Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Examiner’s Note
The Examiner encourages Applicant to schedule an interview to discuss issues related to, for example, the rejections noted below under 35 U.S.C § 101 and § 103, for moving forward allowance.
Providing supporting paragraph(s) for each limitation of amended/new claim(s) in Remarks is strongly requested for clear and definite claim interpretations by Examiner.
Priority
Acknowledgment is made of applicant's claim for the present application filed on 12/14/2021.
Response to Arguments
Applicant's arguments filed on 11/26/2025 have been fully considered but they are not persuasive.
In Remarks, pp. 8-13, Applicant contends:
While Applicant does not concede to this interpretation of claim 1, Applicant submits that even if claim 1 were directed to such a concept, the alleged abstract idea does not fall within any of the subject matter groupings of abstract ideas enumerated in the 2019 Guidance (i.e. "Mathematical concepts," "Certain methods of organizing human activity," or "Mental Processes"). Therefore, for at least this reason, claim 1 does not recite an abstract idea, failing Prong One of the USPTO's required analysis.
…
Applicant respectfully submits that claim l's features are integrated into the practical application of a addressing future bottlenecks of the computing system. For example, claim 1 recites "feeding, by the one or more processors, a first part of the contention-related data to a first machine-learning system," and "feeding, by the one or more processors, a second part of the contention-related data scaled with the first impact values to a second machine-learning system." By integrating these features to the environment of a system having different workload groups, claim 1 integrates the features - and any associated abstract ideas - into a practical application (e.g. addressing bottlenecks), thereby rendering claim 1 subject matter eligible.
…
As shown, the present specification discloses several improvements in how performance of a computer is improved by identifying a cause of a performance anomaly. For example, the specification discloses how embodiments of the present invention "pin-point[] the most probable reason for the resource contention, such as a performance bottleneck that prevents a system from achieving its average level of performance." Claim 1 embodies these improvements at least by training a model by "feeding, by the one or more processors, a first part of the contention-related data to a first machine-learning system," and "feeding, by the one or more processors, a second part of the contention-related data scaled with the first impact values to a second machine learning system." (emphasis added).
Examiner’s response:
The examiner understands the applicant’s assertion.
However, it appears that each processing step is just applying the abstract idea to a general field of endeavor with additional elements. In addition, improvements to technology or technical field are not necessarily reflected in the claims. Thus, the claim does not integrate the judicial exception into a practical application, and the claim does not amount to significantly more than the judicial exception.
The examiner understands the applicant’s assertion “the alleged abstract idea does not fall within any of the subject matter groupings of abstract ideas enumerated in the 2019 Guidance”.
However, the application has not explained why claim 1 is not an abstract idea. As rejected under Claim Rejections - 35 USC § 101, it appears that each processing step is just applying the abstract idea to a general field of endeavor with additional elements. Explaining the reason in detail may help overcome the existing rejections.
The examiner understands the applicant’s assertion “claim 1 recites "feeding, by the one or more processors, a first part of the contention-related data to a first machine-learning system," and "feeding, by the one or more processors, a second part of the contention-related data scaled with the first impact values to a second machine-learning system." By integrating these features to the environment of a system having different workload groups, claim 1 integrates the features - and any associated abstract ideas - into a practical application (e.g. addressing bottlenecks), thereby rendering claim 1 subject matter eligible” and “For example, the specification discloses how embodiments of the present invention "pin-point[] the most probable reason for the resource contention, such as a performance bottleneck that prevents a system from achieving its average level of performance."”.
However, as rejected under Claim Rejections - 35 USC § 101, the “feeding” steps may be interpreted as providing data to machine learning systems with different inputs. Even though the impact values from the first ML system are used for scaling the input to the second ML system, it is not clear how/why the feature provides the alleged improvements. Explaining the feature in detail may help overcome the existing rejections. In addition, it is not clear how claim 1 pin-points the most probable reason for the resource contention. Explaining how claim 1 does prevent a performance bottleneck may help overcome the existing rejections.
The limitations do not clearly show e.g., improvements in computer technology and improvements to other technical fields. Rather, the improvements in Remarks are about just improving the abstract ideas of the independent claims. It doesn’t seem that the specification and/or the independent claims clearly show how the inventive concept of the claims enables improvements and how they are tied together. The applicant may need to amend the claims to show how the claim languages and improvements are tied together.
To find a valid improvement to a technology, MPEP 2106.04(d)(1) says the specification must explain the improvement and that the claim must reflect the disclosed improvement. Furthermore, the improvement should not be merely a consequence of the abstract idea. See MPEP 2106.05(a). An improvement in the abstract idea itself is not an improvement to technology.
For at least these reasons, Applicant's arguments are not convincing.
The Examiner encourages Applicant to schedule an interview to discuss issues related to, for example, the rejections noted below under 35 U.S.C § 101.
Applicant’s arguments regarding 35 USC § 103 with respect to the independent claims have been considered but are moot because the arguments are directed to amended limitation(s) that has/have not been previously examined.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Regarding claim 1
The claim is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: The claim recites a method; therefore, it falls into the statutory category of processes.
Step 2A Prong 1:
The limitations of
“… for identifying a cause of a performance anomaly of a computer system executing workloads in different workload groups, the method comprising:
…;
separating, …, contention-related data and non-contention related data within the received system performance data;
…, … provides a prediction of first contention instances and related first impact values as output; and
… for predicting second contention instances and related second impact values for the different workload groups as output”, as drafted, under its broadest reasonable interpretation, covers performance of the limitation in the mind. That is, nothing in the claim element precludes the step from practically being performed in the mind. For example, the limitations in the context of this claim encompass the user mentally thinking with a physical aid (e.g., pencil and paper).
If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
The claim recites additional elements that are mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea. See MPEP 2106.05(f). In particular, the claim recites an additional element(s) (“A computer-implemented”, “by one or more processors”, “wherein the first machine-learning model”, “by the one or more processors”) – using a device and a model to process data. The device and the model in each step are recited at a high-level of generality (i.e., as a generic computer performing a generic computer function of processing data) such that it amounts no more than mere instructions to apply the exception using a generic computer component. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
In particular, the claim recites an additional element(s) (“receiving, by one or more processors, system performance data”) – the act of receiving data. The claim is adding an insignificant extra-solution activity to the judicial exception – see MPEP 2106.05(g). The act of receiving data is recited at a high-level of generality (i.e., as a generic act of receiving performing a generic act function of receiving data) such that it amounts no more than a mere act to apply the exception using a generic act of receiving. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
In particular, the claim recites an additional element(s) (“feeding, by the one or more processors, a first part of the contention-related data to a first machine-learning system comprising a trained first machine-learning model”, “feeding, by the one or more processors, a second part of the contention-related data scaled with the first impact values to a second machine-learning system comprising a trained second machine-learning model”) – the act of providing (i.e. inputting) data. The claim is adding an insignificant extra-solution activity to the judicial exception – see MPEP 2106.05(g). The act of inputting data is recited at a high-level of generality (i.e., as a generic act of inputting performing a generic act function of inputting data) such that it amounts no more than a mere act to apply the exception using a generic act of inputting. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
In particular, the claim recites an additional element (“wherein the first contention instances are system-wise contention instances and the second contention instances are workgroup contention instances.”). This is a recitation of a particular type or source of model/data to be used in performing the abstract idea. Limiting the abstract idea to a particular type or source of model/data is an attempt to limit the abstract idea to a particular field of use or technological environment, which does not integrate the abstract idea into a practical application. See MPEP 2106.05(h)
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
As discussed above, with respect to integration of the abstract idea into a practical application, the additional elements of using a generic computer component to perform each step amount to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The claim is not patent eligible. MPEP 2106.05(f).
As discussed above, the claim recites the additional element(s) of receiving data at a high-level of generality and is adding an insignificant extra-solution activity – see MPEP 2106.05(g). However, the addition of insignificant extra-solution activity does not amount to an inventive concept, particularly when the activity is well-understood, routine, and conventional. See MPEP 2106.05(d)(II) – “Receiving or transmitting data over a network” or “Storing and retrieving information in memory”. Accordingly, this additional element does not provide an inventive concept and significantly more than the abstract idea. Thus, the claim is not patent eligible.
As discussed above, the claim recites the additional element(s) of inputting data at a high-level of generality and is adding an insignificant extra-solution activity – see MPEP 2106.05(g) – “Mere Data Gathering”. However, the addition of insignificant extra-solution activity does not amount to an inventive concept, particularly when the activity is well-understood, routine, and conventional. See MPEP 2106.05(d)(II) – “Receiving or transmitting data over a network” or “Storing and retrieving information in memory”. Accordingly, this additional element does not provide an inventive concept and significantly more than the abstract idea. Thus, the claim is not patent eligible.
This is a recitation of a particular type or source of model/data to be used in performing the abstract idea. Limiting the abstract idea to a particular type or source of model/data is an attempt to limit the abstract idea to a particular field of use or technological environment, which does not amount to significantly more than the abstract idea. See MPEP 2106.05(h).
Regarding claim 2
The claim is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: The claim recites a method; therefore, it falls into the statutory category of processes.
Step 2A Prong 1:
The limitations of
“analyzing, …, performance metric values for non-contention cases by fitting a number of Gaussian components to the performance metric values for non-contention cases of each of the different workload groups; and
in response to determining a workload group comprising more than one Gaussian components, splitting, …, the workload group into two workload groups”, as drafted, are a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. That is, nothing in the claim element precludes the step from practically being performed in the mind. For example, the limitations in the context of this claim encompass the user mentally thinking with a physical aid (e.g., pencil and paper).
If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
The claim recites additional elements that are mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea. See MPEP 2106.05(f). In particular, the claim recites an additional element(s) (“by the one or more processors”) – using a device and a model to process data. The device and the model in each step are recited at a high-level of generality (i.e., as a generic computer performing a generic computer function of processing data) such that it amounts no more than mere instructions to apply the exception using a generic computer component. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
As discussed above, with respect to integration of the abstract idea into a practical application, the additional elements of using a generic computer component to perform each step amount to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The claim is not patent eligible. MPEP 2106.05(f).
Regarding claim 3
The claim is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: The claim recites a method; therefore, it falls into the statutory category of processes.
Step 2A Prong 1: The claim recites the abstract idea identified above regarding claim 1.
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
In particular, the claim recites an additional element(s) (“feeding, …, a first part of contention-related training data as input to a Tree-structured Parzen Estimator …”) – the act of providing (i.e. inputting) data. The claim is adding an insignificant extra-solution activity to the judicial exception – see MPEP 2106.05(g). The act of inputting data is recited at a high-level of generality (i.e., as a generic act of inputting performing a generic act function of inputting data) such that it amounts no more than a mere act to apply the exception using a generic act of inputting. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
The claim recites additional elements that are mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea. See MPEP 2106.05(f). In particular, the claim recites an additional element(s) (“by the one or more processors”) – using a device and a model to process data. The device and the model in each step are recited at a high-level of generality (i.e., as a generic computer performing a generic computer function of processing data) such that it amounts no more than mere instructions to apply the exception using a generic computer component. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
In particular, the claim recites an additional element(s) (“to adapt hyper-parameter values of the first machine-learning model”). The additional element is recited at such a high level without any details as to how a model is trained such that it amounts to only the idea of a solution or outcome because it fails to recite details of how a solution to a problem is accomplished, and, therefore, represents no more than mere instructions to apply the judicial exception on a computer (see MPEP 2106.05(f)). Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
As discussed above, the claim recites the additional element(s) of inputting data at a high-level of generality and is adding an insignificant extra-solution activity – see MPEP 2106.05(g) – “Mere Data Gathering”. However, the addition of insignificant extra-solution activity does not amount to an inventive concept, particularly when the activity is well-understood, routine, and conventional. See MPEP 2106.05(d)(II) – “Receiving or transmitting data over a network” or “Storing and retrieving information in memory”. Accordingly, this additional element does not provide an inventive concept and significantly more than the abstract idea. Thus, the claim is not patent eligible.
As discussed above, with respect to integration of the abstract idea into a practical application, the additional elements of using a generic computer component to perform each step amount to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The claim is not patent eligible. MPEP 2106.05(f).
The additional elements regarding training are recited at such a high level without any details as to how a model is trained such that it amounts to only the idea of a solution or outcome because it fails to recite details of how a solution to a problem is accomplished, and, therefore, represents no more than mere instructions to apply the judicial exception on a computer (see MPEP 2106.05(f)). Accordingly, this additional element does not amount to significantly more than the abstract idea. The claim is directed to an abstract idea.
Regarding claim 4
The claim is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: The claim recites a method; therefore, it falls into the statutory category of processes.
Step 2A Prong 1: The claim recites the abstract idea identified above regarding claim 1.
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
In particular, the claim recites an additional element (“wherein the first machine-learning system is a Gradient Boosted Tree machine-learning system”). This is a recitation of a particular type or source of data to be used in performing the abstract idea. Limiting the abstract idea to a particular type or source of data is an attempt to limit the abstract idea to a particular field of use or technological environment, which does not integrate the abstract idea into a practical application. See MPEP 2106.05(h)
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
This is a recitation of a particular type or source of model/data to be used in performing the abstract idea. Limiting the abstract idea to a particular type or source of model/data is an attempt to limit the abstract idea to a particular field of use or technological environment, which does not amount to significantly more than the abstract idea. See MPEP 2106.05(h).
Regarding claim 5
The claim is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: The claim recites a method; therefore, it falls into the statutory category of processes.
Step 2A Prong 1: The claim recites the abstract idea identified above regarding claim 1.
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
In particular, the claim recites an additional element (“wherein the second machine-learning system is a Gradient Boosted Tree machine-learning system”). This is a recitation of a particular type or source of data to be used in performing the abstract idea. Limiting the abstract idea to a particular type or source of data is an attempt to limit the abstract idea to a particular field of use or technological environment, which does not integrate the abstract idea into a practical application. See MPEP 2106.05(h)
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
This is a recitation of a particular type or source of model/data to be used in performing the abstract idea. Limiting the abstract idea to a particular type or source of model/data is an attempt to limit the abstract idea to a particular field of use or technological environment, which does not amount to significantly more than the abstract idea. See MPEP 2106.05(h).
Regarding claim 6
The claim is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: The claim recites a method; therefore, it falls into the statutory category of processes.
Step 2A Prong 1: The claim recites the abstract idea identified above regarding claim 1.
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
In particular, the claim recites an additional element(s) (“feeding, …, the output of the first machine-learning system, the output of the second machine-learning system and performance metric values to a performance visualization system”) – the act of providing (i.e. inputting) data. The claim is adding an insignificant extra-solution activity to the judicial exception – see MPEP 2106.05(g). The act of inputting data is recited at a high-level of generality (i.e., as a generic act of inputting performing a generic act function of inputting data) such that it amounts no more than a mere act to apply the exception using a generic act of inputting. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
The claim recites additional elements that are mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea. See MPEP 2106.05(f). In particular, the claim recites an additional element(s) (“by the one or more processors”) – using a device and a model to process data. The device and the model in each step are recited at a high-level of generality (i.e., as a generic computer performing a generic computer function of processing data) such that it amounts no more than mere instructions to apply the exception using a generic computer component. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
As discussed above, the claim recites the additional element(s) of inputting data at a high-level of generality and is adding an insignificant extra-solution activity – see MPEP 2106.05(g) – “Mere Data Gathering”. However, the addition of insignificant extra-solution activity does not amount to an inventive concept, particularly when the activity is well-understood, routine, and conventional. See MPEP 2106.05(d)(II) – “Receiving or transmitting data over a network” or “Storing and retrieving information in memory”. Accordingly, this additional element does not provide an inventive concept and significantly more than the abstract idea. Thus, the claim is not patent eligible.
As discussed above, with respect to integration of the abstract idea into a practical application, the additional elements of using a generic computer component to perform each step amount to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The claim is not patent eligible. MPEP 2106.05(f).
Regarding claim 7
The claim is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: The claim recites a method; therefore, it falls into the statutory category of processes.
Step 2A Prong 1:
The limitations of
“predicting, …, a possible contention case using a time-series analysis of the second impact values and first contention instances and/or second contention instances”, as drafted, under its broadest reasonable interpretation, covers performance of the limitation in the mind. That is, nothing in the claim element precludes the step from practically being performed in the mind. For example, the limitations in the context of this claim encompass the user mentally thinking with a physical aid (e.g., pencil and paper).
If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
The claim recites additional elements that are mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea. See MPEP 2106.05(f). In particular, the claim recites an additional element(s) (“by the one or more processors”) – using a device and a model to process data. The device and the model in each step are recited at a high-level of generality (i.e., as a generic computer performing a generic computer function of processing data) such that it amounts no more than mere instructions to apply the exception using a generic computer component. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
As discussed above, with respect to integration of the abstract idea into a practical application, the additional elements of using a generic computer component to perform each step amount to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The claim is not patent eligible. MPEP 2106.05(f).
Regarding claim 8
The claim is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: The claim recites a method; therefore, it falls into the statutory category of processes.
Step 2A Prong 1:
The limitations of
“determining, …, that two or more workload groups perform concurrently worse than a predefined number of standard deviations from an average performance of the two or more workload groups within a predefined period of time”, as drafted, under its broadest reasonable interpretation, covers performance of the limitation in the mind. That is, nothing in the claim element precludes the step from practically being performed in the mind. For example, the limitations in the context of this claim encompass the user mentally thinking with a physical aid (e.g., pencil and paper).
If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
The claim recites additional elements that are mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea. See MPEP 2106.05(f). In particular, the claim recites an additional element(s) (“by the one or more processors”) – using a device and a model to process data. The device and the model in each step are recited at a high-level of generality (i.e., as a generic computer performing a generic computer function of processing data) such that it amounts no more than mere instructions to apply the exception using a generic computer component. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
As discussed above, with respect to integration of the abstract idea into a practical application, the additional elements of using a generic computer component to perform each step amount to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The claim is not patent eligible. MPEP 2106.05(f).
Regarding claim 9
The claim is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: The claim recites a method; therefore, it falls into the statutory category of processes.
Step 2A Prong 1: The claim recites the abstract idea identified above regarding claim 1.
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
In particular, the claim recites an additional element (“wherein the predefined period of time is at a minimum one minute”). This is a recitation of a particular type or source of data to be used in performing the abstract idea. Limiting the abstract idea to a particular type or source of data is an attempt to limit the abstract idea to a particular field of use or technological environment, which does not integrate the abstract idea into a practical application. See MPEP 2106.05(h)
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
This is a recitation of a particular type or source of model/data to be used in performing the abstract idea. Limiting the abstract idea to a particular type or source of model/data is an attempt to limit the abstract idea to a particular field of use or technological environment, which does not amount to significantly more than the abstract idea. See MPEP 2106.05(h).
Regarding claim 10
The claim is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: The claim recites a method; therefore, it falls into the statutory category of processes.
Step 2A Prong 1:
The limitations of
“referring, …, to a normalized performance index metric for each of the different workload groups as part of the second part of the contention-related data”, as drafted, under its broadest reasonable interpretation, covers performance of the limitation in the mind. That is, nothing in the claim element precludes the step from practically being performed in the mind. For example, the limitations in the context of this claim encompass the user mentally thinking with a physical aid (e.g., pencil and paper).
If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
The claim recites additional elements that are mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea. See MPEP 2106.05(f). In particular, the claim recites an additional element(s) (“by the one or more processors”) – using a device and a model to process data. The device and the model in each step are recited at a high-level of generality (i.e., as a generic computer performing a generic computer function of processing data) such that it amounts no more than mere instructions to apply the exception using a generic computer component. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
As discussed above, with respect to integration of the abstract idea into a practical application, the additional elements of using a generic computer component to perform each step amount to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The claim is not patent eligible. MPEP 2106.05(f).
Regarding claim 11
The claim recites “A computer system for contention detection and cause identification of a performance anomaly of a computer system executing workloads including different workload groups, the system comprising: one or more processors; and one or more computer-readable storage media, and program instructions stored on the one or more computer-readable storage media, wherein the program instructions, when executed, enable the one or more processors to” to perform precisely the method of Claim 1. As performance of an abstract idea on generic computer components (see MPEP 2106.05(f)) and “Storing and retrieving information in memory” (see MPEP 2106.05(g) on Insignificant Extra-Solution Activity, and MPEP 2106.05(d) on Well-Understood, Routine, Conventional Activity) and “Field of Use and Technological Environment” (see MPEP 2106.05(h)) cannot integrate the abstract idea into a practical application nor provide significantly more than the abstract idea itself, the claim is rejected for reasons set forth in the rejection of Claim 1.
Regarding claim 12
The claim is rejected for the reasons set forth in the rejection of Claim 2 under 35 U.S.C. 101, mutatis mutandis, as reciting an abstract idea without integrating the judicial exception into a practical application nor providing significantly more than the judicial exception.
Regarding claim 13
The claim is rejected for the reasons set forth in the rejection of Claim 3 under 35 U.S.C. 101, mutatis mutandis, as reciting an abstract idea without integrating the judicial exception into a practical application nor providing significantly more than the judicial exception.
Regarding claim 14
The claim is rejected for the reasons set forth in the rejection of Claim 4 under 35 U.S.C. 101, mutatis mutandis, as reciting an abstract idea without integrating the judicial exception into a practical application nor providing significantly more than the judicial exception.
Regarding claim 15
The claim is rejected for the reasons set forth in the rejection of Claim 5 under 35 U.S.C. 101, mutatis mutandis, as reciting an abstract idea without integrating the judicial exception into a practical application nor providing significantly more than the judicial exception.
Regarding claim 16
The claim is rejected for the reasons set forth in the rejection of Claim 6 under 35 U.S.C. 101, mutatis mutandis, as reciting an abstract idea without integrating the judicial exception into a practical application nor providing significantly more than the judicial exception.
Regarding claim 17
The claim is rejected for the reasons set forth in the rejection of Claim 7 under 35 U.S.C. 101, mutatis mutandis, as reciting an abstract idea without integrating the judicial exception into a practical application nor providing significantly more than the judicial exception.
Regarding claim 18
The claim is rejected for the reasons set forth in the rejection of Claim 8 under 35 U.S.C. 101, mutatis mutandis, as reciting an abstract idea without integrating the judicial exception into a practical application nor providing significantly more than the judicial exception.
Regarding claim 19
The claim is rejected for the reasons set forth in the rejection of Claim 9 under 35 U.S.C. 101, mutatis mutandis, as reciting an abstract idea without integrating the judicial exception into a practical application nor providing significantly more than the judicial exception.
Regarding claim 20
The claim recites “A computer program product for identifying a cause of a performance anomaly of a computer system executing workloads in different workload groups, the computer program product comprising: a computer readable storage medium having program instructions embodied therewith, the program instructions, when executed, cause the program instructions to” to perform precisely the method of Claim 1. As performance of an abstract idea on generic computer components (see MPEP 2106.05(f)) and “Storing and retrieving information in memory” (see MPEP 2106.05(g) on Insignificant Extra-Solution Activity, and MPEP 2106.05(d) on Well-Understood, Routine, Conventional Activity) cannot integrate the abstract idea into a practical application nor provide significantly more than the abstract idea itself, the claim is rejected for reasons set forth in the rejection of Claim 1.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
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, 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.
Claim(s) 1, 11, 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over MITRA et al. (US 20190354388 A1) Zhang et al. (Cascaded Random Forest for Hyperspectral Image Classification)
Regarding claim 1
MITRA teaches
A computer-implemented method for identifying a cause of a performance anomaly of a computer system executing workloads in different workload groups, the method comprising:
(MITRA [fig(s) 5-6] [par(s) 3] “Embodiments of the present disclosure relate to tenant-side detection and mitigation of performance degradation resulting from interference generated by a noisy neighbor in a distributed computing environment. Generally, a first machine-learning model such as a k-means nearest neighbor classifier can be operated by a tenant to detect an anomaly with a computer system emulator (e.g., a virtual machine, software container, etc.) in the distributed computing environment resulting from a co-located noisy neighbor. A second machine-learning model such as a multi-class classifier can be operated by the tenant to identify a contended resource associated with the anomaly. A corresponding trigger signal can be generated and provided to trigger various mitigation responses, including an application/framework-specific mitigation strategy (e.g., triggered approximations in application/framework performance, best-efforts paths, run-time changes, etc.), load-balancing, scaling out, updates to a scheduler to avoid impacted nodes, and others.” [par(s) 4] “In order to generate training data for the models, interference is simulated and application and/or framework-level metrics are collected from the nodes of the tenant (including the nodes on which the identified co-located emulators reside). The collected metrics can be normalized across all associated emulators (e.g., across all virtual machines performing a particular distributed task), for example, with respect to a representative characteristic of the distributed task. Nodes without impacted metrics are labeled healthy, and nodes with impacted metrics are labeled anomalies;
Examiner notes that paragraph 2 of the Instant Specification describes “Thereby, each service class has underlying related goals and priorities that provide essential information for the workload management tool and how to manage the different jobs (i.e., workloads).”)
receiving, by one or more processors, system performance data;
(MITRA [fig(s) 5-6] [par(s) 4] “In order to generate training data for the models, interference is simulated and application and/or framework-level metrics are collected from the nodes of the tenant (including the nodes on which the identified co-located emulators reside). The collected metrics can be normalized across all associated emulators (e.g., across all virtual machines performing a particular distributed task), for example, with respect to a representative characteristic of the distributed task. Nodes without impacted metrics are labeled healthy, and nodes with impacted metrics are labeled anomalies.”;)
separating, by the one or more processors, contention-related data and non-contention related data within the received system performance data;
(MITRA [fig(s) 5-6] [par(s) 4] “In order to generate training data for the models, interference is simulated and application and/or framework-level metrics are collected from the nodes of the tenant (including the nodes on which the identified co-located emulators reside). The collected metrics can be normalized across all associated emulators (e.g., across all virtual machines performing a particular distributed task), for example, with respect to a representative characteristic of the distributed task. Nodes without impacted metrics are labeled healthy, and nodes with impacted metrics are labeled anomalies.”;)
feeding, by the one or more processors, a first part of the contention-related data to a first machine-learning system comprising a trained first machine-learning model, wherein the first machine-learning model provides a prediction of first contention instances and related first impact values as output; and
(MITRA [fig(s) 5-6] [par(s) 4] “In order to generate training data for the models, interference is simulated and application and/or framework-level metrics are collected from the nodes of the tenant (including the nodes on which the identified co-located emulators reside). The collected metrics can be normalized across all associated emulators (e.g., across all virtual machines performing a particular distributed task), for example, with respect to a representative characteristic of the distributed task. Nodes without impacted metrics are labeled healthy, and nodes with impacted metrics are labeled anomalies.” [par(s) 5] “A first machine learning model (e.g., k-nearest neighbor classifier) can accept as an input a feature vector generated from the normalized, collected metrics. The first machine learning model can be trained using the labeled training data to identify an anomalous node with performance degradation resulting from interference from a noisy neighbor.” [par(s) 23-26] “multiple anomaly detectors can be applied, for example, with a single anomaly detector dedicated to identifying anomalies in one or more nodes and another anomaly detector dedicated to identifying anomalies in the remaining nodes.” and “the output of the anomaly detector and/or resource classifier can indicate if an application is suffering from interference, and if so, which resource is being contended. If the anomaly detector and/or resource classifier indicates resource contention, a corresponding signal can be generated to trigger a designated mitigation response to reduce pressure on a contended resource and achieve low latency in the presence of interference.” [par(s) 38] “These benchmark tests can be performed on various physical machines (e.g., node) on which the tenant operates (e.g., nodes used by or otherwise associated with a particular application or frame work), including the nodes on which the identified co located computer system emulators operate.” [par(s) 41-48] “Generally, the models described herein ( e.g., anomaly detector 110 and/or resource classifier 112) can be used to detect and classify interference issues associated with operating a particular application in a distributed computing environment. … a tenant or customer thereof can automatically detect, classify, and mitigate performance degradation resulting from a noisy neighbor. These techniques can enable tenants to detect impacted applications, identify a contended resource, and take steps to mitigate the contention. Various application-level and/or framework-level metrics can be used to build machine-learning models to automatically detect and classify contentions.”; e.g., “identify an anomalous node with performance degradation resulting from interference from a noisy neighbor” along with “detect and classify interference issues” and “detect impacted applications, identify a contended resource” read(s) on “prediction of first contention instances and related first impact values as output”.)
(Note: Hereinafter, if a limitation has bold brackets (i.e. [·]) around claim languages, the bracketed claim languages indicate that they have not been taught yet by the current prior art reference but they will be taught by another prior art reference afterwards.)
feeding, by the one or more processors, a second part of the contention-related data [scaled with] the first impact values to a second machine-learning system comprising a trained second machine-learning model for predicting second contention instances and related second impact values for the different workload groups as output, wherein the first contention instances are system-wise contention instances and the second contention instances are workgroup contention instances.
(MITRA [fig(s) 5-6] [par(s) 4] “In order to generate training data for the models, interference is simulated and application and/or framework-level metrics are collected from the nodes of the tenant (including the nodes on which the identified co-located emulators reside). The collected metrics can be normalized across all associated emulators (e.g., across all virtual machines performing a particular distributed task), for example, with respect to a representative characteristic of the distributed task. Nodes without impacted metrics are labeled healthy, and nodes with impacted metrics are labeled anomalies.” [par(s) 5] “Training data for a second machine-learning model can be generated by determining the Euclidean distance between a feature vector reflecting the collected application-level and/or framework-level metrics and the centroids of corresponding clusters, and generating labels according to the corresponding resource contention that was simulated. The resulting training data can be used to train the second machine-learning model to identify the particular resource being contended.” [par(s) 23-25] “the output of the anomaly detector and/or resource classifier can indicate if an application is suffering from interference, and if so, which resource is being contended. If the anomaly detector and/or resource classifier indicates resource contention, a corresponding signal can be generated to trigger a designated mitigation response to reduce pressure on a contended resource and achieve low latency in the presence of interference.” [par(s) 38] “These benchmark tests can be performed on various physical machines (e.g., node) on which the tenant operates (e.g., nodes used by or otherwise associated with a particular application or frame work), including the nodes on which the identified co located computer system emulators operate.” [par(s) 41-48] “Generally, the models described herein ( e.g., anomaly detector 110 and/or resource classifier 112) can be used to detect and classify interference issues associated with operating a particular application in a distributed computing environment. … a tenant or customer thereof can automatically detect, classify, and mitigate performance degradation resulting from a noisy neighbor. These techniques can enable tenants to detect impacted applications, identify a contended resource, and take steps to mitigate the contention. Various application-level and/or framework-level metrics can be used to build machine-learning models to automatically detect and classify contentions.”; e.g., “identify the particular resource being contended” along with “the output of the anomaly detector and/or resource classifier can indicate if an application is suffering from interference”, “detect and classify interference issues” and “detect impacted applications, identify a contended resource” read(s) on “predicting second contention instances and related second impact values”. In addition, e.g., “Nodes without impacted metrics are labeled healthy, and nodes with impacted metrics are labeled anomalies” read(s) on “system-wise contention instances” and “workgroup contention instances”.)
However, MITRA does not appear to explicitly teach:
feeding, by the one or more processors, a second part of the contention-related data [scaled with] the first impact values to a second machine-learning system comprising a trained second machine-learning model for predicting second contention instances and related second impact values for the different workload groups as output.
(Note: Hereinafter, if a limitation has one or more bold underlines, the one or more underlined claim languages indicate that they are taught by the current prior art reference, while the one or more non-underlined claim languages indicate that they have been taught already by one or more previous art references.)
Zhang teaches
feeding, by the one or more processors, a second part of the contention-related data scaled with the first impact values to a second machine-learning system comprising a trained second machine-learning model for predicting second contention instances and related second impact values for the different workload groups as output.
(Zhang [fig(s) 1] “Training samples”, “Decision trees”, “Error”, “Weight” and “Updating sample weight” [sec(s) III] “We can summarize the training process of CRF method as follows: 1) all feature significance are computed by neighborhood rough sets theory; 2) the mean value of these feature significance ε is calculated, and the feature set is divided into two subsets B1 and B2 according to threshold ε; 3) S subsets of training samples are selected by probability weight distribution w(t)i; 4) select S feature subsets corresponding to S training subsets, each feature subset is obtained by randomly selecting half number of the m features from B1 and B2 , respectively; 5) train S decision trees on S training subsets corresponding to their feature subsets; 6) compute the OOB error of each decision tree OEts and obtain the weight of each decision tree αts ; 7) compute the OOB error of S decision trees Et and obtain the weight of S decision trees βt; 8) update the weights of samples w(t+1)i according to βt; and 9) repeat step 3 to step 7 T times. The detailed training and classification steps of CRF is described in Algorithm 1.”;)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the system of MITRA with the scaling of Zhang.
One of ordinary skill in the art would have been motived to combine in order to enhance the strength of individual classifier by means of improved feature selection or enhanced resampling methods, and improve the strength of the individual classifier while increasing the diversity between each two classifiers simultaneously.
(Zhang [sec(s) III] “In most cases, these methods only enhance the strength of individual decision trees by means of improved feature selection or enhanced resampling methods. Few of them consider improving the strength of the individual classifier and increasing the diversity between each two classifiers simultaneously. … In this paper, our proposed CRF method achieves the above two aspects at the same time. In particular, HRSM used for feature selection can improve the strength of decision trees and increase the diversity between each two of the random forests. Besides, minimization of the OOB error in the procedure of Boosting iteration can increase the strength of decision trees iteratively. The flow of CRF is illustrated in Fig. 1.”)
Regarding claim 11
The claim is a system claim corresponding to the method claim 1, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejections of the method claim.
Regarding claim 20
The claim is a computer program product claim corresponding to the method claim 1, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejections of the method claim.
Claim(s) 2, 12 is/are rejected under 35 U.S.C. 103 as being unpatentable over MITRA et al. (US 20190354388 A1) Zhang et al. (Cascaded Random Forest for Hyperspectral Image Classification) in view of Zhu et al. (Estimating power consumption of servers using gaussian mixture model) in view of KHOSHKBARFOROUSHHA et al. (Distribution Based Workload Modelling of Continuous Queries in Clouds)
Regarding claim 2
The combination of MITRA, Zhang teaches claim 1.
MITRA further teaches
analyzing, by the one or more processors, performance metric values for non-contention cases [by fitting a number of Gaussian components to] the performance metric values for non-contention cases of each of the different workload groups; and
(MITRA [par(s) 4] “In order to generate training data for the models, interference is simulated and application and/or framework-level metrics are collected from the nodes of the tenant (including the nodes on which the identified co-located emulators reside). The collected metrics can be normalized across all associated emulators (e.g., across all virtual machines performing a particular distributed task), for example, with respect to a representative characteristic of the distributed task. Nodes without impacted metrics are labeled healthy, and nodes with impacted metrics are labeled anomalies.”;)
However, the combination of MITRA, Zhang does not appear to explicitly teach:
analyzing, by the one or more processors, performance metric values for non-contention cases [by fitting a number of Gaussian components to] the performance metric values for non-contention cases of each of the different workload groups; and
in response to determining a workload group comprising more than one Gaussian components, splitting, by the one or more processors, the workload group into two workload groups.
Zhu teaches
analyzing, by the one or more processors, performance metric values for non-contention cases by fitting a number of Gaussian components to the performance metric values for non-contention cases of each of the different workload groups; and
(Zhu [algorithm 1] “centroids, ccov, p k, logL ← gmm.em_gm(TrainingSet[X, Y], K, Iterations = 800, clustering = kmeans)” [sec(s) 1] “Main idea of scheduling strategies is to find out switching on/off which servers and managing resources at which usage level to keep energy consumption in proportionality with computing demand.” [sec(s) III.B] “Non-linear means that the change rate of power consumption is distinct for different ranges of resource usage. For instance, power consumption grows slowly as resource usage increase from idle [17] or resources meet the bandwidth bottleneck for intensive workloads [13]; power consumption could rise when medium resource usage. … GMM [18] is an unsupervised machine learning model, usually used as information classification. Each classification creates a cluster. GMM is a parametric probability density function represented as a weighted sum of Gaussian cluster densities. The relationship between power consumption and resource features at different levels could be better represented using multi-cluster parameters. … We train models using different values of K, and choose one of values which causes the least error on the VS. After that, we can train our GMM model and calculate its errors according to the pseudocodes shown in Algorithm 1. At first, we get TrS and TeS from a complete data set of one server (line 4-5). Second, we use iterative Expectation-Maximization [20] algorithm to find GMM clusters. The parameters of each cluster include cluster center centroid, cluster co-variances matrices ccov and mixing coefficients p_k (line 6). Third, given the value of one feature vector from data matrix in terms of resource features TestSet[X], this feature vector’s conditional probability density function can be obtained based on the cluster parameters in second step [21]. We get parameters of conditional probability density function for given TestSet[X]: cluster center, cluster covariances matrices and mixing coefficients (line 7). Finally, the power consumption Estimated Power for this given TestSet[X] can be calculated as the expected value of the conditional probability density function, which is the sum of the mix coefficient for each cluster multiplied by the center of the cluster (line 8).” [sec(s) IV.B] “We train all the models and estimate power on a server node with Intel E5620 processor (2.4GHz) and 24 GB memory”;)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the system of MITRA, Zhang with the Gaussian component fitting of Zhu.
One of ordinary skill in the art would have been motived to combine in order to provide the best computing resource estimation accuracy compared to the conventional approaches.
(Zhu [sec(s) V] “Besides GMM, we focus on other non-linear models. We see that the non-linear models do improve the estimation accuracy of the linear models using basic resource features. We prove that models trained using the system-level full features have the highest accuracy comparing to only use part of features. The GMM model has the best power estimation accuracy compared with the neural network models and the linear models in the I/O or CPU-intensive cluster.”)
However, the combination of MITRA, Zhang, Zhu does not appear to explicitly teach:
in response to determining a workload group comprising more than one Gaussian components, splitting, by the one or more processors, the workload group into two workload groups.
KHOSHKBARFOROUSHHA teaches
in response to determining a workload group comprising more than one Gaussian components, splitting, by the one or more processors, the workload group into two workload groups.
(KHOSHKBARFOROUSHHA [fig(s) 3] “Our approach builds an MDN model based on the historical logs of queries to pedict distribution of new incoming workloads” [fig(s) 10] “Sample PDF predictions for (a) CPU and (b) memory usage from TPC-H workload” [fig(s) 5] [sec(s) IV.C] “Our approach employs MDN [7], a special type of Artificial Neural Network (ANN), in which the target (e.g., CPU usage) is represented as a conditional PDF. The conditional distribution represents a complete description of data generation. An MDN fuses a mixture model with an ANN. We utilize a Gaussian Mixture Model (GMM) based MDN where the conditional density functions are represented by a weighted mixture of Gaussians. … zαi, zσi, and zμi are the outputs of the neural network corresponding to the mixture weights, variance, and mean for the ith Gaussian component in the GMM, given x [7].” [sec(s) VI.A] “All queries were executed on m1.large instance size with 8 GB RAM, 4VCPU, and the same OS. The hypervisor is KVM, and the nodes are connected with 10 GB Ethernet. In our cloud each physical machine has 16 cores of Intel(R) Xeon(R) CPU 2.20 GHz with hyper threading enabled which the OS sees as 32 cores (CPU threads). Therefore, 4 VCPU map to 4 CPU threads and 2 full CPU cores. … There are a number of hyper-parameters including the number of Gaussian components or number of neurons in MLP that need to be specified beforehand. We evaluated several settings and assessed the trade-off between accuracy, training time and overhead. We concluded that a GMM with three components and two neurons per feature in the input vector provide an acceptable accuracy within a tolerable overhead” [sec(s) VI.D] “Figure 10 plots 14 random sample predicted PDFs for CPU and memory consumption in which they were selected from the model with 3 and 5 GMM components respectively. The histograms of the resource usage of the whole test dataset are also depicted. Each PDF may (not) belong to different queries as they were selected randomly from the test datasets, mean that they are conditioned on different inputs. The dotted vertical line shows the observation value.”; e.g., “We evaluated several settings and assessed the trade-off between accuracy, training time and overhead. We concluded that a GMM with three components and two neurons per feature in the input vector provide an acceptable accuracy within a tolerable overhead” read(s) on “determining a workload group comprising more than one Gaussian components”. In addition, e.g., fig 3 along with “zαi, zσi, and zμi are the outputs of the neural network corresponding to the mixture weights, variance, and mean for the ith Gaussian component in the GMM, given x” read(s) on “splitting, by the one or more processors, the workload group into two workload groups”.)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the system of MITRA, Zhang, Zhu with the workload group splitting of KHOSHKBARFOROUSHHA.
One of ordinary skill in the art would have been motived to combine in order to provide a complete description of the statistical properties of the resource utilization through which the system is not only able to capture the observation point, but also the whole spectrum of the resource usage.
(KHOSHKBARFOROUSHHA [sec(s) 1] “To illustrate one of the possible advantages of using the proposed approach, consider Figure 2. It displays a sample predicted PDF and actual CPU usage in terms of normalized histogram and fitted Kernel Density Estimation (KDE) for one of the experiments on linear road benchmark [5] queries. As we can see, the estimated PDF approximates the actual resource usage PDF closely. The predicted PDF provides a complete description of the statistical properties of the CPU utilization through which we are not only able to capture the observation point, but also the whole spectrum of the resource usage.”)
Regarding claim 12
The claim is a system claim corresponding to the method claim 2, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejections of the method claim.
Claim(s) 3, 13 is/are rejected under 35 U.S.C. 103 as being unpatentable over MITRA et al. (US 20190354388 A1) Zhang et al. (Cascaded Random Forest for Hyperspectral Image Classification) in view of Moradi et al. (uPredict: A User-Level Profiler-Based Predictive Framework for Single VM Applications in Multi-Tenant Clouds)
Regarding claim 3
The combination of MITRA, Zhang teaches claim 1.
MITRA teaches
feeding, by the one or more processors, a first part of contention-related training data as input to [a Tree-structured Parzen Estimator to adapt hyper-parameter] values of the first machine-learning model.
(MITRA [fig(s) 5-6] [par(s) 4] “In order to generate training data for the models, interference is simulated and application and/or framework-level metrics are collected from the nodes of the tenant (including the nodes on which the identified co-located emulators reside). The collected metrics can be normalized across all associated emulators (e.g., across all virtual machines performing a particular distributed task), for example, with respect to a representative characteristic of the distributed task. Nodes without impacted metrics are labeled healthy, and nodes with impacted metrics are labeled anomalies.” [par(s) 5] “A first machine learning model (e.g., k-nearest neighbor classifier) can accept as an input a feature vector generated from the normalized, collected metrics. The first machine learning model can be trained using the labeled training data to identify an anomalous node with performance degradation resulting from interference from a noisy neighbor.” [par(s) 23] “multiple anomaly detectors can be applied, for example, with a single anomaly detector dedicated to identifying anomalies in one or more nodes and another anomaly detector dedicated to identifying anomalies in the remaining nodes.”;)
However, the combination of MITRA, Zhang does not appear to explicitly teach:
feeding, by the one or more processors, a first part of contention-related training data as input to [a Tree-structured Parzen Estimator to adapt hyper-parameter] values of the first machine-learning model.
Moradi teaches
feeding, by the one or more processors, a first part of contention-related training data as input to a Tree-structured Parzen Estimator to adapt hyper-parameter values of the first machine-learning model.
(Moradi [fig(s) 1] [sec(s) 4.2] “The key step in the training phase is to learn the relationship between the application execution time and the resource contention represented by the micro-benchmarks’ access counter values from the collected data tuples and derive a predictive model. … To automatically optimize NN structures, we employed hyperparameter optimization techniques, including Tree-structured Parzen Estimator (TPE) approach and Bayesian Optimization [9, 54]. Hyperparameters refer to the NN parameters defining the number of layers and the number of neurons in each layer of a NN model. Both optimization techniques conduct a search in the optimization search space of the NN models to find a structure with good accuracy. This search space defines the maximum number of layers and neurons per layer that can be used when training NN models. The TPE technique explores the search space using a tree structure following the accuracy distribution obtained from previously sampled points in the search space. The Bayesian Optimization searches for the high-accuracy NN structures through the Gaussian Process (GP), which is a non-linear regression technique. Bayesian optimization uses GP to build a regression model with the already explored NN structures and their accuracies. The regression model is then used to predict a potentially better NN structure until a fixed number of NN structures are searched. … Note that, in uPredict, the hyperparameter optimization for NN models is applied on the training data sets themselves instead of separate cross-validation data sets. Theoretically, training and optimizing on the same data set may lead to over-fitting and thus low prediction accuracy. However, our experiment results show that using the same data set for optimizing the NN models can still provide high accuracy for predicting the performance of the considered benchmark applications and VMs on the multi-tenant clouds”;)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the system of MITRA, Zhang with the hyperparameters of Moradi.
One of ordinary skill in the art would have been motived to combine in order to achieve a reduction in average application execution times and a reduction in their turnaround times, and show the feasibility and effectiveness of the performance predictive framework for ordinary cloud users.
(Moradi [sec(s) V] “The results show that the uPredict based load-balancing can achieve about 19% reduction in average application execution times and 10% reduction in their turnaround times. … The evaluation results show the feasibility and effectiveness of uPredict for ordinary cloud users.”)
Regarding claim 13
The claim is a system claim corresponding to the method claim 3, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejections of the method claim.
Claim(s) 4-7, 14-17 is/are rejected under 35 U.S.C. 103 as being unpatentable over MITRA et al. (US 20190354388 A1) Zhang et al. (Cascaded Random Forest for Hyperspectral Image Classification) in view of POOLE et al. (US 20210073658 A1)
Regarding claim 4
The combination of MITRA, Zhang teaches claim 1.
However, the combination of MITRA, Zhang does not appear to explicitly teach:
wherein the first machine-learning system is a Gradient Boosted Tree machine-learning system.
POOLE teaches
wherein the first machine-learning system is a Gradient Boosted Tree machine-learning system.
(POOLE [fig(s) 3] [par(s) 87-91] “FIG. 3 illustrates the use of the Quantile Loss Gradient Boosted Trees (QLGBT) model-based multi-agent system that can be used to generate a tree map visual. For each category of metric (e.g., checkouts), a set of highly correlated metrics as independent regressors is determined. For each correlated pair of metrics, a time-independent generalized linear model (GLM) with polynomial relationships is fitted, as referenced by 302. Models other than a generalized linear model can be used as well. Using the regression error as a target along with temporal and exogenous variables, two gradient boosted tree (GBT) models are fitted around the prediction error, an example of which is referenced by 304. One gradient boosted tree (GBT) model can correspond to an upper bound and another gradient boosted tree (GBT) model can correspond to a lower bound.”;)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the system of MITRA, Zhang with the Gradient Boosted Tree of POOLE.
One of ordinary skill in the art would have been motived to combine in order to provide an improved way of representing a large number of metrics being monitored for a platform, and expose actionable and useful information associated with the platform in a manner that can be effectively interpreted by a user.
(POOLE [par(s) 6] “Therefore, via the display of the visuals and the implementation of the machine learning algorithms, the techniques described herein provide an improved way of representing a large number of metrics (e.g., hundreds, thousands, etc.) being monitored for a platform. Moreover, the techniques are configured to expose actionable and useful information associated with the platform in a manner that can be effectively interpreted by a user.”)
Regarding claim 5
The combination of MITRA, Zhang teaches claim 1.
However, the combination of MITRA, Zhang does not appear to explicitly teach:
wherein the second machine-learning system is a Gradient Boosted Tree machine-learning system.
POOLE teaches
wherein the second machine-learning system is a Gradient Boosted Tree machine-learning system.
(POOLE [fig(s) 3] [par(s) 87-91] “FIG. 3 illustrates the use of the Quantile Loss Gradient Boosted Trees (QLGBT) model-based multi-agent system that can be used to generate a tree map visual. For each category of metric (e.g., checkouts), a set of highly correlated metrics as independent regressors is determined. For each correlated pair of metrics, a time-independent generalized linear model (GLM) with polynomial relationships is fitted, as referenced by 302. Models other than a generalized linear model can be used as well. Using the regression error as a target along with temporal and exogenous variables, two gradient boosted tree (GBT) models are fitted around the prediction error, an example of which is referenced by 304. One gradient boosted tree (GBT) model can correspond to an upper bound and another gradient boosted tree (GBT) model can correspond to a lower bound.”;)
The combination of MITRA, Zhang is combinable with POOLE for the same rationale as set forth above with respect to claim 4.
Regarding claim 6
The combination of MITRA, Zhang teaches claim 1.
However, the combination of MITRA, Zhang does not appear to explicitly teach:
feeding, by the one or more processors, the output of the first machine-learning system, the output of the second machine-learning system and performance metric values to a performance visualization system.
POOLE teaches
feeding, by the one or more processors, the output of the first machine-learning system, the output of the second machine-learning system and performance metric values to a performance visualization system.
(POOLE [fig(s) 2] “Machine Learning-Based Agents 226” and “Multi-Agent Voting System 228”, “RADAR-BASED VISUAL” and “TREE MAP VISUAL” [fig(s) 3] “Observed Data”, “Lower Bound”, “Upper Bound” and “Confidence Interval” [par(s) 67-86] “The tree map visual 204 can be generated using machine learning-based agents 226 (e.g., hundreds of agents, thousands of agents, etc.) that constitute a multi-agent voting system 228. In one embodiment, a machine learning-based agent 226 can include a polynomial regression model coupled with a Quantile Loss Gradient Boosted Trees (QLGBT) machine learning model. The monitoring system 102 is configured to analyze the metrics being monitored and determine which metrics are highly correlated to distribute amongst the agents 226. The monitoring system 102 may ultimately determine hundreds or thousands of correlations between the metrics being monitored. For instance, independent regressors can be used to determine that a first metric and a second metric satisfy a correlation threshold, and therefore are highly correlated. If a first metric and a second metric are highly correlated, the assigned agent can confidently predict a data value for the second metric from the first metric.” [par(s) 87-91] “Using the regression error as a target along with temporal and exogenous variables, two gradient boosted tree (GBT) models are fitted around the prediction error, an example of which is referenced by 304. One gradient boosted tree (GBT) model can correspond to an upper bound and another gradient boosted tree (GBT) model can correspond to a lower bound. Using the gradient of a quantile loss/objective function in the boosting process, intervals on the predicted error are obtained. The end result is a pipeline that provides thresholds on the error, which is calculated from the predicted data value for the metric compared to the actual data value for the metric, as referenced by 306”;)
The combination of MITRA, Zhang is combinable with POOLE for the same rationale as set forth above with respect to claim 4.
Regarding claim 7
The combination of MITRA, Zhang teaches claim 1.
MITRA teaches
predicting, by the one or more processors, a possible contention case using [a time-series analysis of] the second impact values and first contention instances and/or second contention instances.
(MITRA [fig(s) 5-6] [par(s) 5] “The first machine learning model can be trained using the labeled training data to identify an anomalous node with performance degradation resulting from interference from a noisy neighbor. Training data for a second machine-learning model can be generated by determining the Euclidean distance between a feature vector reflecting the collected application-level and/or framework-level metrics and the centroids of corresponding clusters, and generating labels according to the corresponding resource contention that was simulated. The resulting training data can be used to train the second machine-learning model to identify the particular resource being contended.” [par(s) 23-25] “the output of the anomaly detector and/or resource classifier can indicate if an application is suffering from interference, and if so, which resource is being contended. If the anomaly detector and/or resource classifier indicates resource contention, a corresponding signal can be generated to trigger a designated mitigation response to reduce pressure on a contended resource and achieve low latency in the presence of interference.” [par(s) 41-48] “Generally, the models described herein ( e.g., anomaly detector 110 and/or resource classifier 112) can be used to detect and classify interference issues associated with operating a particular application in a distributed computing environment. … a tenant or customer thereof can automatically detect, classify, and mitigate performance degradation resulting from a noisy neighbor. These techniques can enable tenants to detect impacted applications, identify a contended resource, and take steps to mitigate the contention. Various application-level and/or framework-level metrics can be used to build machine-learning models to automatically detect and classify contentions.”; e.g., “identify the particular resource being contended” along with “the output of the anomaly detector and/or resource classifier can indicate if an application is suffering from interference”, “detect and classify interference issues” and “detect impacted applications, identify a contended resource” read(s) on “second impact values” and “second contention instances”.
Examiner notes that paragraph 49 of the Instant Specification describes “Another embodiment of the present invention comprises predicting (e.g., by use of the workload management tool), one or more potential future contention cases using a timeseries analysis of the second impact values, (i.e., second SHAP values), and data about first contention instances and/or data about second contention instances. The use of the time-series analysis techniques may address future bottlenecks of the computing system and include a feedback loop closure for system optimization”)
However, the combination of MITRA, Zhang does not appear to explicitly teach:
predicting, by the one or more processors, a possible contention case using [a time-series analysis of] the second impact values and first contention instances and/or second contention instances.
POOLE teaches
predicting, by the one or more processors, a possible contention case using a time-series analysis of the second impact values and first contention instances and/or second contention instances.
(POOLE [fig(s) 2] “Machine Learning-Based Agents 226” and “Multi-Agent Voting System 228” [fig(s) 3] “Observed Data”, “Lower Bound”, “Upper Bound” and “Confidence Interval” [fig(s) 4] [par(s) 18-22] “A user can notice when the object starts to move from a darkly-shaded blue region near the center of the radar-based visual, which can be a strong signal of normal activity for the combination of metrics represented by the object, towards a darkly-shaded red region near the periphery of the radar-based visual, which can be a strong signal of anomalous activity for the combination of metrics represented by the object. Another signal of an anomaly can include an increase in a size of the object (e.g., the size of a dot). A size of the object represents a degree to which the real-time data for the combination of metrics is anomalous to the observed historic data” [par(s) 23-27] “Each section in the tree map visual can be associated with a specific attribute used to compose one or more of the metrics being monitored ( e.g., the "checkout" metrics). A size and/or a color of an individual section can be used to indicate anomalous activity for the specific attribute.” [par(s) 28-37] “an agent is a regression model coupled with a Quantile Loss Gradient Boosted Trees (QLGBT) machine learning model for vote-decision making. … The agent then uses its own prediction error and/or other exogenous factors, such as temporal factors, holiday factors, etc., to generate upper and lower quantile limits, or bounds, on the error using QLGBT. If the error in a predicted data value falls outside a confidence interval ( e.g., the upper and lower bounds) when compared to the actual data value, then the agent provides a vote that signals an anomaly. … Once the votes are received from all the agents associated with a large correlated set of metrics being monitored (e.g., hundreds, thousands, etc.), the system can analyze the agents determined to be associated with the voted metric, and localize a problem to a specific attribute.” [par(s) 67-86] “The tree map visual 204 can be generated using machine learning-based agents 226 (e.g., hundreds of agents, thousands of agents, etc.) that constitute a multi-agent voting system 228. In one embodiment, a machine learning-based agent 226 can include a polynomial regression model coupled with a Quantile Loss Gradient Boosted Trees (QLGBT) machine learning model.”;)
The combination of MITRA, Zhang is combinable with POOLE for the same rationale as set forth above with respect to claim 4.
Regarding claim 14
The claim is a system claim corresponding to the method claim 4, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejections of the method claim.
Regarding claim 15
The claim is a system claim corresponding to the method claim 5, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejections of the method claim.
Regarding claim 16
The claim is a system claim corresponding to the method claim 6, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejections of the method claim.
Regarding claim 17
The claim is a system claim corresponding to the method claim 7, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejections of the method claim.
Claim(s) 8-10, 18-19 is/are rejected under 35 U.S.C. 103 as being unpatentable over MITRA et al. (US 20190354388 A1) Zhang et al. (Cascaded Random Forest for Hyperspectral Image Classification) in view of Martin et al. (US 10254970 B1)
Regarding claim 8
The combination of MITRA, Zhang teaches claim 1.
wherein the separating contention-related data and non-contention related data further comprises: (See claim 1)
However, the combination of MITRA, Zhang does not appear to explicitly teach:
determining, by the one or more processors, that two or more workload groups perform concurrently worse than a predefined number of standard deviations from an average performance of the two or more workload groups within a predefined period of time.
Martin teaches
determining, by the one or more processors, that two or more workload groups perform concurrently worse than a predefined number of standard deviations from an average performance of the two or more workload groups within a predefined period of time.
(Martin [fig(s) 1] [fig(s) 14] [fig(s) 17] [fig(s) 43] [col 27 ln 5– col 27 ln 52] “Physical devices (PDs) comprising the storage tiers may be included in logical groupings referred to as pools or storage pools (SPs). The storage tiers, and also the SPs, may be classified based on criteria including performance characteristics such as expected average response time (RT) for completing an I/O operation.” [col 33 ln 54– col 34 ln 38] “Referring to FIG. 14, shown is an example illustrating the number of observed I/Os having an observed RT meeting the RT objective for the different SPs in an embodiment in accordance with techniques herein. The example 1550 may also be referred to herein as the SP or pool CDF (cumulative distribution function) denoting the percentage of observed I/Os in each pool meeting SP-specific RT objectives. The example 1550 denotes the different SP RT objectives in 1552 where SPs A, B, C, D and E, respectively, have RT objectives of 2 ms, 4 ms, 6 ms, 8 ms, and 14 ms. Also illustrated, the SPs A, B, C, D, and E, respectively, have percentages of 70 (1554), 90 (1556), 99 (1558), 90 (1560) and 55 (1562) where each of the percentages denote a percentage of all I/Os directed to that SP that have an observed RT less than the SP-specific RT objective denoted in 1552 (e.g., at or below the expected average RT denoted by the SP's RT objective).” [col 58 ln 23– col 59 ln 62] “An SP may be determined as stable, for example, if the standard deviation for the SP with respect to the average RT for the SP is determined to be within an acceptable range. Furthermore, standard deviation may be determined for each individual RT bin for a given SP to determine whether particular observed RT ranges have acceptable standard deviation with respect to the average RT of the bin. The standard deviation of an RT bin may be used in connection with determining performance goals such as the target performance range for an SP whose performance characterization and capabilities are unknown other than through observing performance results.” [col 82 ln 1– col 83 ln 3] “processors”; e.g., fig 14 along with “standard deviation for the SP with respect to the average RT for the SP” read(s) on “determining, by the one or more processors, that two or more workload groups perform concurrently worse than a predefined number of standard deviations from an average performance of the two or more workload groups within a predefined period of time”.)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the system of MITRA, Zhang with the workload performances of Martin.
One of ordinary skill in the art would have been motived to combine in order to provide more effective, efficient and better optimization of resource usage.
(Martin [col 27 ln 5– col 27 ln 52] “However, even in such cases, actual expected performance may still vary, for example, depending on the particular vendor and possibly other factors. In such cases, classifying storage tiers based on expected average RT may prove more effective, efficient and for better optimization of resource usage.”)
Regarding claim 9
The combination of MITRA, Zhang, Martin teaches claim 8.
Martin further teaches
wherein the predefined period of time is at a minimum one minute.
(Martin [col 10 ln 20– col 12 ln 2] “It should be noted that the techniques herein may be used in connection with flash devices comprising what may be characterized as enterprise-grade or enterprise-class flash drives (EFDs) with an expected lifetime (e.g., as measured in an amount of actual elapsed time such as a number of years, months, and/or days) based on a number of guaranteed write cycles, or program cycles, and a rate or frequency at which the writes are performed.” [col 30 ln 1– col 32 ln 26] “As described in more detail below, an embodiment in accordance with techniques herein may maintain a histogram of RTs for each SP. The histogram may be converted into a cumulative distribution function (CDF). The slope of the CDF may indicate how much variation there is in the population and correspond to a wider QOS Target Range %. During each sample period, such as at each 10 minute or other time interval, information may be collected for each SP for an RT histogram instance, along with other data, such as total number of reads, total number of writes, total amount of data written and/or read, based on accumulated data for the time period. The accumulated information for each SP over multiple sampling periods may be included in a QOS matrix described in more detail herein. The QOS matrix may be used, for example, to model expected average RT performance for a proposed data movement, to map or convert between RT ranges and corresponding workloads (e.g., IOPS (I/Os per second) or number of I/Os), and the like. In a similar manner as described for SPs, data may be collected and maintained for each SG and used in connection with modeling data movements and assessing impact of such proposed data movements to determine whether SG performance goals are met, or below specified SG performance levels, or above specified SG performance goals” [col 37 ln 1– col 38 ln 7] “The SP state machine may timeout automatically, for example, after 30 minutes or some other suitable time period in order to ensure that the current budget and associated SP state do not become outdated due to real-time changes in the system. ”;)
The combination of MITRA, Zhang, Martin is combinable with Martin for the same rationale as set forth above with respect to claim 8.
Regarding claim 10
The combination of MITRA, Zhang, Martin teaches claim 8.
wherein the determining that two or more workload groups perform concurrently worse than a predefined number of standard deviations from their average performance within the predefined period of time, further comprises: (See claim 8)
MITRA further teaches
referring, by the one or more processors, to a normalized performance index metric for each of the different workload groups as part of the second part of the contention-related data.
(MITRA [fig(s) 5-6] [par(s) 4] “In order to generate training data for the models, interference is simulated and application and/or framework-level metrics are collected from the nodes of the tenant (including the nodes on which the identified co-located emulators reside). The collected metrics can be normalized across all associated emulators (e.g., across all virtual machines performing a particular distributed task), for example, with respect to a representative characteristic of the distributed task. Nodes without impacted metrics are labeled healthy, and nodes with impacted metrics are labeled anomalies.”;)
Regarding claim 18
The claim is a system claim corresponding to the method claim 8, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejections of the method claim.
Regarding claim 19
The claim is a system claim corresponding to the method claim 9, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejections of the method claim.
Prior Art
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
SOULY et al. (US 20230051237 A1) teaches a separate module that generates crossmodal attention data and provides the crossmodal attention data to the transformer network.
Wang et al. (Static and Moving Object Detection Using Flux Tensor with Split Gaussian Models) teaches Split Gaussian Models for images.
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 nonprovisional extension fee (37 CFR 1.17(a)) 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 SEHWAN KIM whose telephone number is (571)270-7409. The examiner can normally be reached Mon - Thu 7:00 AM - 5:00 PM.
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, Michael J Huntley can be reached on (303) 297-4307. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/S.K./Examiner, Art Unit 2129
3/27/2026
/MICHAEL J HUNTLEY/Supervisory Patent Examiner, Art Unit 2129