, and this is
“hardwired” for all instances and not selectively
applied. Nor does Allen et al. teach the dynamic
decision making of claim 6. Neither Alur et al. nor
Folkert et al. teaches parallel decoding.
App. Br. 18.
Appellant’s argument is not persuasive. Appellant’s argument fails to
take into account the Examiner’s proposed combination of the teachings of
Allen and Folkert.
The Examiner relied on Allen for disclosure of beginning to
decompress one of the two or more compressed files before completion of
the decompression of the one compressed file. See Ans. 6 (citing Allen, col.
25, ll. 45-60). The Examiner further relied on Folkert for disclosure of
dynamically determining when there is enough additional processing power
to add more work (additional files to be decompressed in parallel), and when
there is additional processing power, to add those additional files in parallel.
See Ans. 22; see also Folkert, col. 21, l. 65-col. 22, l. 2. Appellant has not
apprised us of error regarding the Examiner’s stated findings or conclusions.
Accordingly, for the foregoing reasons, the Examiner’s rejection of
claim 6 as unpatentable over Alur, Allen and Folkert is sustained.
Appeal 2011-011116
Application 11/758,860

10
Claim 8
Appellant contends that
[c]laim 8 depends on claim 7 and
distinguishes over Alur et al., Allen et al. and
Folkert et al. for the same reasons that claim 7
distinguishes thereover, and further distinguishes
over Alur et al., Allen et al. and Folkert et al. for
the same reasons that claim 6 further distinguishes
thereover.
App. Br. 18. As discussed above, these arguments are not persuasive.
Accordingly, for similar reasons, the Examiner’s rejection of claim 8
as unpatentable over Alur, Allen and Folkert is sustained.
Obviousness over Alur, Iwatsu and Strasser
Claim 9
Claim 9 calls for a method for processing two or more data update
files at a primary server for transmission to a backup server including the
step of “the primary server determining if the primary server has more than a
predetermined level of available processor power.” App. Br. 34, Clms.
App’x.
The Examiner found that Alur discloses the limitations of claim 9,
including the step of “the primary server determining if the primary server
has more than a predetermined level of available processor power” (citing
Alur, para. [0077]) except “Alur does not explicitly teach if the network has
more than a predetermined level of available bandwidth.” Ans. 9-10. The
Examiner relied on Strasser for disclosing “if the network has more than a
predetermined level of available bandwidth.” Id. at 10 (citing Strasser,
Abstract). The Examiner concluded that
[i]t would have been obvious to one of
ordinary skill in the art at the time of invention to
Appeal 2011-011116
Application 11/758,860

11
modify the teachings of Alur to account for
network bandwidth as disclosed by Strasser.
Doing so would provide flexible parallel backup
that [could] use all the theoretically available
system resources such as bandwidth.
Ans. 10. The Examiner further found that “if Alur does not explicitly
[disclose] the primary server determining if the primary server has more than
a predetermined level of available processor power . . . Iwatsu discloses the
same in (par.[0094]-par.[0098).” Id. at 11 (emphasis omitted). The
Examiner concluded that
it would have been obvious to one of ordinary skill
in the art at the time of invention to choose
sequential processing instead of parallel as
disclosed by Iwatsu. Doing so would eliminate[]
the time interval during which the processing of
decryption/demodulation for the current contents
data and the tamper check processing for the next
contents data are carried out simultaneously in
parallel to prevent the processing load caused by
concurrent execution of the tamper check
processing and the processing of
decryption/demodulation from increasing.
Id.
Appellant contends that neither Alur nor Iwatsu teaches or suggests
that the decision whether to compress two or more files in parallel or
sequentially is based on the level of available processor power. See App. Br.
19, 26.
Appellant’s arguments are persuasive. Alur discloses a system to
optimize available resources to meet application data availability and
business objectives. See Alur, Abstract. Alur uses business objectives
which include “Service Level Agreements,” such as Gold, Bronze, etc., and
these Service Level Agreements define a specific level of service that the
Appeal 2011-011116
Application 11/758,860

12
provider will provide to the client. See Alur, para. [0078]. Alur discusses
the system generating intelligent and optimal schedules to deliver the desired
quality of service (QoS). See Alur, para. [0035]. Alur is not changing the
level of service to provide more when it is available, but rather is providing
the required amount based on the service level paid for by the client
(Bronze, Gold, etc.). See Alur, para. [0078]. In other words, Alur is not
providing more than the agreed-upon amount of service or determining if
more processing power beyond that included in the level of service should
be used for a specific job.
Iwatsu discusses completing multiple different tasks in parallel. See
Iwatsu, Abstract, para. [0071]. Furthermore, the portion of Iwatsu relied on
by the Examiner (i.e., paras. [0094]-[0098]) (see Ans. 11) refers to the state
of high Central Processing Unit (CPU) take-up ratio possibly causing
interruptions in the current work (reproducing output speech). The
Examiner has failed to explain how completing multiple different tasks in
parallel or the state of high CPU take-up ratio relates to “the primary server
determining if the primary server has more than a predetermined level of
available processor power” as claimed. See Ans. 11, 24-25. Moreover, we
could not find any discussion in Iwatsu pertaining to determining if the
primary server has more than a predetermined level of available processor
power. Consequently, the Examiner has failed show by a preponderance of
the evidence that the combined teachings of Alur, Iwatsu and Strasser
disclose the step of “the primary server determining if the primary server has
more than a predetermined level of available processor power,” as required
by claim 9.
Accordingly, for the foregoing reasons, the Examiner’s rejection of
Appeal 2011-011116
Application 11/758,860

13
claim 9 as unpatentable over Alur, Iwatsu and Strasser cannot be sustained.
Claim 11
Similar to independent claim 9, independent claim 11 requires
“second program instructions, for execution at the primary server, to
determine if the primary server has more than a predetermined level of
available processor power.” App. Br. 36, Clms. App’x. The Examiner
relied on Alur and/or Iwatsu for this limitation. Ans. 10. As discussed
above, neither Alur nor Iwatsu provide any disclosure relating to
determining if the primary server has more than a predetermined level of
available processor power.
Accordingly, for similar reasons, the Examiner’s rejection of
independent claim 11 as unpatentable over Alur, Iwatsu and Strasser cannot
be sustained.
Claims 10 and 12
Claim 10 depends from claim 9. App. Br. 35 Clms. App’x. Claim 12
depends from claim 11. Id. at 37. The Examiner again relies on Alur and/or
Iwatsu for disclosing the limitation “[determining/determine] if the primary
server has more than a predetermined level of available processor power.”
Ans. 10-11. The Examiner cites Strasser for its disclosure of a
predetermined level of available bandwidth. Id. at 10. This does not cure
the above-discussed deficiency in the Examiner’s citation of Alur and/or
Iwatsu as meeting the aforementioned claim limitation.
Accordingly, for the foregoing reasons, the Examiner’s rejection of
claims 10 and 12 as unpatentable over Alur, Iwatsu and Strasser cannot be
sustained.

Appeal 2011-011116
Application 11/758,860

14
DECISION
We AFFIRM the decision of the Examiner to reject claims 5-8 as
being unpatentable over Alur, Allen and Folkert.
We REVERSE the decision of the Examiner to reject claims 9-12 as
being unpatentable over Alur, Iwatsu and Strasser.
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-IN-PART

mls