DETAILED ACTION
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Status of Claims
The following claim(s) is/are pending in this office action: 1-20
The following claim(s) is/are amended: 3, 5, 8-10, 12, 14-17, 19
The following claim(s) is/are cancelled: -
The following claim(s) is/are new: -
Claim(s) 1-20 is/are rejected. This rejection is FINAL.
Previous Rejections Withdrawn
The 35 USC 112(b) rejection to claim(s) 5, 8-20 is/are withdrawn based on the amendment.
Response to Arguments
Applicant’s arguments filed in the amendment filed 5/4/2026, have been fully considered but are moot in view of new grounds of rejection. The reasons set forth below.
Applicant’s Invention as Claimed
Claim Rejections - 35 USC § 103
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102 of this title, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-20 are rejected under 35 U.S.C. 103(a) as being unpatentable over Goyal (US Pub. 2020/0159568) in view of Hamilton (US Pub. 2003/0182464), and further in view of Lavian (US Pub. 2014/0105025).
With respect to Claim 1, Goyal teaches a network message processing method, comprising: reading a first network message from a first queue maintained in a first server; (A first queue will be taught later. Fig. 1a, paras. 31-35; Data center includes data processing units (access nodes) connected to a plurality of servers. Paras. 34-37; The access nodes may process streams of data written to or read from the servers. Each access node may act as network or storage I/O or a gateway for multiple servers.)
querying a queue mapping table maintained in a data processing unit to determine a destination second queue to which the first network message is to be transmitted, and storing the first network message in the destination second queue, (para. 37-43; DPU includes multiple cores for performing processing on packet streams. Paras. 45-46; DPU receives stream and converts it to work units. Paras. 59-61; upon receiving a flow, DPU performs a lookup in a flow to map it to a particular core by using a queue manager to enqueue the work unit in the queue associated with that particular core. See also Fig. 6, para. 69; WU queue for each core.)
and calling a message distribution module in the data processing unit, and reading, through the message distribution module, the first network message from the destination second queue and transmitting the first network message to a destination second server through a network. (para. 80; DPU transmits data packets to one or more external devices. Fig. 1a, paras. 31-35; Data center includes data processing units (access nodes) connected to a plurality of servers. Paras. 34-37; The access nodes may process streams of data written to or read from the servers. Para. 187; allow packet to pass through device.)
But Goyal does not explicitly teach a quantity of the first queues is greater than a quantity of the second queues.
Hamilton, however, does teach a first queue maintained in a first server (Fig. 1, paras. 51-54; network of server computers with queueing. Fig. 5, para. 67-69; Queues can be accessed locally or remotely. Member queues of a macro queue can be located outside of a local computer. Fig. 4, paras. 62-66; each process has a queue that feeds into a macro queue. Therefore, Hamilton teaches a technique where queues on multiple devices such as servers can be accessed by a remote device.)
and a quantity of the first queues is greater than a quantity of the second queues; (Fig. 4, para. 66; four producer queues feed into a macro queue. Therefore, the art can map multiple inputs to a single queue.)
It would have been obvious to one of ordinary skill prior to the effective filing date to combine the method of Goyal with the quantity of the first queues is greater than a quantity of the second queues in order to allow a device to retrieve messages from multiple sources as if they were a single source. (Hamilton, para. 5)
But modified Goyal does not explicitly teach a mapping relationship between first queues maintained in the first server and second queues maintained in the data processing unit.
Lavian, however, does teach wherein the queue mapping table is used to record a mapping relationship between first queues maintained in the first server and second queues maintained in the data processing unit, (Goyal may teach on its own. First see Goyal, paras. 59-60; system includes a flow table that maps a flow to a core. When a flow is seen for the first time the system writes the flow into the flow table. Subsequent times the flow can be looked up to find which core (and therefore which queue) is responsible for the flow. That is a mapping between the incoming data (from the first queue) and the second queues. However, Goyal does not explicitly teach mapping a server to a queue, i.e., assigning a particular server to a particular queue, and it is unclear if the claim language calls for that. Consequently, Examiner will cite Lavian, paras. 27-28, 36; AR table includes source address, and packets may be specifically enqueued in particular queues based upon source address.)
It would have been obvious to one of ordinary skill prior to the effective filing date to combine the method of modified Goyal with the mapping relationship between first queues maintained in the first server and second queues maintained in the data processing unit in order to allow a device to prioritize data from particular devices. (Lavian, para. 36)
With respect to Claim 2, modified Goyal teaches the network message processing method according to claim 1, and Goyal also teaches wherein the calling the message distribution module in the data processing unit, and reading, through the message distribution module, the first network message from the destination second queue and transmitting the first network message to the destination second server through the network comprises: calling the message distribution module in the data processing unit, and controlling the message distribution module to use a multi-thread mode, wherein one second queue in the data processing unit is allocated for each thread, and different threads correspond to different second queues; (para. 4, 108-109; DPU accelerators are multithreaded for parallel processing. para. 37-43; DPU includes multiple cores for performing processing on packet streams. Fig. 6, para. 69; WU queue for each core. Para. 95; DPU performs thread scheduling)
And Hamilton also teaches and calling a first target thread corresponding to the destination second queue, and reading, through the first target thread, the first network message and transmitting the first network message to the destination second server through the network. (Fig. 5, para. 67-69; Queues can be accessed locally or remotely. Member queues of a macro queue can be located outside of a local computer.)
The same motivation to combine as the independent claim applies here.
With respect to Claim 3, modified Goyal teaches the network message processing method according to claim 2, and Lavian also teaches wherein the reading, through the first target thread, the first network message and transmitting the first network message to the destination second server through the network comprises: determining the first queue from which the first network message comes, and determining a second target thread having a mapping relationship with the first queue, (paras. 27-28, 36; AR table includes source address, and packets may be specifically enqueued in particular queues based upon source address. See also Goyal, paras. 59-60; system includes a flow table that maps a flow to a core. When a flow is seen for the first time the system writes the flow into the flow table. Subsequent times the flow can be looked up to find which core (and therefore which queue) is responsible for the flow.)
each first queue of the first server is provided with a corresponding thread; (paras. 27-28, 36; AR table includes source address, and packets may be specifically enqueued in particular queues based upon source address. See also Goyal, paras. 45-46; DPU receives stream and converts it to work units. Paras. 59-61; upon receiving a flow, DPU performs a lookup in a flow to map it to a particular core by using a queue manager to enqueue the work unit in the queue associated with that particular core. See also Fig. 6, para. 69; WU queue for each core.)
The same motivation to combine as the independent claim applies here.
And Goyal also teaches wherein the second target thread is a thread on the data processing unit and (para. 4, 108-109; DPU accelerators are multithreaded for parallel processing. para. 37-43; DPU includes multiple cores for performing processing on packet streams. Fig. 6, para. 69; WU queue for each core.)
if the second target thread is not a same thread as the first target thread, reading, through the first target thread, the first network message and forwarding the first network message to the second target thread; (Para. 151; dispatcher invokes functions in a series. paras. 96, 107-108, 151; DPU may employ service chaining, where a work unit has multiple functions performed on it. A processor may relinquish control of work units to be passed to next processors in line. A dispatcher may invoke next functions.)
and transmitting, through the second target thread, the first network message to the destination second server through the network. (para. 80; DPU transmits data packets to one or more external devices. Fig. 1a, paras. 31-35; Data center includes data processing units (access nodes) connected to a plurality of servers. Paras. 34-37; The access nodes may process streams of data written to or read from the servers. Para. 187; allow packet to pass through device.)
With respect to Claim 4, modified Goyal teaches the network message processing method according to claim 3, and Goyal also teaches further comprising: determining a target second queue corresponding to the second target thread; (Fig. 6, para. 69; WU queue for each core. Para. 151; dispatcher invokes functions in a series.)
and updating, in the queue mapping table, a mapping relationship between the first queue and the destination second queue to a mapping relationship between the first queue and the target second queue. (paras. 96, 107-108, 151; DPU may employ service chaining, where a work unit has multiple functions performed on it. A processor may relinquish control of work units to be passed to next processors in line. A dispatcher may invoke next functions. Paras. 59-61; upon receiving a flow, DPU performs a lookup in a flow to map it to a particular core by using a queue manager to enqueue the work unit in the queue associated with that particular core. See also Lavian, para. 45; updating of queue assignment.)
With respect to Claim 5, modified Goyal teaches the network message processing method according to claim 1, and Goyal also teaches further comprising: calling the message distribution module in the data processing unit, and obtaining a second network message, (Paras. 45-46; DPU receives stream, which suggests a plurality of messages.)
and reading the second network message from the second queue, identifying the target first queue to which the second network message is to be transmitted, and transmitting the second network message to the target first queue. (para. 80; DPU transmits data packets to one or more external devices. Fig. 1a, paras. 31-35; Data center includes data processing units (access nodes) connected to a plurality of servers. Paras. 34-37; The access nodes may process streams of data written to or read from the servers. Para. 187; allow packet to pass through device.)
and Lavian also teaches wherein the second network message carries a target first queue to which the second network message is to be transmitted; (para. 16-20; destination address for a message.)
The same motivation to combine as the independent claim applies here.
And Hamilton also teaches storing the second network message in a second queue in the data processing unit; (para. 64; enqueueing in a round-robin fashion to a plurality of queues. Examiner also notes that enqueueing based on assignment or enqueueing based on source address would result in storage of “in a.” See Goyal, paras. 59-61; upon receiving a flow, DPU performs a lookup in a flow to map it to a particular core by using a queue manager to enqueue the work unit in the queue associated with that particular core. See Lavian, Lavian, paras. 27-28, 36; AR table includes source address, and packets may be specifically enqueued in particular queues based upon source address.)
The same motivation to combine as the independent claim applies here.
With respect to Claim 6, modified Goyal teaches the network message processing method according to claim 1, and Goyal also teaches wherein the queue mapping table maintains a mapping relationship between a plurality of first queues (Fig. 1a, paras. 31-35; Data center includes data processing units (access nodes) connected to a plurality of servers. Paras. 34-37; The access nodes may process streams of data written to or read from the servers. Each access node may act as network or storage I/O or a gateway for multiple servers.)
and a plurality of second queues, (para. 37-43; DPU includes multiple cores for performing processing on packet streams. Paras. 45-46; DPU receives stream and converts it to work units. Paras. 59-61; upon receiving a flow, DPU performs a lookup in a flow to map it to a particular core by using a queue manager to enqueue the work unit in the queue associated with that particular core. See also Fig. 6, para. 69; WU queue for each core.)
and Lavian also teaches wherein one first queue is mapped to only one second queue, (paras. 27-28, 36; AR table includes source address, and packets may be specifically enqueued in particular queues based upon source address.)
The same motivation to combine as the independent claim applies here.
And Hamilton also teaches and one second queue supports mapping of a plurality of first queues. (Fig. 4, para. 66; four producer queues feed into a macro queue.)
The same motivation to combine as the independent claim applies here.
With respect to Claim 7, modified Goyal teaches the network message processing method according to claim 1, and Goyal also teaches wherein the reading, through the message distribution module, the first network message from the destination second queue and transmitting the first network message to the destination second server through the network comprises: reading, through the message distribution module, the first network message from the destination second queue, identifying the destination second server to which the first network message is to be transmitted, and transmitting the first network message to the destination second server through the network. (para. 80; DPU transmits data packets to one or more external devices. Fig. 1a, paras. 31-35; Data center includes data processing units (access nodes) connected to a plurality of servers. Paras. 34-37; The access nodes may process streams of data written to or read from the servers. Para. 187; allow packet to pass through device.)
With respect to Claim 8, it is substantially similar to Claim 1 and is rejected in the same manner, the same art and reasoning applying. Further, Goyal also teaches an electronic device, comprising: at least one processor and a memory; wherein the memory stores computer-execution instructions; and when the at least one processor executes the computer-execution instructions stored in the memory, the at least one processor is enabled to: (para. 31, 37; processor. para. 70; memory.)
With respect to Claims 9-14, they are substantially similar to Claims 2-7, respectively, and are rejected in the same manner, the same art and reasoning applying.
With respect to Claim 15, it is substantially similar to Claim 1 and is rejected in the same manner, the same art and reasoning applying. Further, Goyal also teaches a non-transitory computer-readable storage medium, wherein the computer-readable storage medium stores computer-execution instructions, and when a processor executes the computer-execution instructions, the following operations are implemented: (para. 207; non-transitory media read by processors.)
With respect to Claims 16-20, they are substantially similar to Claims 2-6, respectively, and are rejected in the same manner, the same art and reasoning applying.
Remarks
Applicant amends Claim 3 and similar claims. They are taught above. Applicant argues at Remarks, pg. 10 that the amendments to other claims fixes the previous 112b issues with them. Examiner agrees and withdraws the rejection.
With respect to the independent claims, Applicant stands on argument. Applicant argues at Remarks, pgs. 11-13, that while Goyal may teach a DPU and one queue, it does not teach two queues and therefore fails to teach entirety of the second limitation of Claim 1. Applicant argues at Remarks, pgs. 13-14, that while Hamilton teaches two queues, it fails to teach a DPU and therefore fails to teach the entirety of the second limitation of Claim 1. Applicant argues at Remarks, pgs. 15-16, that while Lavian teaches a mapping, Lavian’s AR table “is in fact related to the priority of a packet” and it fails to teach a DPU and therefor fails to teach the entirety of the second limitation of Claim 1.
The argument is unpersuasive because Applicant improperly piecemeal attacks the rejection, as evidenced by the manner in which Applicant takes each reference individually in a serial fashion and complains that a feature that is obviously present in the primary reference (a DPU) is missing from the other references. When the teachings are considered in combination the claims are obvious over the sum of the teachings.
Specifically, Goyal teaches assigning a message to one of a plurality of destination second queues, but differs in that the message does not come from a first queue, where the number of first queues is greater than the number of second queues. Hamilton teaches that there can be a plurality of first queues and gives an example where the number of first queues is larger than the number of second queues. Therefore, rather than just assigning streams of data to destination (second) queues, in combination a person of ordinary skill would have recognized that one could have had a plurality of source (first) queues, and then dequeue from those queues to fill the destination queues.
But which sources go to which destinations? Examiner initially asserts that Goyal renders the mapping between first/second queues obvious. Goyal has a queue for each core, and a flow table that maps flows to cores. Examiner asserts that is a “queue mapping table [that] is used to record a mapping relationship between first queues [] and second queues [].” However, to make the teaching more clear, Examiner cites Lavian, which specifically enqueues based on a source address. Therefore, in combination, a person of ordinary skill would have recognized that one could map to second queues based on a source, which is a mapping table that records a mapping relationship between first queues and second queues.
Consequently, the argument that Goyal lacks two queue types and that Hamilton and Lavian lack a DPU is irrelevant, because no reference is required to anticipate the entirety of the second limitation. The only question is whether the limitation is obvious over the sum knowledge of the art. It was obvious prior to the instant effective filing to map sources to destinations and to employ queues at both ends of the transmission. That should be an unsurprising conclusion, as Applicant admits that conventionally DPUs map from source to destination queues, see Spec, Background, paras. 3-4. In other words, in addition to being an improper serialized individual analysis of the references the instant arguments are belied by the Specification, which takes as a given that DPUs have mappings between two sets of queues. Rather, Applicant’s Specification asserts that the problem with the prior art is a one-to-one mapping that overly consumes resources (Spec, paras. 3-5). But a many-to-one (or a plurality-to-smaller plurality, as that is just a duplication of a many-to-one system) enqueuing system was old in the art. Hamilton (a 2003 reference) explicitly draws a many-to-one queuing system in Fig. 4. And although Applicant argues that the motivation in Lavian for enqueuing in different queues is “priority,” Applicant does not dispute that Lavian teaches enqueuing in a different queue (a first priority queue or a second best-effort queue) based on a source address, which is a source to destination mapping that enables a many-to-one enqueuing. These teachings are on top of the fact that Goyal teaches a flow table that maps to a particular core (and therefore a particular second queue) based on the flow and that there is no logical limitation on the number of flows, i.e., the natural reading of Goyal is that a many to one mapping could exist because the flow mapping table suggests that any number of flows could be guided to a particular queue.
All claims remain rejected.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any extension fee pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to NICHOLAS P CELANI whose telephone number is (571)272-1205. The examiner can normally be reached on M-F 9-5.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Vivek Srivastava can be reached on 571-272-7304. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/NICHOLAS P CELANI/Examiner, Art Unit 2449
Prosecution Insights