DETAILED ACTION
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 .
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 1–19 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Regarding independent claims 1, 8, and 14
Step 1 — whether the claim falls within any statutory category. See MPEP 2106.03
Claim 1 is drawn to a method claim; claim 8 is drawn to a system claim reciting a processor and memory; claim 14 is drawn to a computer program product claim reciting a non-transitory computer-readable medium. Therefore, each of these claims falls under one of the four categories of statutory subject matter, namely a process, a machine, and a manufacture, respectively. (Step 1: YES).
Step 2A Prong 1 — whether the claim recites a judicial exception. See MPEP 2106.04, subsection II
Regarding independent claim 1, the claim is directed to a method comprising monitoring statistics associated with a file-transfer workflow, detecting sustained low resource utilization, forecasting predicted values, checking the predicted values against trigger conditions, and triggering the workflow.
The limitation of "monitoring a plurality of statistics corresponding to a plurality of parameters associated with a workflow to transfer files from one of a deduplicated storage appliance or a cloud storage to another of the deduplicated storage appliance or the cloud storage" recites an abstract idea because it describes observing and collecting values of parameters. This limitation falls within the mental processes grouping of abstract ideas, i.e., concepts performed in the human mind including observation, evaluation, judgment, and opinion. See MPEP § 2106.04(a)(2), subsection III.
The limitation of "detecting that resource utilization on the deduplicated storage appliance has remained below a threshold resource utilization for a period of time" recites an abstract idea because it describes comparing an observed value against a threshold and forming a judgment as to whether the condition has persisted. A human being can observe utilization figures in a log and judge whether they have remained below a threshold. This limitation falls within the mental processes grouping of abstract ideas. See MPEP § 2106.04(a)(2), subsection III.
The limitation of "upon the detecting, forecasting predicted values corresponding to the plurality of parameters over a duration of time that the workflow is expected to run" recites an abstract idea because it describes calculating future values from past values. Under the broadest reasonable interpretation, the claim does not recite any particular forecasting technique or model, and the plain meaning of "forecasting" encompasses both a mathematical calculation and an evaluation or judgment practically performed in the human mind. This limitation falls within the mathematical concepts grouping and the mental processes grouping of abstract ideas. See MPEP §§ 2106.04(a)(2), subsections I and III.
The limitation of "checking the predicted values against a plurality of trigger conditions" recites an abstract idea because it describes comparing values against conditions and evaluating whether the conditions are met. This limitation falls within the mental processes grouping of abstract ideas. See MPEP § 2106.04(a)(2), subsection III.
Accordingly, independent claim 1 recites an abstract idea under Step 2A Prong One. Therefore, the analysis proceeds to Step 2A Prong Two.
Independent claim 8 is a system claim reciting limitations similar to claim 1 and is directed towards the abstract idea for similar reasons.
Independent claim 14 is a computer program product claim reciting limitations similar to claim 1 and is directed towards the abstract idea for similar reasons.
Step 2A Prong 2 — whether the claim as a whole integrates the recited judicial exception into a practical application of the exception or whether the claim is "directed to" the judicial exception. This evaluation is performed by (1) identifying whether there are any additional elements recited in the claim beyond the judicial exception, and (2) evaluating those additional elements individually and in combination to determine whether the claim as a whole integrates the exception into a practical application. See MPEP 2106.04(d)
Regarding independent claim 1, this claim recites additional elements of:
"a deduplicated storage appliance"
"a cloud storage"
"upon the predicted values satisfying at least some of the plurality of trigger conditions, triggering the workflow"
The recited deduplicated storage appliance and cloud storage are recited at a high level of generality and merely serve as the environment in which the abstract idea is performed. Limiting the abstract idea of monitoring, forecasting, and comparing to a deduplicated storage appliance and cloud storage amounts to no more than generally linking the use of the judicial exception to a particular technological environment or field of use. See MPEP § 2106.05(h).
The limitation "triggering the workflow" amounts to no more than instructions to apply the judicial exception by invoking a conventional operation. The claim does not recite how the workflow is triggered or how the transfer is carried out, and the specification acknowledges that cloud tiering is a pre-existing operation that "is generally scheduled by the end-user for periodic automatic invocation" (Specification, ¶ [0031]). This limitation therefore constitutes mere instructions to apply the exception. See MPEP § 2106.05(f).
The claim does not recite a particular improvement to the functioning of the storage appliance, the file system, or any other technology or technical field. See MPEP § 2106.05(a). The claim also does not recite a particular machine that is integral to the claimed process, a transformation of a particular article to a different state or thing, or any other meaningful limitation that applies the exception in a manner beyond merely implementing the abstract idea in a computer environment. See MPEP §§ 2106.05(b), 2106.05(c), and 2106.05(e).
Accordingly, the additional elements, individually and in combination, do not integrate the recited judicial exception into a practical application. Therefore, independent claim 1 is directed to the abstract idea under Step 2A, Prong Two. (Step 2A: YES).
Regarding independent claim 8, this claim is drawn to a system claim reciting similar limitations to claim 1 and is rejected under the same rationale. Claim 8 also recites additional elements of "a processor" and "memory configured to store one or more sequences of instructions which, when executed by the processor, cause the processor to carry out the steps of." These limitations recite a generic computer and generic computer components that merely act as a tool on which the method operates, and thus fail to integrate the exception into a practical application. See MPEP § 2106.05(f).
Regarding independent claim 14, this claim is drawn to a computer program product claim reciting similar limitations to claim 1 and is rejected under the same rationale. Claim 14 also recites additional elements of "a non-transitory computer-readable medium having a computer-readable program code embodied therein" and "the computer-readable program code adapted to be executed by one or more processors." These limitations recite a generic computer-readable medium and generic processors that merely act as a tool on which the method operates, and thus fail to integrate the exception into a practical application. See MPEP § 2106.05(f).
Step 2B — whether the claim amounts to significantly more than the judicial exception. See MPEP § 2106.05
Regarding independent claim 1, the claim recites the additional elements of a deduplicated storage appliance, a cloud storage, and triggering the workflow.
These additional elements, individually and in combination, do not amount to significantly more than the judicial exception. The recited deduplicated storage appliance and cloud storage are generic components performing generic functions of storing and transferring data, and such implementation is well-understood, routine, and conventional when recited at this level of generality. See MPEP §§ 2106.05(d), 2106.07(a)(III). The specification itself describes the deduplicated file system and the cloud tiering workflow as pre-existing (Specification, ¶¶ [0023], [0027], [0031]).
The step of triggering the workflow amounts to invoking a conventional data-movement operation and does not provide an inventive concept. See MPEP § 2106.05(f). Limiting the abstract idea to a deduplicated storage appliance and cloud storage merely links the abstract idea to a particular technological environment or field of use. See MPEP § 2106.05(h).
Accordingly, claim 1 does not include additional elements that amount to significantly more than the judicial exception. (Step 2B: NO). Claim 1 is not eligible.
Independent claims 8 and 14 recite the additional elements of a processor, memory, and a non-transitory computer-readable medium. These are generic computer components performing generic computer functions of storing instructions, executing instructions, processing data, and producing an output, and do not amount to significantly more than the judicial exception for the reasons set forth with respect to claim 1. See MPEP §§ 2106.05(d), 2106.05(f).
Regarding dependent claims 2–7, 9–13, and 15–19
Claims 2–7, 9–13, and 15–19 merely narrow the previously cited abstract idea limitations. For the reasons described above with respect to independent claims 1, 8, and 14, these judicial exceptions are not meaningfully integrated into a practical application, nor do they amount to significantly more than the abstract ideas.
Step 2A Prong 1
Regarding claims 2, 9, and 15, these claims recite the limitations of "training a statistically-based seasonal-autoregressive integrated moving average (SARIMA) model on a portion of the monitored statistics to predict the values of the plurality of parameters," "testing the predicted values on another portion of the monitored statistics," "measuring an accuracy of the model," "determining that the accuracy of the model is below a threshold accuracy," "upon the determination, recalibrating orders of the model to retrain the model, forecasting new predicted values, and retesting the new predicted values using the retrained model," and "repeating the recalibrating, forecasting, and retesting until at least one of the threshold accuracy is reached or a threshold number of times the model has been retrained is reached."
These limitations are directed towards the abstract idea of a mathematical concept, specifically the use of a mathematical model and mathematical calculations to compute predicted values and to compute and compare an accuracy metric. See MPEP § 2106.04(a)(2), subsection I. The limitations of measuring accuracy, determining that the accuracy is below a threshold, and repeating until a threshold is reached additionally fall within the mental processes grouping, as they describe evaluation and judgment. See MPEP § 2106.04(a)(2), subsection III.
Regarding claims 3, 10, and 16, these claims recite the limitations of "monitoring the workflow" and "suspending the workflow when one or more of a duration of the workflow exceeds the duration of time that the workflow is expected to run occurs, the resource utilization on the deduplicated storage appliance exceeds the threshold resource utilization, or the workflow overlaps with another job on the deduplicated storage appliance having priority over the workflow."
The monitoring and the evaluation of the recited conditions are directed towards the abstract idea of a mental process, i.e., observation, evaluation, and judgment as to whether a stated condition has occurred. See MPEP § 2106.04(a)(2), subsection III.
Regarding claims 4, 11, and 17, these claims recite the limitations of "maintaining a schedule specifying when the migration of the files should occur" and "triggering the migration of the files according to the schedule and regardless of whether the resource utilization on the deduplicated storage appliance has remained below the threshold resource utilization for the period of time."
Maintaining a schedule is directed towards the abstract idea of a mental process, as a human being can maintain a schedule specifying when an operation should occur, including with the aid of pen and paper. See MPEP § 2106.04(a)(2), subsection III.
Regarding claims 5, 12, and 18, these claims recite the limitation of "triggering another migration of the files when all of the predicted values satisfy the plurality of trigger conditions, and an amount of time elapsed since a last migration of the files is greater than a threshold time."
The evaluation of whether the predicted values satisfy the conditions and whether the elapsed time exceeds a threshold is directed towards the abstract idea of a mental process, i.e., comparison, evaluation, and judgment. See MPEP § 2106.04(a)(2), subsection III.
Regarding claims 6, 13, and 19, these claims recite the limitation of "triggering a second migration when capacity of the deduplicated storage appliance is predicted to cross a threshold capacity, wherein the second migration is triggered regardless of whether all of the predicted values satisfy the plurality of trigger conditions."
The prediction of capacity and the comparison of the predicted capacity against a threshold are directed towards the abstract idea of a mathematical concept and a mental process. See MPEP § 2106.04(a)(2), subsections I and III.
Regarding claim 7, this claim recites the limitation of "triggering a second migration when a total size of the files to be migrated reaches a threshold size, wherein the second migration is triggered regardless of whether all of the predicted values satisfy the plurality of trigger conditions."
Determining a total size and comparing it against a threshold is directed towards the abstract idea of a mental process, i.e., evaluation and judgment, and of a mathematical concept, i.e., calculating a sum. See MPEP § 2106.04(a)(2), subsections I and III.
Step 2A Prong 2 and Step 2B
Regarding claims 2–7, 9–13, and 15–19, these claims recite no additional elements beyond those recited in independent claims 1, 8, and 14, namely the deduplicated storage appliance, the cloud storage, the processor, the memory, and the non-transitory computer-readable medium.
These limitations amount to no more than generally linking the use of the judicial exception to a particular technological environment or field of use, and to mere instructions to apply the exception using generic computer components. See MPEP §§ 2106.05(f) and 2106.05(h). The recited SARIMA model of claims 2, 9, and 15 is used only as a tool to compute predicted values, without reciting any particular improvement to the model itself or to machine learning technology. See MPEP § 2106.05(f).
Accordingly, claims 2–7, 9–13, and 15–19 do not integrate the recited judicial exception into a practical application and do not amount to significantly more than the judicial exception, and are rejected under 35 U.S.C. § 101.
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 (i.e., changing from AIA to pre-AIA ) 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.
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.
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1, 3, 4, 5, 8, 10, 11, 12, 14, 16, 17, and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Lakshmikantha et al. (Lakshmikantha), US 11,514,317 B2, in view of Eda et al. (Eda), US 2019/0155534 A1, and further in view of Kottomtharayil (Kottomtharayil), US 8,769,048 B2.
As to independent Claim 1, Lakshmikantha teaches:
"A method comprising:"
"monitoring a plurality of statistics corresponding to a plurality of parameters associated with a workflow"
"detecting that resource utilization on the deduplicated storage appliance has remained below a threshold resource utilization for a period of time;"
"upon the detecting, forecasting predicted values corresponding to the plurality of parameters"
"checking the predicted values against a plurality of trigger conditions; and"
"upon the predicted values satisfying at least some of the plurality of trigger conditions, triggering the workflow."
Regarding "A method comprising," Lakshmikantha discloses a method of efficiently managing a plurality of resources of a protection storage system, comprising registering requests from file system services, collecting resource utilization data, training a machine learning model, scheduling execution based on predicted utilization patterns, and conducting monitoring (Lakshmikantha, col. 18, lines 36–52).
Regarding "monitoring a plurality of statistics corresponding to a plurality of parameters associated with a workflow," Lakshmikantha discloses that "the resource utilization collection unit is responsible for collecting utilization data of the storage system resources… Resource utilization may include a measurement of disk activity, the utilization of disk, disk space, memory, CPU, network, or other resource over a period of time" (Lakshmikantha, col. 6, lines 53–60). The plurality of parameters is further enumerated as "% disk read time…, % disk time…, % disk write time…, % idle time…, current disk queue length…, disk reads/sec…, disk writes/sec…, split IO/sec" (Lakshmikantha, col. 6, line 65 – col. 7, line 10), together with CPU utilization and disk space (Lakshmikantha, col. 7, lines 11–21). The statistics are collected as performance logs maintained by the resources of the storage system (Lakshmikantha, col. 6, lines 63–65). The workflow is a file system service executed on the deduplication file system, such as file verification 157, garbage collection 160, or opportunistic caching 163 (Lakshmikantha, col. 5, lines 46–50).
Regarding "detecting that resource utilization on the deduplicated storage appliance has remained below a threshold resource utilization for a period of time," Lakshmikantha discloses a data protection storage system having a deduplication file system in which "the deduplication engine is responsible for deduplicating data entering the file system" (Lakshmikantha, col. 5, lines 18–20). Lakshmikantha further discloses that the predictions "include a set of availability windows 190. An availability window may include a timetable specifying particular days including starting and ending times, e.g., time slot, when utilization of a particular resource (e.g., disk) is predicted to be available (e.g., low)" (Lakshmikantha, col. 7, lines 29–35), and that "these windows indicate when resource utilization is low and for how long" (Lakshmikantha, col. 16, lines 53–58). The sustained sub-threshold condition over a period of time is disclosed as "a predicted utilization pattern includes a timetable showing low disk utilization on Sunday from 2:00 AM to 4:00 AM" and "a low disk busy percent of 9% on Sunday at 2:00 AM" (Lakshmikantha, col. 10, lines 22–36). Lakshmikantha additionally discloses sampling the actual resource utilization and comparing it against a threshold (Lakshmikantha, col. 10, lines 15–21).
Regarding "upon the detecting, forecasting predicted values corresponding to the plurality of parameters," Lakshmikantha discloses that "the machine learning model is built and trained according to machine learning algorithms using utilization data collected by the resource utilization collection unit… The output of the learning algorithm is a machine learning model that can make predictions" (Lakshmikantha, col. 7, lines 22–28), and that the model may "be trained with multi-features such as disk, CPU, memory utilization, or combinations of these" (Lakshmikantha, col. 17, lines 19–22).
Regarding "checking the predicted values against a plurality of trigger conditions," Lakshmikantha discloses that "the scheduler executes an algorithm that attempts to match a resource requirement of a particular file system service to an availability window associated with a corresponding resource and indicating low utilization of the corresponding resource during a time period specified by the availability window" (Lakshmikantha, col. 8, lines 36–41). The plurality of trigger conditions is disclosed as a first availability window specifying a first time slot where utilization of a first resource is to be low and a second availability window specifying a second time slot where utilization of a second resource is to be low, each matched against the corresponding registered requirement (Lakshmikantha, col. 8, lines 42–54), and as further factors including "a priority level of a file system service, current resource usage levels, amount of time elapsed since a file system service was last executed" (Lakshmikantha, col. 8, lines 55–60), and a threshold of "96 percent of disk space used" (Lakshmikantha, col. 9, lines 1–6).
Regarding "upon the predicted values satisfying at least some of the plurality of trigger conditions, triggering the workflow," Lakshmikantha discloses that "the scheduler receives the file system service registration information from the registration unit and the predicted availability windows from the machine learning model. The scheduler analyzes the received information to output a file system services execution schedule 196" (Lakshmikantha, col. 8, lines 18–22), that "the scheduler may then schedule the first file system service for execution during the first time slot, and the second file system service for execution during the second time slot" (Lakshmikantha, col. 8, lines 50–54), and that "the services scheduler can call the registered service during the next available window based on a priority scheme" (Lakshmikantha, col. 17, lines 7–11).
Lakshmikantha teaches something related to the claimed workflow, in that the scheduled workflow is a data-movement or batch-processing operation executed on a deduplication file system, including migration and replication processes (Lakshmikantha, col. 5, lines 54–56). However, Lakshmikantha does not teach "to transfer files from one of a deduplicated storage appliance or a cloud storage to another of the deduplicated storage appliance or the cloud storage."
In the same field of endeavor, Eda teaches "to transfer files from one of a deduplicated storage appliance or a cloud storage to another of the deduplicated storage appliance or the cloud storage." Eda discloses "current day deduplication-enabled and 'cloud-as-backup' tier enabled storage environments" (Eda, page 1, ¶ [0002]), and defines the workflow as follows: "'Cloud-as-tier' (or cloud-as-backup tier) generally is a feature that provides a 'cloud as an extended storage tier' option to an on-premises filesystem (or storage system/device)… that works by using an Information Life Cycle Management (ILM) migrate policy to determine appropriate candidate data for migration to the cloud… Each file migrated to the cloud leaves a stub on the local filesystem, is stored as two separate objects (i.e., one for data and one for metadata), and creates a container for itself" (Eda, page 1, ¶ [0016]). Eda further discloses the reverse transfer direction, in that "when a read/write event is performed on the stubs… these events are captured, and a recall script is triggered which reconstructs the file based on data and metadata objects stored in the cloud" (Eda, page 1, ¶ [0016]). Eda characterizes the environment as "a local filesystem and a cloud-as-tier (e.g., to/from cloud-as-tier, backup/restore files stored on an on-premises filesystem to/from public/private cloud storage), in a deduplication-enabled storage environment" (Eda, page 2, ¶ [0023]).
Lakshmikantha and Eda are analogous to the claimed invention as both are from the same field of endeavor of managing data-movement operations in deduplication-enabled storage systems. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the resource-utilization monitoring and prediction-based scheduling of Lakshmikantha with the cloud-as-tier file migration workflow of Eda. The motivation to combine Lakshmikantha and Eda is that Lakshmikantha teaches that batch-processing operations performed on a deduplication file system "consume[] resources including disk, CPU, memory, and other computer resources" and that scheduling such operations without regard to resource availability may cause a backup job to "not be completed within the specified backup window" (Lakshmikantha, col. 6, lines 22–35), so that one of ordinary skill would apply the same scheduling technique to the cloud migration workflow of Eda to avoid contention with client backup workloads.
The combination of Lakshmikantha and Eda, however, does not teach "over a duration of time that the workflow is expected to run."
In the same field of endeavor, Kottomtharayil teaches "over a duration of time that the workflow is expected to run." Kottomtharayil discloses that "the system looks at various criteria to determine an (optimal or near optimal) period for performing a storage operation. The criteria can include one or more of the following: job priority, types of data within the job, network traffic or load, a disk load, CPU usage, expected backup duration, and so on" (Kottomtharayil, col. 7, lines 13–19). Kottomtharayil further discloses matching the forecast horizon to the expected run length of the job: "The system may review historical data or information about previously performed jobs or usage of storage operation resources, and predict when to start a job… The data may identify a time when the usage resources are predicted to be at a 'low' usage level (e.g., below a threshold level of a load on the resources), and start the job at that time. For example, the system may determine that the usage level for certain resources drops during a typical lunch hour (e.g., 12:00 to 1:00 P.M.), and select a job to complete during that time that is expected to be completed within an hour" (Kottomtharayil, col. 8, lines 12–24).
Lakshmikantha, Eda, and Kottomtharayil are analogous to the claimed invention as all are from the same field of endeavor of scheduling data protection and data-movement operations based on storage system resource utilization. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the prediction-based scheduling of Lakshmikantha and the cloud-as-tier migration workflow of Eda with the expected-duration criterion of Kottomtharayil. The motivation to combine Lakshmikantha, Eda, and Kottomtharayil is as recited by Kottomtharayil, which teaches selecting a job "expected to be completed within an hour" for a window in which resources are predicted to remain below a threshold load (Kottomtharayil, col. 8, lines 19–24), such that forecasting the parameters over the duration the workflow is expected to run ensures the workflow completes within the low-utilization window rather than extending into a period of high resource demand.
As to Claim 3, the limitations of Claim 1 from which Claim 3 depends are rejected under the same rationale set forth above with respect to Claim 1.
Regarding the additional limitations of Claim 3, Lakshmikantha teaches:
"monitoring the workflow; and"
"suspending the workflow when one or more of a duration of the workflow exceeds the duration of time that the workflow is expected to run occurs"
Regarding "monitoring the workflow; and," Lakshmikantha discloses that "the monitor is responsible for monitoring during the execution of the file system services according to the execution schedule. Parameters that the monitor may monitor may include a performance of the backup engine (e.g., backup throughput or speed at which the storage system has ingested backup data as measured in megabytes per second), actual resource utilization patterns, latency, other parameters, or combinations of these" (Lakshmikantha, col. 9, lines 7–14). Lakshmikantha further discloses that "data for the actual utilization patterns may be collected by periodically polling the resources during execution of the file system services" (Lakshmikantha, col. 9, lines 22–25).
Regarding "suspending the workflow when one or more of a duration of the workflow exceeds the duration of time that the workflow is expected to run occurs," Lakshmikantha discloses that "the file system services execution schedule may include a timetable specifying time slots when each file system service should be executed (and when the file system service should be stopped or suspended)" (Lakshmikantha, col. 8, lines 22–26). Lakshmikantha further discloses the suspension occurring where the workflow runs past its expected duration, in that "the file system service is scheduled in the 2:00 AM to 4:00 AM window. Note that the service will stop at 4:00 AM but may not have completed what it originally set out to" (Lakshmikantha, col. 13, lines 10–14), and that "a file system service that is not completed within its assigned time slot may be resumed at a later time… A bookmark may be stored indicating a progress of the file system service. The scheduler may then update the execution schedule and assign another time slot to the file system service" (Lakshmikantha, col. 8, lines 26–34).
Lakshmikantha teaches something related to suspending a workflow on the basis of a resource threshold and a job priority, in that the scheduler "may override the priority level of a file system service during scheduling… when a measurement of used disk space exceeds a threshold (e.g., 96 percent of disk space used)" (Lakshmikantha, col. 9, lines 1–6). However, Lakshmikantha does not teach:
"the resource utilization on the deduplicated storage appliance exceeds the threshold resource utilization"
"or the workflow overlaps with another job on the deduplicated storage appliance having priority over the workflow"
In the same field of endeavor, Kottomtharayil teaches "the resource utilization on the deduplicated storage appliance exceeds the threshold resource utilization." Kottomtharayil discloses that "once a job starts, the system may also monitor the usage of the resources and make decisions based on the monitored usage. For example, the system may start a job and after a certain time the job (and other jobs or operations) may cause the load on the resources to exceed a threshold value. Such an occurrence may cause the system to stop the job until the load drops below the threshold value (or, until the predicted load drops below the threshold value). At this time, the system may restart the job using the resources" (Kottomtharayil, col. 8, lines 25–33).
Kottomtharayil further teaches "or the workflow overlaps with another job on the deduplicated storage appliance having priority over the workflow." Kottomtharayil discloses the suspension arising from concurrent jobs, in that it is "the job (and other jobs or operations)" that causes the load to exceed the threshold value and thereby causes the system to stop the job (Kottomtharayil, col. 8, lines 27–31). Kottomtharayil further discloses the priority relationship among the overlapping jobs, in that "the priority system instructs a storage system to perform all hard coded jobs first (jobs that must be completed within a time window). The hard coded jobs may be certain daily backups or other data required to be copied within a time window. Once these jobs are completed, the storage system performs all soon to expire jobs second… and performs all other jobs based on the flexible and dynamically determined schedule of operations described herein" (Kottomtharayil, col. 8, lines 1–9), and that "the system may assign a higher priority to partially completed jobs, which may affect the threshold value" (Kottomtharayil, col. 8, lines 33–36).
Lakshmikantha and Kottomtharayil are analogous to the claimed invention as both are from the same field of endeavor of scheduling and controlling data protection operations based on storage system resource utilization. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the monitoring and suspension of the scheduled workflow of Lakshmikantha with the load-threshold and job-priority suspension criteria of Kottomtharayil. The motivation to combine Lakshmikantha and Kottomtharayil is that Lakshmikantha teaches that a scheduled service executing concurrently with unanticipated backup jobs degrades performance (Lakshmikantha, col. 10, lines 3–8), and Kottomtharayil provides the known technique of stopping the job when the load exceeds the threshold and restarting it once the load drops, thereby protecting the higher-priority production workload.
As to Claim 4, the limitations of Claim 1 from which Claim 4 depends are rejected under the same rationale set forth above with respect to Claim 1.
Regarding the additional limitations of Claim 4, Lakshmikantha teaches:
"maintaining a schedule specifying when the migration of the files should occur; and"
Regarding "maintaining a schedule specifying when the migration of the files should occur; and," Lakshmikantha discloses that "many system services are manually scheduled or run on periodic intervals via a fixed schedule" (Lakshmikantha, col. 1, lines 38–41), and that "rather than having system services manually scheduled or set to run on periodic intervals via a fixed schedule, ML techniques can be applied to provide more efficient use of resources" (Lakshmikantha, col. 17, lines 22–26). Lakshmikantha further discloses that the schedule is maintained by the system as "a file system services execution schedule 196," which "may include a timetable specifying time slots when each file system service should be executed" (Lakshmikantha, col. 8, lines 20–24), and that the system may "revert back to a manual scheduling of the file system services where, for a temporary collection period, the user is prompted to specify the execution schedule of the file system services" (Lakshmikantha, col. 11, lines 20–25).
Lakshmikantha teaches something related to the claimed migration, in that the deduplication file system performs "migration processes" among its data movement operations (Lakshmikantha, col. 5, lines 54–56). However, Lakshmikantha does not teach:
"wherein the workflow comprises migrating the files from the deduplicated storage appliance to the cloud storage"
"triggering the migration of the files according to the schedule and regardless of whether the resource utilization on the deduplicated storage appliance has remained below the threshold resource utilization for the period of time."
In the same field of endeavor, Eda teaches "wherein the workflow comprises migrating the files from the deduplicated storage appliance to the cloud storage." Eda discloses that "'Cloud-as-tier' (or cloud-as-backup tier) generally is a feature that provides a 'cloud as an extended storage tier' option to an on-premises filesystem (or storage system/device)… that works by using an Information Life Cycle Management (ILM) migrate policy to determine appropriate candidate data for migration to the cloud. File details migrated (or transferred) to the cloud are stored in a database (e.g., on the local filesystem)… Each file migrated to the cloud leaves a stub on the local filesystem" (Eda, page 1, ¶ [0016]), in a "deduplication-enabled storage environment" (Eda, page 2, ¶ [0023]).
Lakshmikantha and Eda are analogous to the claimed invention as both are from the same field of endeavor of managing data-movement operations in deduplication-enabled storage systems. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to apply the maintained execution schedule of Lakshmikantha to the cloud-as-tier file migration workflow of Eda. The motivation to combine Lakshmikantha and Eda is that Eda teaches that migration to the cloud tier is governed by a migrate policy determining candidate data (Eda, page 1, ¶ [0016]), so that maintaining a schedule specifying when that migration occurs, as taught by Lakshmikantha, provides a defined time at which the policy is executed.
The combination of Lakshmikantha and Eda, however, does not teach "triggering the migration of the files according to the schedule and regardless of whether the resource utilization on the deduplicated storage appliance has remained below the threshold resource utilization for the period of time."
In the same field of endeavor, Kottomtharayil teaches "triggering the migration of the files according to the schedule and regardless of whether the resource utilization on the deduplicated storage appliance has remained below the threshold resource utilization for the period of time." Kottomtharayil discloses a scheduled class of jobs that executes irrespective of the resource-utilization assessment, in that "the priority system instructs a storage system to perform all hard coded jobs first (jobs that must be completed within a time window). The hard coded jobs may be certain daily backups or other data required to be copied within a time window. Once these jobs are completed, the storage system performs all soon to expire jobs second (that is, any jobs having a deadline of completion less than a threshold deadline, or previously postponed jobs), and performs all other jobs based on the flexible and dynamically determined schedule of operations described herein" (Kottomtharayil, col. 8, lines 1–9). Kottomtharayil expressly distinguishes these scheduled jobs from the jobs whose timing depends on predicted low resource usage, the latter being started at "a time when the usage resources are predicted to be at a 'low' usage level (e.g., below a threshold level of a load on the resources)" (Kottomtharayil, col. 8, lines 15–19). Kottomtharayil further discloses that the schedule is defined by request criteria including "a desired frequency (e.g., once per hour) that the request should be performed" (Kottomtharayil, col. 7, lines 28–31), and that the flexible priority system "ensures, among other things, that a data storage system can complete all data storage operations when they are required to be completed, even when certain operations are not completed when originally scheduled" (Kottomtharayil, col. 8, lines 9–13).
Lakshmikantha and Kottomtharayil are analogous to the claimed invention as both are from the same field of endeavor of scheduling data protection operations based on storage system resource utilization. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the resource-utilization-based triggering of Lakshmikantha with the hard-coded scheduled job class of Kottomtharayil. The motivation to combine Lakshmikantha and Kottomtharayil is as recited by Kottomtharayil, which teaches that the scheduled class "ensures… that a data storage system can complete all data storage operations when they are required to be completed, even when certain operations are not completed when originally scheduled" (Kottomtharayil, col. 8, lines 9–13), such that the migration is guaranteed to occur notwithstanding that the predicted low-utilization condition is never satisfied.
As to Claim 5, the limitations of Claim 1 from which Claim 5 depends are rejected under the same rationale set forth above with respect to Claim 1. The limitations "wherein the workflow comprises migrating the files from the deduplicated storage appliance to the cloud storage and the method further comprises:", "maintaining a schedule specifying when the migration of the files should occur;" and "triggering the migration of the files according to the schedule; and" are rejected under the same rationale set forth above with respect to Claim 4.
Regarding the additional limitations of Claim 5, Lakshmikantha teaches:
"triggering another migration of the files when all of the predicted values satisfy the plurality of trigger conditions, and an amount of time elapsed since a last migration of the files is greater than a threshold time."
Regarding "triggering another migration of the files when all of the predicted values satisfy the plurality of trigger conditions," Lakshmikantha discloses that "the scheduler executes an algorithm that attempts to match a resource requirement of a particular file system service to an availability window associated with a corresponding resource and indicating low utilization of the corresponding resource during a time period specified by the availability window" (Lakshmikantha, col. 8, lines 36–41). Lakshmikantha discloses that the match is made across the plurality of conditions, in that "a first availability window associated with the first resource specifies a first time slot where utilization of the first resource is to be low as predicted by the machine learning model. A second availability window associated with the second resource specifies a second time slot where utilization of the second resource is to be low as predicted by the machine learning model. The scheduler may then schedule the first file system service for execution during the first time slot, and the second file system service for execution during the second time slot" (Lakshmikantha, col. 8, lines 45–54).
Regarding "and an amount of time elapsed since a last migration of the files is greater than a threshold time," Lakshmikantha discloses that "the scheduler may additionally weigh one or more other factors when generating the execution schedule of the file system services. Such factors may include a priority level of a file system service, current resource usage levels, amount of time elapsed since a file system service was last executed, other factors, or combinations of these" (Lakshmikantha, col. 8, lines 55–60).
Lakshmikantha teaches something related to the elapsed-time criterion in that the amount of time elapsed since the service was last executed is weighed in generating the schedule (Lakshmikantha, col. 8, lines 57–60). However, Lakshmikantha does not teach "an amount of time elapsed since a last migration of the files is greater than a threshold time."
In the same field of endeavor, Kottomtharayil teaches "an amount of time elapsed since a last migration of the files is greater than a threshold time." Kottomtharayil discloses that the storage operation request includes "a desired frequency (e.g., once per hour) that the request should be performed" (Kottomtharayil, col. 7, lines 28–31), and that the criteria applied in determining when to perform the operation include "frequency of use, particularly with respect to particular data, data sources, departments, etc." (Kottomtharayil, col. 7, lines 56–59). Kottomtharayil further discloses that the system evaluates a job against a deadline expressed as an elapsed-time threshold, in that the storage system "performs all soon to expire jobs second (that is, any jobs having a deadline of completion less than a threshold deadline, or previously postponed jobs)" (Kottomtharayil, col. 8, lines 5–8), and that the job criterion "710 may include a metric attributed to the time window in which to complete the job" (Kottomtharayil, col. 8, lines 55–58).
Lakshmikantha and Kottomtharayil are analogous to the claimed invention as both are from the same field of endeavor of scheduling data protection operations based on storage system resource utilization. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the elapsed-time scheduling factor of Lakshmikantha with the frequency and threshold-deadline criteria of Kottomtharayil. The motivation to combine Lakshmikantha and Kottomtharayil is that Kottomtharayil teaches specifying a desired frequency at which a storage operation should be performed (Kottomtharayil, col. 7, lines 28–31), such that comparing the elapsed time since the last execution against a threshold time enforces that frequency and avoids performing the operation more often than required.
Claim 8 recites limitations identical to those of Claim 1 and is rejected under the same rationale set forth above with respect to Claim 1.
Claim 10 recites limitations identical to those of Claim 3 and is rejected under the same rationale set forth above with respect to Claim 3.
Claim 11 recites limitations identical to those of Claim 4 and is rejected under the same rationale set forth above with respect to Claim 4.
Claim 12 recites limitations identical to those of Claim 5 and is rejected under the same rationale set forth above with respect to Claim 5.
Claim 14 recites limitations identical to those of Claims 1 and 8 and is rejected under the same rationale set forth above with respect to Claims 1 and 8.
Claim 16 recites limitations identical to those of Claims 3 and 10 and is rejected under the same rationale set forth above with respect to Claims 3 and 10.
Claim 17 recites limitations identical to those of Claims 4 and 11 and is rejected under the same rationale set forth above with respect to Claims 4 and 11.
Claim 18 recites limitations identical to those of Claims 5 and 12 and is rejected under the same rationale set forth above with respect to Claims 5 and 12.
Claims 2, 9, and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Lakshmikantha et al. (Lakshmikantha), US 11,514,317 B2, in view of Eda et al. (Eda), US 2019/0155534 A1, further in view of Kottomtharayil (Kottomtharayil), US 8,769,048 B2, and further in view of Nagpal et al. (Nagpal), US 2020/0034745 A1.
As to Claim 2, the limitations of Claim 1 from which Claim 2 depends are rejected under the same rationale set forth above with respect to Claim 1.
Regarding the additional limitations of Claim 2, Lakshmikantha teaches:
"testing the predicted values on another portion of the monitored statistics;"
"measuring an accuracy of the model;"
"determining that the accuracy of the model is below a threshold accuracy;"
Regarding "testing the predicted values on another portion of the monitored statistics," Lakshmikantha discloses that "a first portion of the collected resource utilization dataset is to be used for retraining the machine learning model and a second portion of the collected resource utilization dataset is to be used to validate the model… The accuracy of the predicted resource utilization pattern can be evaluated against the second portion of the dataset" (Lakshmikantha, col. 11, lines 40–50). Lakshmikantha further discloses "validate the model and its accuracy by using a portion (e.g., 75 percent) of the tuples for training and validate the model using the remaining portion (e.g., 25 percent) of tuples as a validation set" (Lakshmikantha, col. 14, Table A, step 7), and that "25% of the data was used to validate the selected model" (Lakshmikantha, col. 15, lines 58–62).
Regarding "measuring an accuracy of the model," Lakshmikantha discloses computing accuracy by root mean square error, defining "Root Square Error for one sample value=[(P−V)/V]^2" and "Root Mean Square Error for entire data set=Mean of all Root Square Error values," where "V=actual sample value" and "P=prediction of the same sample value" (Lakshmikantha, col. 16, Table B). Lakshmikantha further discloses that "accuracy of prediction for each of the models is computed" (Lakshmikantha, col. 15, lines 65–67), reporting a best accuracy of 78% for linear regression (Lakshmikantha, col. 16, lines 8–10), 91% for random forests (Lakshmikantha, col. 16, lines 11–15), and 95% for LSTM (Lakshmikantha, col. 16, lines 31–34).
Regarding "determining that the accuracy of the model is below a threshold accuracy," Lakshmikantha discloses that "if, through monitoring, accuracy falls below a certain threshold on the validation set, then it indicates that the pattern of usage has changed and hence retraining of the model may be required" (Lakshmikantha, col. 16, lines 46–49). Lakshmikantha further discloses that "the monitor computes a difference between the predicted utilization patterns and the actual utilization patterns. The difference is compared against a pre-determined threshold. If the difference (or a prediction error) is greater than the pre-determined threshold, then the machine learning model may be retrained" (Lakshmikantha, col. 9, lines 28–34), and that "if the accuracy or error in prediction does not meet a desired threshold, the temporary period may be extended" (Lakshmikantha, col. 11, lines 52–56).
Lakshmikantha teaches something related to training a model on a portion of the monitored statistics, in that "75% of the data was used to train a selected model" (Lakshmikantha, col. 15, lines 58–60), and something related to retraining and retesting, in that "the machine learning module may be retrained using data collected over a second period of time having a duration greater than the first period of time" (Lakshmikantha, col. 12, lines 6–10). However, Lakshmikantha does not teach:
"training a statistically-based seasonal-autoregressive integrated moving average (SARIMA) model on a portion of the monitored statistics to predict the values of the plurality of parameters;"
"upon the determination, recalibrating orders of the model to retrain the model, forecasting new predicted values, and retesting the new predicted values using the retrained model; and"
"repeating the recalibrating, forecasting, and retesting until at least one of the threshold accuracy is reached or a threshold number of times the model has been retrained is reached."
In the same field of endeavor, Nagpal teaches "training a statistically-based seasonal-autoregressive integrated moving average (SARIMA) model on a portion of the monitored statistics to predict the values of the plurality of parameters." Nagpal discloses "systems, methods, and computer program products for seasonal time series analysis and forecasting using a distributed tournament selection process" (Nagpal, page 2, ¶ [0031]), in which "computer models can be used within a computing framework for mathematically representing predicted future values based on previously measured values. Examples of such models include autoregressive models, integrated models, moving average models, etc. and/or various combinations of the foregoing… For instance, popular models include the autoregressive integrated moving average (ARIMA) models, exponential smooth (ETS) models, seasonal trend decomposition using loess (STL) models, neural networks, random walk models, seasonal naïve, mean and/or linear regression models" (Nagpal, page 1, ¶ [0005]). Nagpal further discloses training on a portion of the collected data, in that the training process includes "a determination of prediction model parameters based on a first past time window, an evaluation of those prediction models using a second past time window" (Nagpal, page 11, ¶ [0117]), the prediction being of parameter values in that "a time series generator 129 calculates a plurality of expected values for the parameter (e.g., predicted time series 133) using the selected one or more of the plurality of evaluated prediction models" (Nagpal, page 8, ¶ [0086]).
Nagpal further teaches "upon the determination, recalibrating orders of the model to retrain the model, forecasting new predicted values, and retesting the new predicted values using the retrained model." Nagpal discloses that "the prediction model parameters are determined using the first past time window. Any suitable approach can be used to remove noise/randomness, adjust for seasonality, and otherwise fit the model to the data" (Nagpal, page 11, ¶ [0120]), and that determining the correct model order includes differencing, in that "to choose the correct ARIMA model, a calculation module might check if the time series is stationary using tests for the stationary property, and if so, use the ARIMA model, otherwise the time series may have to be differenced" (Nagpal, page 1, ¶ [0006]). Nagpal further discloses that "individual models may be selected or removed from consideration based on various characteristics of the data. For instance, initially, or at any point during training, a model may be dropped from consideration if, for instance, an abrupt change in trend or seasonal data is identified according to known techniques such that a model is presumed to provide less then acceptable results" (Nagpal, page 11, ¶ [0120]), and that the models are evaluated "using parameter data captured over a second time series" against "applicability and/or accuracy criteria" (Nagpal, page 8, ¶ [0086]).
Nagpal further teaches "repeating the recalibrating, forecasting, and retesting until at least one of the threshold accuracy is reached or a threshold number of times the model has been retrained is reached." Nagpal discloses that "multiple iterations of the tournament may be performed. For instance, if the number of models or model combinations is greater than the number of nodes, then the tournament may need to execute in multiple rounds. In this case the process may collect the results for later analysis for operation in multiple rounds" (Nagpal, page 13, ¶ [0146]), the iterations being evaluated by accuracy in that "the prediction models for each individual node may have varying levels of accuracy. Therefore, the results of each individual node may need to be weighted so as to favor prediction models that better predicted the actual past time series data usage over lower performing models" (Nagpal, page 13, ¶ [0152]), and the accuracy being measured by "symmetric mean absolute percentage error (sMAPE), mean absolute scaled error (MASE), root mean square error (RMSE), random error (RE), absolute error (AE), etc." (Nagpal, page 13, ¶ [0145]). The termination conditions are further taught by Lakshmikantha, which discloses that the process "may be repeated until a threshold or an acceptable level of prediction accuracy is achieved" (Lakshmikantha, col. 11, lines 56–60), and that "retraining the machine learning model is not triggered until the predicted utilization patterns have failed to meet their desired accuracy target a threshold number of consecutive times… For example, the number misses may be configured to be 2, 3, 4, 5, or more than 5 times" (Lakshmikantha, col. 12, lines 14–25).
Lakshmikantha and Nagpal are analogous to the claimed invention as both are from the same field of endeavor of forecasting resource utilization by time series analysis. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to substitute the seasonal ARIMA model and the model-parameter determination and evaluation process of Nagpal for the machine learning model of Lakshmikantha. The motivation to combine Lakshmikantha and Nagpal is that Lakshmikantha identifies a recurring seasonal pattern in the monitored statistics, disclosing that "in certain 7-day usage cycles, backups take place on weekends. Thus, disk utilization tends to be on the higher side during weekends" (Lakshmikantha, col. 15, lines 53–57), and Nagpal teaches that its seasonal models permit "adjusting for seasonal workload variations and quality of service" (Nagpal, page 3, ¶ [0040]), such that one would obtain more accurate forecasts of the weekly-cyclical utilization recognized by Lakshmikantha.
Claim 9 recites limitations identical to those of Claim 2 and is rejected under the same rationale set forth above with respect to Claim 2.
Claim 15 recites limitations identical to those of Claims 2 and 9 and is rejected under the same rationale set forth above with respect to Claims 2 and 9.
Claims 6, 13, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Lakshmikantha et al. (Lakshmikantha), US 11,514,317 B2, in view of Eda et al. (Eda), US 2019/0155534 A1, further in view of Kottomtharayil (Kottomtharayil), US 8,769,048 B2, and further in view of Perneti et al. (Perneti), US 11,409,453 B2.
As to Claim 6, the limitations of Claim 1 from which Claim 6 depends are rejected under the same rationale set forth above with respect to Claim 1. The limitations "wherein the workflow comprises migrating the files from the deduplicated storage appliance to the cloud storage and the method further comprises:" and "maintaining a schedule specifying when the migration of the files should occur;" are rejected under the same rationale set forth above with respect to Claim 4.
Regarding the additional limitations of Claim 6, Lakshmikantha teaches:
"triggering a first migration according to the schedule; and"
Regarding "triggering a first migration according to the schedule; and," Lakshmikantha discloses that "the scheduler analyzes the received information to output a file system services execution schedule 196. The file system services execution schedule may include a timetable specifying time slots when each file system service should be executed" (Lakshmikantha, col. 8, lines 20–24), and that "the services scheduler can call the registered service during the next available window based on a priority scheme" (Lakshmikantha, col. 17, lines 7–11).
Lakshmikantha teaches something related to triggering a second operation on the basis of a capacity threshold, in that "the scheduler may override the priority level of a file system service during scheduling. For example, the second file system service (e.g., garbage collection) may be scheduled for execution earlier than the first file system service (e.g., file verification) when a measurement of used disk space exceeds a threshold (e.g., 96 percent of disk space used)" (Lakshmikantha, col. 9, lines 1–6). However, Lakshmikantha does not teach:
"triggering a second migration when capacity of the deduplicated storage appliance is predicted to cross a threshold capacity, wherein the second migration is triggered regardless of whether all of the predicted values satisfy the plurality of trigger conditions."
In the same field of endeavor, Perneti teaches "triggering a second migration when capacity of the deduplicated storage appliance is predicted to cross a threshold capacity, wherein the second migration is triggered regardless of whether all of the predicted values satisfy the plurality of trigger conditions."
Perneti discloses the predicted capacity, in that "the process begins with step 200, generating, for at least one storage system that is part of an active tier of a storage environment, at least one storage capacity forecast comprising probability values for storage capacity of the at least one storage system over a designated time period. The storage environment may comprise a deduplication-based storage backup environment" (Perneti, col. 9, lines 28–35).
Perneti discloses the prediction crossing a threshold capacity, in that "in step 202, a determination is made as to whether any of the probability values for storage capacity of the at least one storage system over the designated time period exceed at least one designated storage capacity threshold. Responsive to determining that a given one of the probability values for storage capacity of the at least one storage system over the designated time period exceeds the at least one designated storage capacity threshold, one or more remedial actions are selected in step 204 for freeing up space on the active tier of the storage environment" (Perneti, col. 9, lines 37–48).
Perneti discloses that the action so triggered is a migration to cloud storage, in that "the first remedial action associated with the first storage capacity threshold may comprise identifying and migrating a portion of data stored on the at least one storage system from the active tier of the storage environment to a cloud tier of the storage environment" (Perneti, col. 9, lines 56–62), and that "step 206 may comprise migrating the identified portions of data from the at least one storage system in the active tier of the storage environment to a cloud tier of the storage environment" (Perneti, col. 10, lines 25–30).
Perneti discloses that this second migration is triggered on the capacity forecast alone, and thus regardless of whether the other predicted values satisfy their conditions, in that the remedial actions are selected and initiated solely "responsive to determining that a given one of the probability values for storage capacity… exceeds the at least one designated storage capacity threshold" (Perneti, col. 9, lines 41–48), and are initiated "to prevent the at least one storage system from reaching a designated storage capacity error condition" (Perneti, col. 9, lines 48–51).
Lakshmikantha and Perneti are analogous to the claimed invention as both are from the same field of endeavor of forecasting resource and capacity utilization of a deduplication-based backup storage system in order to schedule data-movement operations. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the capacity-threshold override of Lakshmikantha with the forecast-based capacity threshold and cloud-tier migration remedial action of Perneti. The motivation to combine Lakshmikantha and Perneti is that Lakshmikantha triggers its override only upon a measurement of used disk space already exceeding the threshold (Lakshmikantha, col. 9, lines 1–6), whereas Perneti teaches acting on a forecast so as "to prevent the at least one storage system from reaching a designated storage capacity error condition" (Perneti, col. 9, lines 48–51), such that one of ordinary skill would act on the predicted capacity rather than the measured capacity in order to free space before the appliance runs out of storage.
Claim 13 recites limitations identical to those of Claim 6 and is rejected under the same rationale set forth above with respect to Claim 6.
Claim 19 recites limitations identical to those of Claims 6 and 13 and is rejected under the same rationale set forth above with respect to Claims 6 and 13.
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Lakshmikantha et al. (Lakshmikantha), US 11,514,317 B2, in view of Eda et al. (Eda), US 2019/0155534 A1, further in view of Kottomtharayil (Kottomtharayil), US 8,769,048 B2, and further in view of Xu et al. (Xu), US 8,943,032 B1.
As to Claim 7, the limitations of Claim 1 from which Claim 7 depends are rejected under the same rationale set forth above with respect to Claim 1. The limitations "wherein the workflow comprises migrating the files from the deduplicated storage appliance to the cloud storage and the method further comprises:", "maintaining a schedule specifying when the migration of the files should occur;" and "triggering a first migration according to the schedule; and" are rejected under the same rationale set forth above with respect to Claim 6.
Regarding the additional limitation of Claim 7, Lakshmikantha teaches:
"wherein the second migration is triggered regardless of whether all of the predicted values satisfy the plurality of trigger conditions."
Lakshmikantha discloses that "the scheduler may override the priority level of a file system service during scheduling. For example, the second file system service (e.g., garbage collection) may be scheduled for execution earlier than the first file system service (e.g., file verification) when a measurement of used disk space exceeds a threshold (e.g., 96 percent of disk space used)" (Lakshmikantha, col. 9, lines 1–6). The override is performed on the threshold alone, and thus without regard to whether the availability windows for the several resources have been matched to the registered resource requirements as otherwise required by the scheduler (Lakshmikantha, col. 8, lines 36–41).
Lakshmikantha teaches something related to triggering a second operation upon a quantity of stored data reaching a threshold (Lakshmikantha, col. 9, lines 1–6). However, Lakshmikantha does not teach "triggering a second migration when a total size of the files to be migrated reaches a threshold size."
In the same field of endeavor, Xu teaches:
"triggering a second migration" — Xu discloses "migrating the candidate files from the source storage tier to the target storage tier using the selected data movement method" (Xu, col. 19, lines 5–8), where the storage system is "implemented as part of an archive and/or backup system such as a deduplication storage system available from EMC® Corporation of Hopkinton, Mass." (Xu, col. 4, lines 3–7), and where the target tier may be cloud storage, in that "tiers include different storage technologies (e.g., tape, hard drives, semiconductor-based memories, optical drives, etc.), different locations (e.g., local computer storage, local network storage, remote network storage, distributed storage, cloud storage, archive storage, vault storage, etc.), or any other appropriate storage for a tiered data storage system" (Xu, col. 5, lines 17–26).
"when a total size of the files to be migrated" — Xu discloses "identifying a list of files stored in a source storage tier as candidate files to be migrated to a target storage tier based on a migration policy" (Xu, col. 19, lines 60–63), that "candidate files that are suitable for migration are identified based on migration policies 112" (Xu, col. 5, lines 47–49), and that "source candidate index 110 is used to identify all of the deduplicated segments of all the candidate files to be migrated" (Xu, col. 5, lines 56–59). The aggregate quantity of that identified set is measured as "a size of the candidate files" (Xu, col. 20, lines 17–19).
"reaches a threshold size" — Xu discloses that "the bulk data movement method is selected if a number of the candidate files is above a first predetermined threshold or a size of the candidate files is above a second predetermined threshold" (Xu, col. 20, lines 12–20), and further discloses the same criterion in the description, in that "a bulk data movement method may be selected if there are many candidate files for migration and/or the candidate files are large… A regular data movement method may be selected if there are few candidate files for migration and/or the candidate files are small" (Xu, col. 3, lines 28–39).
Lakshmikantha and Xu are analogous to the claimed invention as both are from the same field of endeavor of managing data-movement operations on a deduplication storage system. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the threshold-based override trigger of Lakshmikantha with the aggregate candidate-file size threshold of Xu. The motivation to combine Lakshmikantha and Xu is as recited by Xu, which teaches that the setup cost of bulk data movement "is fixed, and will not be amortized across small number of files" (Xu, col. 12, lines 55–60), such that conditioning the migration on the total size of the candidate files reaching a threshold ensures the migration is performed only when enough data has accumulated to justify its fixed cost.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to HUNG VAN LE whose telephone number is (571)270-0164. The examiner can normally be reached 8 a.m. - 5 p.m..
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, Cesar Paula can be reached at (571) 272-4128. 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.
/HUNG VAN LE/Examiner, Art Unit 2145
/CESAR B PAULA/Supervisory Patent Examiner, Art Unit 2145
Prosecution Insights