Ex Parte Clark

Patent Trial and Appeal Board

Sep 29, 2016

12416415 (P.T.A.B. Sep. 29, 2016)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO.
12/416,415
36378 7590
VMWARE, INC.
DARRYL SMITH
FILING DATE
0410112009
10/03/2016
3401 Hillview Ave.
PALO ALTO, CA 94304
FIRST NAMED INVENTOR
Jonathan CLARK
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.
A304 4360
EXAMINER
BUKHARI, SIBTE H
ART UNIT PAPER NUMBER
2449
NOTIFICATION DATE DELIVERY MODE
10/03/2016 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):
ipteam@vmware.com
ipadmin@vmware.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte JONATHAN CLARK
Appeal2013-006696
Application 12/416,415
Technology Center 2400
Before JASON V. MORGAN, TRENTON A. WARD, and
NORMAN H. BEAMER, Administrative Patent Judges.
WARD, Administrative Patent Judge.
DECISION ON APPEAL
Appellant appeals under 35 U.S.C. § 134(a) from the Examiner's
Final Rejection of claims 1-16 and 20-23, which constitute all the claims
pending in this application. Appeal Br. 1. Claims 1 7-19 have been
cancelled. Id. at 26. The real party in interest identified by Appellant is
VMware, Inc. Appeal Br. 3. We have jurisdiction under 35 U.S.C. § 6(b).
We AFFIRM.
Appeal2013-006696
Application 12/416,415
STATEMENT OF THE CASE
The References and Rejections on Appeal
Claims 1, 9-11, 21, and 22 under 35 were rejected under 35 U.S.C.
§ 103 as being unpatentable over Dhawan et al. (US 2007/0079307 Al; pub.
April 5, 2007) ("Dhawan") and Bacher (US 2010/0162252 Al; pub. June 24,
2010) ("Bacher").
Claim 2 was rejected under 35 U.S.C. § 103(a) as unpatentable over
Dhawan, Bacher, and Rosu et al. (US 2006/0230407 Al; pub. Oct. 12, 2006)
("Rosu").
Claims 3-8 and 12-14 were rejected under 35 U.S.C. § 103(a) as
unpatentable over Dhawan, Bacher, and Chauffour et al. (US 2005/0005006
Al; pub. Jan. 6, 2005) ("Chauffour").
Claims 15-16, 20, and 23 were rejected under 35 U.S.C. § 103(a) as
unpatentable over Dhawan and Rueda et al. (US 2002/0112076 Al; pub.
August 15, 2 002) ("Rueda").
The Claimed Invention
The claimed invention is directed to a system in which one or more
processes are offloaded from one virtual machine running on one host
computer to another virtual machine running on another host computer so
that load balancing can be performed on a per process level and physical
servers supporting the virtual machines can be more optimally utilized.
Spec., Abstract. The method involves migrating one or more processes from
a virtual machine running on a first host computer to a second host
computer, including replicating the virtual machine on the second host
computer, terminating the migrated processes from the virtual machine
running on the first host computer, and terminating all user-level processes
2
Appeal2013-006696
Application 12/416,415
running on the replicated virtual machine other than the migrated processes.
Id.
Claimed Subject Matter
Independent claim 1 under appeal is illustrative and is reproduced
below (emphasis added):
1. In a computer system having virtual machines
instantiated on one or more host computers, a method for
offloading one or more user-level processes running on a guest
operating system that is running in a first virtual machine, said
method comprising:
identifying some but not all of the user-level processes
running on the guest operating system of the first virtual
machine to be offloaded;
replicating the first virtual machine to create a second
virtual machine, the second virtual machine having a guest
operating system running therein that is replicated from the
guest operating system of the first virtual machine and user-
level processes running on the guest operating system of the
second virtual machine that are replicated from all of the user-
level processes running on the guest operating system of the
first virtual machine, including some that are to be offloaded
and remaining user-level processes that are not to be offloaded;
terminating the user-level processes running on the guest
operating system of the first virtual machine that were
identified for offloading while continuing to execute all of the
remaining user-level processes running on the guest operating
system of the first virtual machine; and
terminating all of the user-level processes running on the
guest operating system of the second virtual machine that were
replicated from the remaining user-level processes running on
the guest operating system of the first virtual machine while
continuing to execute the user-level processes running on the
guest operating system of the second virtual machine that were
replicated from the user-level processes running on the guest
3
Appeal2013-006696
Application 12/416,415
operating system of the first virtual machine that were identijzed
for offloading.
Appellant's Contentions
1. Appellant contends the Examiner erred in rejecting claims 1, 9-
11, 21, and 22 because Dhawan in view of Bacher fails to teach
or suggest terminating, on a first virtual machine, processes
identified for offloading while continuing to execute remaining
processes, and vice versa for processes replicated on a second
virtual machine. Appeal Br. 11.
2. Appellant also contends that the Examiner erred in rejecting
claims 1, 9-11, 21, and 22 because Dhawan in view of Bacher
fails to teach or suggest replicating both "processes to be
offloaded" and "remaining processes." Appeal Br. 13.
3. Appellant further contends that the Examiner erred in rejecting
claims 15-16; 20; and 23 because Dhawan in view of Rueda
fails to teach or suggest the claim limitation requiring a first
virtual machine running on first host computer and a second
virtual machine running on a second host computer to have
different MAC addresses. Appeal Br. 1 7.
ANALYSIS
We have reviewed the Examiner's rejections in light of Appellant's
contentions that the Examiner has erred.
We disagree with Appellant's arguments. We adopt as our own (1)
the findings and reasons set forth by the Examiner in the action from which
this appeal is taken and (2) the reasons set forth by the Examiner in the
4
Appeal2013-006696
Application 12/416,415
Examiner's Answer in response to Appellant's Appeal Brief As described
below, we concur with the conclusions reached by the Examiner.
A. Obviousness Rejections Based on (1) Dhawan and Bacher,
(2) Dhawan, Bacher, and Rosu, and (3) Dhawan, Bacher, and
Chaujfour
1. Whether Dhawan in view of Bacher fails to teach or suggest the
claimed terminating steps
Appellant contends the Examiner erred in rejecting claims 1, 9-11,
21, and 22 because Dhawan in view of Bacher fails to teach or suggest:
(1) on a first virtual machine, terminating processes identified for offloading
while continuing to execute remaining processes; and (2) on a second virtual
machine, terminating processes replicated from the remaining processes
while continuing to run replicated processes that were identified for
offloading. Appeal Br. 12. Specifically, Appellant argues that a person of
ordinary skill in the art reviewing the cited portions of Bacher would
"understand that the process (present as a 'process stub') in the first platform
(i.e., first kernel of the first system) and its duplicate in the second platform
(i.e., second kernel of the second system) are simultaneously terminated in
response to the 'event trigger to terminate the process."' Id. (emphasis
added).
The Examiner disagrees and responds that Appellant's interpretation
of a process stub running on the first platform as a process itself running on
the first platform is not accurate. Ans. 4. We agree. Bacher discloses a
hybrid system 100 for shifting workloads across platforms, such as by
duplicating the processes on system K to system U. Bacher i-fi-157, 64. As
the Examiner notes in the Answer (Ans. 4), Bacher discloses that
5
Appeal2013-006696
Application 12/416,415
A process stub of the process resides in the system K to
represent, within the kernel K, the duplicate process executing
in the system U but the process stub does not waste clock
cycles of the system K as being excluded from scheduling.
Bacher i-f 66 (emphasis added). Therefore, as described by the Examiner,
through the "process stub" the kernel K manages the duplicate process
running in the system U but does not waste clock cycles of the system K
because it is excluded from scheduling. Ans. 4 (citing Bacher i-f 66).
Therefore, we are not persuaded by Appellant's argument that the process
stub is the process itself running on the first platform.
Appellant further argues that Bacher fails to teach continuing to run
the offloaded process on the second system because Appellant argues that
Bacher teaches simultaneously terminating the process on the first platform
and the second platform in response to an event trigger to terminate the
process. Appeal Br. 12. Contrary to Appellant's argument, Bacher discloses
"stopping scheduling the process in the first platform responsive to receiving
a signal from the second kernel indicating that the second platform schedules
and executes the duplicate process in the second platform." Bacher i-f 8
(emphasis added). Thus, in Bacher, the first process on the first platform is
stopped based on a signal that the second platform has begun execution of
the duplicate process. See id. Therefore, we are not persuaded of error in
the Examiner's finding that Dhawan in view of Bacher teaches the claimed
"terminating the user-level processes running on the guest operating system
of the first virtual machine that were identified for offloading."
Furthermore, we are not persuaded by Appellant's argument that
Bacher teaches simultaneously terminating the process on the first platform
and the duplicate process on the second platform. As discussed, above
6
Appeal2013-006696
Application 12/416,415
Bacher describes a "process stub" on the first platform and a duplicate
process on the second platform. See Ans. 4 (citing Bacher i-f 66). Bacher
discloses that after execution of duplicate process an event trigger may
provide a "notification by the kernel U [the second platform] that the
duplicate process had been halted." Bacher i-f 68. In response to this event
trigger, Bacher discloses that "the kernel K [on the first platform] terminates
the process stub for the system U binary code in the system K and
subsequently signals the kernel U to terminate the duplicate process on the
system U." Bacher i-f 69 (emphasis added). Thus, Bacher teaches
terminating the process stub on the first platform (system U), as the
execution of the original process on the first platform (system U) has already
been terminated. See Bacher i-fi-1 66---69. As the Examiner identifies,
Appellant's proposed simultaneous termination is not practical because if the
original process is to be terminated, it would be impractical for Bacher to
teach duplicating the original process first and then terminating both the
original process and its duplicate. Ans. 4. This is also contrary to Bacher's
disclosure regarding the duplicate process and its process stub. See Bacher
i-fi-1 68---69. Accordingly, we are not persuaded of error in the Examiner's
finding that Dhawan in view of Bacher teaches these limitations of claim 1.
2. Whether Dhawan in view of Bacher fails to teach or suggest
replicating both ''processes to be offloaded" and "remaining
" processes
Appellant contends that the Examiner erred in rejecting claims 1, 9-
11, 21, and 22 because Dhawan in view of Bacher fails to teach or suggest
replicating both "processes to be offloaded" and "remaining processes."
Appeal Br. 13. Specifically, Appellant argues that the disclosure in Dhawan
regarding hot cloning relied upon by the Examiner "does not teach what
7
Appeal2013-006696
Application 12/416,415
processes to transfer" and Dhawan is silent about what processes might be
loaded into the carrier VM in the first place Id. at 14 (emphasis omitted).
The Examiner responds that Appellant's arguments are irrelevant because
the claimed step of replicating is reasonably interpreted as replicating all the
running processes. Ans. 6.
We agree. Claim 1 requires replicating "all of the user-level
processes running on the guest operating system of the first virtual machine"
(emphasis added). Dhawan's hot cloning teaches replicating all of the
processes from a first virtual machine to a second virtual machine. See
Dhawan i-f 20 (describing that in hot cloning, "a clone of the VM is created
in the next participating physical host in the migration path."). As claim 1
requires "all of the user-level processes" to be replicated, we are not
persuaded of error in the Examiner's reliance upon Dhawan's hot cloning to
teach this replication. See Ans. 5-6.
Based on the foregoing, we sustain the Examiner's rejection of claims
1, 9-11, 21, and 22. Appellant relies upon the same arguments as those
made against claims 1, 9-11, 21, and 22 in disputing the rejections of claims
2-8 and 12-14. Br. 17. As with claims 1, 9-11, 21, and 22 we are
unpersuaded equally of error and sustain the rejection of claims 2-8 and 12-
14.
B. Obviousness Rejection Based Dhawan and Rueda
The Examiner rejects claims 15-16, 20, and 23 as unpatentable over
Dhawan in view of Rueda. Final Action 17-22. Appellant argues that the
Examiner erred in rejecting claim 15 because (1) Rueda makes no mention
of virtual machines and (2) contrary to Rueda, Dhawan requires maintaining
8
Appeal2013-006696
Application 12/416,415
the same MAC and IP address in both a virtual machine and a migrated
virtual machine. Appeal Br. 18-19.
First, as to Appellant's argument that Rueda fails to teach a virtual
machine but instead discloses only physical computers, the Examiner
responds that the rejection does not rely upon Rueda alone. Ans. 7. The
Examiner states that claim 15 is rejected over Dhawan in view of Rueda, and
the rejection relies upon Dhawan for its teachings regarding virtual
machines, not Rueda. Id. We are not persuaded by Appellant's argument
that the rejection is improper simply because Rueda fails to disclose virtual
machines.
Second, Appellant argues that Dhawan cannot be properly combined
with Rueda to teach the claim 15 recitation that "the first virtual machine
and the second virtual machine share the same IP address and have different
MAC addresses." Appeal Br. 18. Appellant argues that Rueda requires that
each of the different IP addresses of the client machine be associated with a
unique MAC address. Id. at 19 (citing Rueda i-fi-1283-84). Appellant further
argues that, in contrast to Rueda, Dhawan requires maintaining the same
MAC and IP address in both the virtual machine and the migrated virtual
machine. Id. (citing Dhawan i-f 14). The Examiner responds that Appellant
has not analyzed the particular combination of Dhawan and Rueda set forth
in the Final Office Action. Ans. 9. In the Final Office Action, the Examiner
stated that "given the suggestion by Dhawan for transferring payload using
the same IP and MAC addresses and the teachings by Rueda for mapping IP
sockets to different MAC addresses" it would have been obvious to combine
these teachings by "modifying the load transfer to allow load transfer to a
different MAC address using mapping tables." Final Action 19.
9
Appeal2013-006696
Application 12/416,415
We note that the cited disclosure in Rueda provides that a table is
constructed and the fields in the table are chosen to provide a specific
mapping of an IP socket (IP address and port number) to a MAC address.
Rueda i-f 282. In accordance with the Examiner's rejection, this mapping
table in Rueda would be implemented into the Dhawan system. See Ans. 9.
Therefore, although Dhawan teaches maintaining the same MAC and IP
address in both the virtual machine and the migrated virtual machine, in
accordance with the Examiner's rejection, it would have been obvious to one
of skill in the art to modify Dhawan with Rueda's mapping table for
mapping IP sockets to different MAC addresses. See id. In view of the cited
disclosures, we are not persuaded of error in the Examiner's combination of
Dhawan virtual machine migration with Rueda' s mapping table to teach the
claim 15 recitation that "the first virtual machine and the second virtual
machine share the same IP address and have different MAC addresses."
Based on the foregoing, we sustain the Examiner's rejection of claims 15-
16, 20, and 23 as unpatentable over Dhawan in view of Rueda.
DECISION
The Examiner's decision rejecting claims 1-16 and 20-23 is affirmed.
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).
AFFIRMED
10