Ex Parte Aviles et al

Patent Trial and Appeal Board

Feb 3, 2015

12015197 (P.T.A.B. Feb. 3, 2015)

UNITED STATES PATENT AND TRADEMARK OFFICE
________________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
________________

Ex parte JOAQUIN J. AVILES,
MARK U. CREE, and GREGORY A. DAHL1
________________

Appeal 2012-007325
Application 12/015,197
Technology Center 2400
________________

Before JEAN R. HOMERE, JASON V. MORGAN, and
NATHAN A. ENGELS, Administrative Patent Judges.

MORGAN, Administrative Patent Judge.

DECISION ON APPEAL
Introduction
This is an appeal under 35 U.S.C. § 134(a) from the Examiner’s final
rejection of claims 1–21. We have jurisdiction under 35 U.S.C. § 6(b).
We AFFIRM-IN-PART.
Invention
Appellants invented a cache appliance that transparently monitors
traffic between clients and network attached storage subsystems and caches
files using dynamically adjustable cache policies. Abstract.

1 CacheIQ, Inc., is the Real Party in Interest. Br. 1.
Appeal 2012-007325
Application 12/015,197

2

Exemplary Claims
Claims 1, 5, and 13, reproduced below with key limitations
emphasized, are exemplary:
1. A standalone cache unit for caching data operations
requested from one or more networked data storage devices by
one or more remote clients, the cache unit comprising:
a cache memory for caching data at the standalone cache
unit that is requested by a remote client from the one or more
networked data storage devices;
a high-speed packet processor at the standalone cache
unit coupled to the cache memory and to one or more I/O ports
for splicing connections between the data storage devices and
remote clients, where the high-speed packet processor inspects
network protocol traffic state parameters received on the I/O
ports to determine if a request from a remote client can be
serviced by data from the one or more networked data storage
devices that is stored in the cache memory.

5. The standalone cache unit of claim 4, where the
predetermined cache policy [used by software, executed by the
high-speed packet processor of claim 1, that transparently
monitors network file system and common internet file system
traffic between remote clients and network attached storage
subsystems and caches] comprises a least recently used with
dual time reference algorithm aided with greedy dual size
frequency algorithm.

13. The method of claim 10, where splicing connections
[between a remote client and one or more networked data
storage devices] comprises splicing only a connection state, a
source sequence number and a destination sequence number in
an L4 layer.
Appeal 2012-007325
Application 12/015,197

3

Rejections

The Examiner rejects claims 1–4, 6–12, 14, and 16–21 under 35
U.S.C. § 103(a) as being unpatentable over Jungck (US 7,330,908 B2; Feb.
12, 2008) and Huang (US 2003/0169749 A1; Sept. 11, 2003). Ans. 4–11.
The Examiner rejects claims 5 and 15 under 35 U.S.C. § 103(a) as
being unpatentable over Jungck, Huang, and Sarukkai (US 6,775,695 B1;
Aug. 10, 2004). Ans. 11–12.
The Examiner rejects claim 13 under 35 U.S.C. § 103(a) as being
unpatentable over Jungck, Huang, and Nielsen (US 2007/0124577 A1; May
31, 2007). Ans. 12.
ISSUES
1. Did the Examiner err in finding the combination of Jungck and
Huang teaches or suggests: (1) “a cache memory for caching data at the
standalone cache unit that is requested by a remote client from the one or
more networked data storage devices” and (2) “where the high-speed packet
processor inspects network protocol traffic state parameters received on the
I/O ports to determine if a request from a remote client can be serviced by
data from the one or more networked data storage devices that is stored in
the cache memory,” as recited in claim 1?
2. Did the Examiner err in finding the combination of Jungck, Huang,
and Sarukkai teaches or suggests “where the predetermined cache policy
comprises a least recently used with dual time reference algorithm aided
with greedy dual size frequency algorithm,” as recited in claim 5?
3. Did the Examiner err in finding the combination of Jungck, Huang,
and Nielsen teaches or suggests “where splicing connections comprises
Appeal 2012-007325
Application 12/015,197

4

splicing only a connection state, a source sequence number and a destination
sequence number in an L4 layer,” as recited in claim 13?
ANALYSIS
Except as they pertain to the rejection of claim 13, we agree with and
adopt as our own the Examiner’s findings of facts and conclusions as set
forth in the Answer and in the action from which this appeal was taken. We
have considered Appellants’ arguments with respect to the Examiner’s
rejections of claims 1–12 and 14–21, but do not find them persuasive. We
provide the following explanation primarily for emphasis.
Claims 1–4, 6–12, 14, and 16–21
In rejecting claim 1, the Examiner finds Jungck’s edge server 402,
which can intercept and modify DNS (domain name server) requests,
teaches or suggests a server with a cache and with a high-speed packet
processor that “inspects network protocol traffic state parameters received
on the I/O ports to determine if a request from a remote client can be
serviced by data from the one or more networked data storage devices that is
stored in the cache memory.” Ans. 4–5 (citing, e.g., Jungck col. 17, ll. 32,
47, col. 18, ll. 1–5, and col. 19, ll. 8–17). The Examiner acknowledges
“Jungck does not explicitly teach a standalone cache unit,” and thus relies on
Huang’s caching method on an intelligent router to teach or suggest caching
on a standalone unit. Ans. 5 (citing Huang ¶ 95). The Examiner concludes
it would have been obvious to an artisan of ordinary skill to modify Jungck
using Huang’s teachings and suggestions to “efficiently handle requests
without having to route them to dedicated caches.” Ans. 5.
Appellants contend the Examiner erred because Jungck’s disclosure of
a cache is “nothing more than a generic reference to a cache memory which
Appeal 2012-007325
Application 12/015,197

5

serves the frequently used data used by the Edge Server 402” rather than
“data that is requested by a remote client from one or more networked data
storage devices, as claimed.” Br. 6. Appellants argue that Jungck, rather
than caching requested data on edger server 402, merely routes or forwards
remote cache operations to geographically dispersed cache servers 302, 304.
Id. (citing Jungck col. 19, l. 20–col. 20, l. 10). Appellants contend Huang
does not cure the alleged deficiency of Jungck because “Huang’s disclosure
is merely cumulative of Jungck’s disclosure in describing a generic cache
memory which serves the frequently used data used by the intelligent
router.” Br. 8. Appellants argue “Huang’s disclosed ‘caching method 320’
refers generically to caching a ‘message having data or content, a channel ID
and subjects (step 322)’ without ever indicating that the cached data/content
is data that is ‘requested by a remote client from the one or more networked
data storage devices.[’]” Id. (citing Huang’s Fig. 14, ¶ 95).
Appellants unpersuasively attack Jungck and Huang individually,
rather than showing error in the Examiner’s reliance on their combined
teachings and suggestions. The Examiner acknowledges Jungck teaches or
suggests a server that routes DNS “requests to geographic friendly cache
servers, i.e.[,] the caching and inspecting are not performed at a singular
entity in Jungck.” Ans. 13. However, the Examiner correctly finds Huang
teaches or suggests the local caching of data. Id. (citing ¶ 95). In particular,
Huang’s “intelligent router 92 receives a message having data” and then
“time marks the data . . . and locally caches it.” Huang ¶ 95. That is, Huang
teaches or suggests a single entity that locally caches data obtained from
elsewhere (e.g., received as part of a message). See id.
Appeal 2012-007325
Application 12/015,197

6

Appellants further contend the Examiner erred because “the Examiner
has provided zero evidence of any suggestion or motivation to combine”
Jungck and Huang. Br. 8. Appellants also argue “the Examiner has failed
to articulate a finding that there was a reasonable expectation of success.”
Id. at 9. However, we agree with the Examiner “that both Jungck and Huang
deal with inspection of content requests and servicing these requests by way
of caching if possible” (Ans. 14 (citing Jungck col. 13, ll. 25–35; Huang
Abstract)) and that an artisan of ordinary skill would have recognized
“combining Huang with Jungck would [provide] the advantages of having a
local cache directly at the point of inspection, i.e.[,] immediate servicing if
the site of inspection is close to the client” (Ans. 14). Thus, the Examiner
persuasively both: (1) articulates a reason having a rational underpinning for
the proposed combination of Jungck and Huang and (2) provides findings
showing there would have been a reasonable expectation of success for this
combination.
For these reasons, we agree with the Examiner the combination of
Jungck and Huang teaches or suggests: (1) “a cache memory for caching
data at the standalone cache unit that is requested by a remote client from the
one or more networked data storage devices” and (2) “where the high-speed
packet processor inspects network protocol traffic state parameters received
on the I/O ports to determine if a request from a remote client can be
serviced by data from the one or more networked data storage devices that is
stored in the cache memory,” as recited in claim 1. Ans. 4–5.
Accordingly, we sustain the Examiner’s 35 U.S.C. § 103(a) rejection
of claim 1, and claims 2–4, 6–12, 14, and 16–21, which Appellants do not
argue separately. Br. 4, 7.
Appeal 2012-007325
Application 12/015,197

7

Claims 5 and 15
In rejecting claim 5, the Examiner finds Sarukkai’s size-aware greedy-
dual algorithm and least recently used algorithm teach or suggest where the
predetermined cache policy comprises a least recently used with dual time
reference algorithm aided with greedy dual size frequency algorithm. Ans.
11 (citing Sarukkai col. 1, ll. 52, 65). The Examiner concludes it would
have been obvious to an artisan of ordinary skill to modify the combined
teachings of Huang and Jungck “to improve cache efficiency and to provide
a more versatile cache meeting various content access situations.” Ans. 11.
Appellants contend the Examiner erred because Sarukkai describes
greedy-dual and least recently used algorithms as “as being competing and
conflicting algorithms insofar as the ‘greedy-dual algorithm works best for
increasing document hit rate,’ while ‘a Least Recently Used (LRU)
algorithm has been shown to be reasonably effective in increasing byte-hit
rates, though it performs sub optimally in the case of document hit rates.’”
Br. 10 (citing Sarukkai col. 1, l. 51–col. 2, l. 2) (emphasis added).
In response, the Examiner concludes that, because the Specification
lacks “detail as to what aiding by an algorithm entails . . . . the broadest
reasonable interpretation of ‘aided’ would include situations where the mere
use of both algorithms or even the availability of a second algorithm would
satisfy this claim limitation.” Ans. 15 (emphasis added). We agree with the
Examiner that the Specification does not detail what it means for a least
recently used with dual time reference algorithm to be aided with a greedy
dual size frequency algorithm. The Specification merely discloses “a
predetermined cache policy, such as a least recently used with dual time
reference algorithm aided with greedy dual size frequency algorithm”
Appeal 2012-007325
Application 12/015,197

8

without any explanation as to what it means for one algorithm to aid the
other. Spec. ¶¶ 6, 35. Given this omission, we conclude: (1) the
Examiner’s broad interpretation of the disputed recitation is reasonable in
light of the Specification and (2) the disputed recitation is taught or
suggested by Sarukkai. See Ans. 11, 15.
Appellants argue the Examiner erred by relying, without supporting
evidence on a “conclusory and unsupported assertion that the combination
would be made ‘to improve cache efficiency and to provide a more versatile
cache meeting various content access situations.’” Br. 11 (citing Fin. Act.
9). However, as Appellants’ acknowledge, Sarukkai teaches or suggests the
two algorithms have different performance rates in terms of byte-hit and
document-hit rates. Ans. 10 (citing Sarukkai col. 1, l. 51–col. 2, l. 2). That
is, each algorithm can meet different content access situations (i.e., whether
high byte-hit or high document-hit rates are preferred), thus providing
support for the Examiner’s reason why it would have been obvious to
combine the teachings and suggestions of Jungck, Huang, and Sarukkai in
the manner recited in claim 15.
Appellants also argue the rejection “is entirely silent on the question
of whether there is a reasonable expectation of success.” Br. 11. However,
the Examiner’s findings persuasively show there would have been a
reasonable expectation of success for applying Sarukkai’s caching
algorithms to the standalone cache unit taught or suggested by the
combination of Jungck and Huang. See Ans. 11, 15.
For these reasons, we agree with the Examiner, the combination of
Jungck, Huang, and Sarukkai teaches or suggests “where the predetermined
cache policy comprises a least recently used with dual time reference
Appeal 2012-007325
Application 12/015,197

9

algorithm aided with greedy dual size frequency algorithm,” as recited in
claim 5. Id. at 11.
Accordingly, we sustain the Examiner’s 35 U.S.C. § 103(a) rejection
of claim 5, and claim 15, which Appellants do not argue separately. Br. 10.
Claim 13
In rejecting claim 13, the Examiner finds Jungck and Huang teach or
suggest splicing connections (Ans. 12 (citing Jungck col. 17, l. 26)), but do
“not explicitly teach a connection state, a source sequence number and a
destination sequence number in an L4 layer” (Ans. 12). Thus, the Examiner
relies on Nielsen’s “Source IP, Destination IP, Source Port, Destination Port,
Source Sequence Number, Destination Sequence Number, Protocol, and
Individual Protocol State Numbers” listing to teach or suggest a connection
state, source sequence number, and destination sequence number in an L4
layer. Id. (citing Nielsen ¶ 63).
Appellants contend the Examiner erred because claim 13 requires the
“‘splicing only a connection state, a source sequence number, and a
destination sequence number in an L4 layer.’” Br. 12. Yet, Nielsen’s listing
of connection information contains additional details (e.g., source IP,
destination IP, source port, etc.) without suggesting “that only a subset will
be used to splice connections as claimed.” Id.
We agree the Examiner erred. The Examiner does not present
sufficient findings or explanation showing Nielsen, even in combination
with Jungck and Huang, teaches or suggests splicing connections comprising
splicing only a subset of the items listed in Nielsen. Therefore, we find the
Examiner’s findings do not show the combination of Jungck, Huang, and
Nielsen teaches or suggests “where splicing connections comprises splicing
Appeal 2012-007325
Application 12/015,197

10

only a connection state, a source sequence number and a destination
sequence number in an L4 layer,” as recited in claim 13. Ans. 12.
Accordingly, we do not sustain the Examiner’s 35 U.S.C. § 103(a)
rejection of claim 13.
DECISION
We affirm the Examiner’s decision rejecting claims 1–12 and 14–21.
We reverse the Examiner’s decision rejecting claim 13.
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a). See 37 C.F.R.
§ 41.50(f).
AFFIRMED-IN-PART

Klh