Ex Parte LI et al

Patent Trial and Appeal Board

Jun 29, 2017

13309300 (P.T.A.B. Jun. 29, 2017)

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.
13/309,300 12/01/2011 Fulu LI ORA120096-US -NP 1474
55498 7590 07/03/2017
Vista IP Law Group, LLP (Oracle)
2160 Lundy Avenue
Suite 230
San Jose, CA 95131
EXAMINER
BLACK, LINH
ART UNIT PAPER NUMBER
2163
NOTIFICATION DATE DELIVERY MODE
07/03/2017 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):
docketing@viplawgroup.com
ev@viplawgroup.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte FULU LI, ATSUSHI MORIMURA,
MICHAEL ZOLL, VINEET MARWAH, and AMIT GANESH
Appeal 2017-004517
Application 13/309,3001
Technology Center 2100
Before JOHN A. JEFFERY, BRUCE R. WINSOR, and
JUSTIN BUSCH, Administrative Patent Judges.
BUSCH, Administrative Patent Judge.
DECISION ON APPEAL
Appellants seek our review under 35 U.S.C. § 134 of the Examiner’s
final decision rejecting claims 1—5, 7—12, 14—18, and 20. Claims 6, 13, and
19 were canceled previously. We have jurisdiction under 35 U.S.C. § 6(b).
We reverse.
1 According to Appellants, the real party in interest is Oracle International
Corporation. App. Br. 3.
Appeal 2017-004517
Application 13/309,300
CLAIMED SUBJECT MATTER
Claims 1, 8, and 15 are independent claims. The claims generally
relate to “implementing failover and resume when using ordered sequences
in a multi-instance database environment.” Spec. Abstract.
Claim 1 is representative and reproduced below:
1. A computer implemented method for implementing
failover when using ordered sequences in a multi-instance
database environment, the method comprising:
instantiating a plurality of database instances of a
database, wherein the plurality of database instances comprises
at least a first database instance initially designated as an active
instance, and one or more other database instances initially
designated as passive instances;
identifying a master sequence associated with the
database, wherein access to the master sequence is controlled
through a global lock;
allocating one or more portions of the master sequence to
form a sequence cache, wherein the sequence cache comprises a
plurality of symbols to be accessed in an order for satisfying
sequence value requests from one or more sessions associated
with the first database instance and the one or more other
database instances;
designating the first database instance as master of the
sequence cache, such that access to the sequence cache is
controlled through a first instance lock on the first database
instance;
satisfying a first sequence value request by accessing the
sequence cache through the first instance lock on the first
database instance;
using a second database instance to recognize that the first
database instance has failed, where the second database instance
that recognized that the first database instance has failed is
designated as master of the sequence cache, such that access to
the sequence cache is controlled through a second instance lock
on the second database instance; and
2
Appeal 2017-004517
Application 13/309,300
satisfying a second sequence value request by accessing
the sequence cache through the second instance lock on the
second database instance.
REJECTION
Claims 1—5, 7—12, 14—18, and 202 3stand rejected under 35 U.S.C.
§ 103(a) as being unpatentable over Colrain et al. (US 2006/0190453 Al;
Aug. 24, 2006), Chen et al. (US 7,089,356 Bl; Aug. 8, 2006), and Cruz et al.
(US 2005/0172161 Al; Aug. 4, 2005). Final Act. 3-9.
OPINION
The Examiner finds the combination of Colrain, Chen, and Cruz
teaches or suggests each of the limitations of independent claims 1, 8, and
15. Final Act. 3 8. In particular, the Examiner finds Chen discloses,
inter alia, allocating one or more portions of a master sequence to form a
sequence cache. Id. at 6-1 (citing Chen 4:1 50, 5:5 67; Figures 1, 2). The
Examiner finds Colrain discloses, inter alia, a sequence cache comprising a
plurality of symbols to be accessed in an order for satisfying sequence value
requests from one or more sessions associated with database instances. Id.
at 5 (citing Colrain || 12, 13, 36, 48, 53 56; Figures 8 and 9). The Examiner
concludes “it would have been obvious to one of ordinary skill in the art at
the time of the invention to combine the teachings of Colrain and [Chen] in
order to effectively and continuously [monitor and perform] sequences of
database events without interruption of a database failure.” Id. at 7.
2 Although the Examiner includes cancelled claim 19 in the statement of the
rejection and its corresponding discussion (Final Act. 3, 9), we nonetheless
omit that claim and present the correct claim listing here for clarity.
3
Appeal 2017-004517
Application 13/309,300
Appellants contend “[t]he proposed combination of references . . .
would render Colrain unsatisfactory for its intended purpose.” App. Br. 17.
Specifically, Appellants assert Colrain discloses that a monitored database
instance, during normal operation, acquires and holds a semaphore on a
pseudo object using a holding process. Id. at 17 18 (citing
Colrain Figures 5, 6). Appellants argue a second process acquiring the
pseudo object controlled by the semaphore would eviscerate Colrain’s
intended purpose because “Colrain is specifically premised upon an
approach where [the] second process is prevented from acquiring [the
pseudo object] controlled by [the] semaphore during normal operation.”
Id. at 17.
It is well established that, where the proposed modification would
render the prior art invention being modified unsatisfactory for its intended
purpose, the proposed modification would not have been obvious. See Tec
Air Inc. v. Denso Mfg. Michigan Inc., 192 F.3d 1353, 1360 (Fed. Cir. 1999);
In re Gordon, 733 F.2d 900, 902 (Fed. Cir. 1984).
We agree with Appellants in concluding the proposed combination is
problematic. First, Colrain is directed to detecting the termination of a
monitored database instance using a lock- or event-based system. Colrain
1112—13. Colrain’s lock-based system begins by forking a lightweight
parent process into a holding routine/process and a waiting routine/process
in serial order. Id. at H 37, 42. Upon the holding routine obtaining an
exclusive lock on a pseudo-object in the monitored database instance, the
holding routine blocks on the pseudo-object and sleeps without releasing the
exclusive lock. Id. at H 45, 46. The waiting routine then connects to the
4
Appeal 2017-004517
Application 13/309,300
monitored database instance and indefinitely attempts to acquire the
exclusive lock on the same pseudo-object. Id. at 147. If the waiting routine
successfully acquires the exclusive lock, a standard application error is
returned. Id. Accordingly, we find one intended purpose of Colrain’s lock-
based system is to not release the exclusive lock on the pseudo-object in the
monitored database instance while the monitored database instance is
operating.
Second, Figure 2 of Chen, relied upon by the Examiner, is illustrative
and reproduced below:
200
5
Appeal 2017-004517
Application 13/309,300
Figure 2 of Chen “depicts the relationships between a sequence
object, multiple database instances and multiple parallel processes.”
Chen 5:11—13. As shown, system (200) includes a sequence object (202)
with a corresponding R lock (203). Id. at 5:16—18. System (200) further
includes a plurality of database instances (204A, 204B, 204N) that call the
sequence object (202) to gain control of the corresponding R lock (203). Id.
at 5:45—55. Chen further teaches processes (e.g., 208A, 208B) in each
database instance similarly gain control of the corresponding R lock (203).
Id. at 6:4—12. Chen teaches the purpose of the R lock (203) is to “[prohibit]
more than one database instance 204A, 204B, 204N, or associated processes,
from concurrently accessing and obtaining sequence values from sequence
object 202.” Id. at 5:38-41. Thus, we find Chen’s sequence object (202),
allowing the database instances (204A, 204B, 204N) and their respective
processes (e.g., 208A, 208B) to gain control of the corresponding R lock
(203) and obtain and cache (206A, 206B) sequence values from sequence
object (202) in a serial order, allowing database instances to maintain
separate respective caches concurrently, runs counter to Colrain’s intended
purpose discussed above.
In response to Appellants’ contention, other than repeating the
findings from the Final Action, the Examiner finds Colrain discloses “a
failsafe environment for database instances” as claimed. Ans. 8. The
Examiner further finds Figure 2 of Chen teaches allocating one or more
portions of a master sequence (202) to form a sequence cache (206A through
206N). Id. at 9. The Examiner explains Chen’s allocation “does not prevent
the active database instance to be monitored in case of failure.” Id.
6
Appeal 2017-004517
Application 13/309,300
The Examiner has not cogently explained how applying Chen’s
architecture to Colrain would not defeat Colrain’s intended purpose of not
releasing an exclusive lock on a pseudo-object in a monitored database
instance while the monitored database instance is operating. Because the
Examiner has supplied insufficient evidence or reasoning to explain why the
skilled artisan would have made the proposed modification, we agree with
Appellants. Accordingly, on this record, we must reverse the 35 U.S.C.
§ 103(a) rejections of claims 1—5, 7—12, 14—18, and 20.
DECISION
For the above reasons, the Examiner’s decision to reject claims 1—5,
7—12, 14—18, and 20 is reversed.
REVERSED
7