FR3087556A1 - METHOD FOR IMPROVING THE EFFICIENCY OF USE OF THE RESOURCES OF AN INFRASTRUCTURE FOR IMPLEMENTING A SCHEDULING PLAN - Google Patents

Abstract

A method of generating an optimized sizing plan for improving the efficiency of resource utilization of an IT infrastructure involved in performing a plurality of batch processes according to a plan of. 'scheduling, said method comprising: - a step (100) of data acquisition, comprising the acquisition of a first dimensioning plan of a predetermined weighting coefficient for each resource of a plurality of resources and of predetermined execution parameters of a plan scheduling; - a step of calculating (200) a first value of efficiency of use of resources; - a step of determining (300) extended execution times of the plurality of batch processes, as a function of values of quantities of available resources, a step of generation (400) of at least one optimized sizing plan comprising selecting new resource quantity values assigned to the at least one optimized sizing plan.

Description

PROCESS FOR IMPROVING THE EFFICIENCY OF USING THE RESOURCES OF AN INFRASTRUCTURE INTENDED TO EXECUTE A SCHEDULE PLAN

The invention relates to the field of batch processing, and more particularly that of improving the use of resources of a computer infrastructure involved in the execution of batch processing.

In particular, the invention relates to a method for improving the efficiency of using resources of an IT infrastructure for performing a plurality of batch processing operations according to a scheduling plan.

Such a method makes it possible to calculate an optimized dimensioning of the IT infrastructure so that the execution of a scheduling plan is done with improved efficiency.

[Prior art

[0002] In general, batch processing is the process executed by a computer system autonomously (e.g. without user interaction).

They are used for the automation of non-interactive tasks that must be performed reliably before certain deadlines.

For example, a batch job can be used to update information at the end of the day, transform data, generate reports, print documents or perform other non-interactive tasks.

Batch processing can also be used to perform various operations with files, including digital images, such as resizing or converting.

These batch jobs are typically put on hold or accumulated during peak hours when computing resources are unavailable or limited by real-time activity.

They are then executed outside peak hours when computer resources are available.

It is therefore very common that, for technical reasons, these treatments are scheduled at night, on weekends or during periods of low business activity.

The constant evolution of information systems, and the multiplication of batch processing, tends to overload the periods during which the processing operations can be executed.

However, these are often long and critical processes that should be monitored in particular in order to guarantee the proper functioning of business applications during “human” hours of use.

Thus, it is known to batch schedulers.

There is in particular a mechanism described in document EP3113022 consisting in using the consumption levels of 2 resources in the scheduling rules for batch processing.

More specifically, this mechanism is based on execution constraints, the resources they consume and the resources available on the infrastructure.

[0004] In addition, the significant fluctuation over time in the volumes of data leads to very volatile resource consumption and processing times.

For example, the same scheduling on the same infrastructure can lead one day to an overload of resources and an interruption of processing while another day, with a few days intervals, no dysfunction will be detected.

All these changes have an impact not only on the execution time of the processing operations, but also on the quantity of the resources used.

In fact, information systems (IS) administrators tend to increase the resources of their infrastructure so that the batch processes are completed in the allotted time and that they do not lead to a saturation of the available resources. .

Indeed, in some cases, production can be stopped if the treatments could not be carried out in time. 15

[0005] These changes generally lead to an oversizing of the IT infrastructures involved in the execution of batch processing according to scheduling plans.

Now, with more and more common uses of these batch processing operations and constraints in the use of resources, it is necessary for IS administrators to have methods or devices available which make it possible to improve the efficiency of the process. the use of resources and reduce their waste.

[0006] In fact, an inappropriate use of certain resources during the execution of batch processing can lead to a slowing down of the execution of the batch processing, incompatibility with other batch processing operations and exhaustion of resources.

In addition, the allocated resources being variable between each batch process, it is difficult to organize, program or more generally to execute the batch processes.

[0007] However, to date there is no solution which makes it possible to calculate a dimensioning of an IT infrastructure allowing the proper execution of the scheduling plan, without wasting resources and with maximum efficiency of use.

Indeed, with the known mechanisms, it can occur a waste of 30 resources.

In addition, the execution of batch processes in virtualized environments becomes generalized and the definition of an optimized sizing with an improved efficiency of use of the resources allowing a good execution of the processes by 3 batches would allow to be able to put in parallel a more large number of batch processes on the same IT infrastructure.

[0008] Thus, there is a need for new methods or devices capable of generating a dimensioning plan allowing the improvement of the efficiency of use of resources for the execution of a plurality of batch processes.

Technical problem!

[0009] The aim of the invention is to remedy the drawbacks of the prior art.

In particular, the object of the invention is to propose a method for improving the efficiency of the use of the resources of an IT infrastructure for the execution of a plurality of batch processing operations according to a scheduling plan, said method being capable of generating an optimized sizing plan relating to a quantity of resources sufficient to carry out the scheduling plan, and of which the efficiency of use of the resources is improved. 15

The invention further aims to provide a device capable of implementing a method of improving the efficiency of use of the resources of a computer infrastructure so that it is capable with a minimum resource to run the batch processes in a predetermined maximum time and with maximum efficiency of the resources used.

20 JBrief description of the invention!

To this end, the invention relates to a method for generating an optimized sizing plan for improving the efficiency of use of the resources of an IT infrastructure involved in the execution of a plurality of batch processing operations according to a scheduling plan, said method comprising: a step of data acquisition, by an acquisition module, comprising the acquisition of a first dimensioning plan, of a predetermined weighting coefficient for each resource of a plurality of resources and of predetermined execution parameters of a scheduling plan, said first dimensioning plan comprising first values of quantity of resources 4 allocated allowing the execution of a plurality of processes in batches according to the scheduling plan over a period corresponding to a first overall duration of execution of the scheduling plan; a step of calculating a first value of efficiency of use of resources, by an optimization module, from the first values of quantity of resources allocated, from the predetermined weighting coefficients and from the first overall duration d 'execution of the scheduling plan; a step of determining extended execution times of the plurality of batch processes, as a function of values of quantities of available resources, comprising: the generation on an IT infrastructure, by a resource scarcity module, of a controlled consumption of at least one resource according to previously stored rarefaction rules so as to leave a quantity of available resources available for a batch processing less than an initial quantity of resources, o execution on the infrastructure data processing, by a batch processing execution module, of each of the batch processing operations of the plurality of batch processing operations, and a step of measuring, by a metrics collection module, of execution times extended in function of values of quantity of resources available for each of the batch processes; and a step of generation, by the optimization module, of at least one optimized sizing plan, said generation step comprising: the definition of resource quantity values, lower than the first 25 allocated resource quantity values , the calculation, for each value of quantity of resources, of a new value of efficiency of use of resources, said calculation being a function of the values of quantity of resources, of the predetermined weighting coefficients and of the extended execution times measured batch processes, and the selection of new values of quantity of resources allocated to the at least one optimized dimensioning plan, said new values of quantity of allocated resources being selected from among the values of quantity of resources including the new one Resource use efficiency value is improved over the first resource use efficiency value 5

The method according to the invention makes it possible to calculate a resource sizing, for an IT infrastructure, capable of supporting the proper execution of a scheduling plan and making it possible to maximize the efficiency of the use of resources.

It makes it possible, for example, to determine the dimensioning (s) comprising a more efficient combination of resources for a given scheduling plan.

Many methods aim to reduce the execution time of batch processes or to secure their execution by ensuring a quantity of resources much greater than needed.

Here, the method makes it possible to generate a dimensioning plan improving the efficiency of use of the resources involved or allocated.

That is, the values of 10 allocated resources are not selected to be as low as possible or to allow the fastest execution, but they are selected to correspond to the quantities of resources that are most suitable with regard to factors specific to its resources (availability, gross quantity, demand, etc.) as modeled by the weighting coefficient. 15

[0013] According to other optional characteristics of the method: the generation of at least one optimized sizing plan further comprises the calculation of execution parameters of the scheduling plan as a function of values of efficiency of use of the resources and in that the new allocated resource quantity values are also selected from among the resource quantity values whose calculated execution parameters respect predetermined execution parameters of the scheduling plan.

Thus, it is for example possible to verify that the predetermined maximum overall execution time of the scheduling plan is not exceeded. 25 - the first sizing plan is obtained according to a sizing plan generation step comprising: oua data acquisition step, by an acquisition module, comprising the acquisition of predetermined execution rules, and of a footprint on a plurality of resources of each batch process 30 or a calculation step, by a dimensioning module, comprising, as a function of the footprint on a plurality of resources of each batch process and according to a scheduling plan respecting the predetermined execution rules, the calculation of maximum values of resource consumption per period by batch processing, and 6 or a step of defining a first dimensioning plan of an IT infrastructure, by the dimensioning module, according to maximum values of resource consumption per calculated periods, said first dimensioning plan comprising first values of quantity of resources a ttributed - it includes the measurement, by the metrics collection module, of a footprint of each of the batch processes and the identification, by the optimization module, of critical resources for each of the batch processes.

This measurement can be carried out upstream of the method, for example prior to the step of determining extended execution times.

It makes it possible to select the resources to be studied within the framework of the optimization of the dimensioning. the generation step further comprises: the calculation, for each quantity values generated of resources, of a new overall execution time of the scheduling plan as a function of the measured extended execution times of the processing operations by batches, and the deletion of the resource quantity values, for which the new overall execution time of the corresponding scheduling plan is greater than a predetermined maximum overall execution time of the scheduling plan. This allows the method according to the invention of generating dimensioning plans exhibiting improved efficiency in the use of resources but also respecting predetermined maximum execution times. - during the generation step, the method comprises the generation of a plurality of optimized dimensioning plan. - It further comprises a step of verifying compliance with the scheduling plan, said step comprising: the execution, by the batch processing execution module and on an IT infrastructure whose available resources correspond to the scheduling plan. optimized sizing, of the plurality of batch processes according to the scheduling plan, 35 o the measurement, by the metrics collection module, of execution parameters of the scheduling plan, and 7 o the comparison, by a module verification, measured execution parameters of the scheduling plan with predetermined execution parameters of the scheduling plan.

Thus, it is possible to verify that the optimized sizing plan generated effectively allows the execution of the scheduling plan under the prescribed conditions.

The comparison can preferably be carried out over the overall execution time of the scheduling plan. the step of verifying compliance with the scheduling plan comprises beforehand the generation on the IT infrastructure, by the resource scarcity module, of a controlled consumption of several resources according to the optimized sizing plan.

This makes it possible to check before going into production on a dedicated infrastructure that the sizing plan is adapted to the scheduling plan and its execution rules. 15 - the verification step further comprises: the calculation, by the verification module, of a difference between a predetermined maximum overall duration of execution of the scheduling plan and a measured overall duration of execution of the plan d 'scheduling, the comparison of the difference calculated, by the verification module, with a predetermined performance threshold, and the validation, by the verification module, of the optimized sizing plan when the predetermined performance threshold is not not exceeded.

Thus, it is possible to check whether it is possible to further optimize the dimensioning plan when the latter still allows, despite the reduction in resources, the execution of the scheduling plan in a period much less than the maximum overall duration. predetermined.

These steps further reduce the waste of resources. 30 - it further comprises a step of modifying the optimized sizing plan if at least one of the following predetermined execution parameters of the scheduling plan is not respected: execution rule, predetermined maximum duration of execution of batch processing and predetermined maximum overall duration of the scheduling plan. 35 8

According to another aspect, the invention relates to a device for generating an optimized sizing plan for improving the efficiency of use of the resources of an IT infrastructure involved in the execution of a plurality of batch processes according to a scheduling plan according to the invention, characterized in that it comprises: 5 - an acquisition module configured to acquire a first dimensioning plan, a predetermined weighting coefficient for each resource of a plurality of resource and predetermined execution parameters of a scheduling plan, said first sizing plan comprising first allocated resource quantity values allowing the execution of a plurality of batch processes according to a plan of scheduling, said execution taking place over a period corresponding to a first overall duration of execution of the scheduling plan; a resource rarefaction module configured to generate, on an IT infrastructure, a controlled consumption of at least one resource 15 according to previously stored rarefaction rules so as to leave a quantity of available resources available for batch processing less than an initial amount of resources, - a batch processing execution module configured to execute, on the IT infrastructure, each of the batch processes of the plurality of 20 batch processes, and - a metrics collection module configured to measure extended execution times as a function of values of quantity of resources available for each of the batch processes, - an optimization module configured to calculate a first value of efficiency of use of resources from the first values of allocated resource quantity, predetermined weighting coefficients and the first overall duration e execution of the scheduling plan, and configured to define resource quantity values lower than the first allocated resource quantity values; calculating, for each resource quantity value a new resource utilization efficiency value, said calculation being a function of the resource quantity values, the predetermined weighting coefficients and the measured extended execution times of the batch processes , and selecting new resource quantity values allocated to the at least one optimized sizing plan, said new allocated resource quantity values being selected from among the resource quantity values including the new efficiency value of. Resource utilization is improved from the first value of resource utilization efficiency.

Like the method according to the invention, such a device makes it possible to calculate the dimensioning of an IT infrastructure, necessary for the proper execution of a scheduling plan while maximizing the efficiency of use. Resource.

Thus, unlike conventional devices or systems aiming to optimize the scheduling plan or to ensure the absence of malfunction during the execution of a scheduling plan, such a device will make it possible to reduce the waste of resources associated with maintaining oversized IT infrastructure dedicated to performing batch processing.

According to another aspect, the invention further relates to a computer program comprising instructions which, when the computer program is loaded into the memory of a computer and is executed by a processor of said computer, cause the implementation by the computer of all the steps of the method according to the invention. 15

According to another aspect, the invention further relates to a tangible recording medium, readable by a computer, on which the program according to the invention is recorded in a non-transient manner.

[0018] Other advantages and characteristics of the invention will become apparent on reading the following description given by way of illustrative and non-limiting example, with reference to the appended figures which represent: - Figure 1, a schematic representation of a method of generating an optimized sizing plan for improving the efficiency of use of the resources of an IT infrastructure involved in the execution of a plurality of batch processes according to the invention.

Steps enclosed by dotted lines are optional; - Figure 2, a schematic representation of a preliminary step for identifying critical resources; FIG. 3, a schematic representation of a preliminary step for generating a dimensioning plan; FIG. 4, a schematic representation of a step for determining extended execution times, of the plurality of batch processes, as a function of values of quantity of resources available according to one embodiment of the invention; - Figure 5A, a schematic representation of a step for generating at least one new optimized dimensioning plan according to one embodiment of the invention - Figure 5B, a schematic representation of a step for generating at least one new optimized sizing plan according to one embodiment of the invention; FIG. 6, a schematic representation of a step for verifying compliance with the scheduling plan according to the invention; - Figure 7, a schematic representation of a step of verifying compliance with the scheduling plan according to one embodiment of the invention - Figure 8, a schematic representation of a step of additional modification of the quantity of allocated resources the optimized dimensioning plan according to the invention; FIG. 9, a schematic view of the device for optimizing the sizing of an IT infrastructure for the execution of a plurality of batch processing operations according to the invention. 20

Aspects of the present invention are described with reference to flowcharts and / or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention.

[0020] In the figures, flowcharts and block diagrams illustrate the architecture, functionality and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. .

In this regard, each block in the flowcharts or block diagrams may represent a system, device, module or code, which comprises one or more executable instructions for implementing the specified logic function (s).

In some implementations, the functions associated with the blocks may appear in a different order than that shown in the figures.

For example, two blocks shown in succession may, in fact, be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order, depending on the functionality involved.

Each block of the block diagrams and / or the flowchart, and combinations of blocks in the block diagrams and / or the flowchart, can be implemented by special hardware systems that perform the specified functions or acts or perform combinations of special equipment and computer instructions. f Description of the invention]. 10

In the remainder of the description, the expression "improved the efficiency of use of resources" within the meaning of the invention is understood to mean the action of maximizing or improving the use of the resources of an IT infrastructure. , within the framework of the execution of a batch processing, compared to a level prior to the implementation of the invention.

This better efficiency is made possible by the choice of quantity of resources so as to favor the most available resources and the use of which results in the lowest possible impact on the IT infrastructure or its administration.

The expression this batch processing "(" batch processing "in English terminology) corresponds to an automatic sequence of several phases by means of a series of commands (process) on a computer without intervention 20 d 'an operator.

Once a process is finished the computer processes a next batch and at a time T there are usually several batch processes running in parallel.

Batch processing ends after all batches in the stack (queue) have been executed.

Batch processes are mainly used for automated tasks, for example for the automatic sending of postal or electronic mail.

A batch process is generally stored in a batch file which is executed by the computer system.

Running the batch file can produce a number of actions as diverse as updating databases, sending emails to users, or producing one or more output files to use by another batch process. 30

The "scheduling plan" within the meaning of the invention, also called "execution plan", corresponds to the order of execution of the various batch processes to be carried out during a period of time. given.

This scheduling is generally calculated as a function of rules and / or prerequisites so as to satisfy resource availability requirements.

The expression “execution parameters of the scheduling plan” within the meaning of the invention corresponds to a set of parameters associated with the execution of the scheduling plan.

This may for example include the overall execution time of the scheduling plan, compliance with execution rules or the execution time of batch processes.

The execution parameters of the scheduling plan can also include: the use of resources by the batch processes over time, or even a volume of data generated or an occupation of the memory space after the end of the treatments by lots.

These values can for example be measured by probes and they can also be predetermined, the invention including in particular the comparison of values of parameters measured with values of predetermined parameters.

The expressions "execution rules" or "predetermined execution rules" within the meaning of the invention correspond to rules for the execution of a scheduling plan, compliance with which allows execution to meet the needs of the SI administrator.

These rules can in particular include sequence orders to be respected for batch processing, schedules to be respected, or resource consumption thresholds to be respected. 20

The expression "execution time" within the meaning of the invention corresponds to the execution time of a batch process or of several batch processes.

In other words, it is the time required for the computer system (e.g. processor) to execute the orders.

The expression “predetermined maximum execution time” within the meaning of the invention corresponds to the execution time that a batch process must not exceed.

It may correspond, for example, to the duration of execution of a batch processing which must not be exceeded so as not to cause a malfunction or a late finalization of a scheduling plan.

The expression "extended execution time" within the meaning of the invention corresponds to the execution time of a batch process or of several batch processes when the available resources have been reduced.

The expression "overall execution time of the scheduling plan" within the meaning of the invention corresponds to the execution time of the scheduling plan.

In other words, 13 is the time necessary for the computer system (e.g. processor) to execute the orders associated with all the batch processes constituting the scheduling plan.

The predetermined maximum overall execution time of the scheduling plan within the meaning of the invention corresponds to the execution time that a scheduling plan must not exceed.

It may correspond, for example, to the time not to be exceeded so as not to cause a malfunction or a late finalization of a scheduling plan.

The expression "IT infrastructure" within the meaning of the invention corresponds to a set of IT structures (i.e. IT devices) capable of operating an application or an application chain.

The computer structure can be one or more servers, it can for example be composed of a presentation server, a business server and a data server.

The IT infrastructure can also be a “test” server with the same resources as a production server or a clone of a production virtual server.

Preferably, a computer structure is a server. 15

The expression "dimensioning plan" within the meaning of the invention corresponds to the quantities of resources granted to, or held by, an IT infrastructure in particular within the framework of the execution of a scheduling plan.

By "probe" or "computer probe" is meant within the meaning of the invention software associated with equipment which makes it possible to perform, manage and send back to computer equipment measurements of the values of parameters. the use of resources, batch processing execution parameters or the operating state of resources.

[0032] The term “resource” within the meaning of the invention corresponds to parameters, capacities or functions of computer devices allowing the operation of a batch process or of an application.

A single computer device is generally associated with several resources.

Likewise, the same resource can be shared between several batch processes.

A resource is generally associated with a unique identifier allowing it to be identified within an IT infrastructure.

For example, the term "resource" can include: network disks characterized for example by their inputs / outputs, reading / writing on disks, the rate of memory utilization, a network characterized by its bandwidth, a processor characterized for example by its use (in percent) or the occupancy rate of its caches, a random access memory characterized by the amount allocated.

The term “resource use” is understood to mean the consumption of a resource by a batch processing or by a business application.

The expression “critical resource” within the meaning of the invention corresponds to the resource (s) whose use under nominal (ie usual) conditions is closest to its limit use (predetermined maximum use threshold) or to its use. saturation or else a resource identified as such by a user.

Thus, the scarcity of a critical resource means that the execution conditions of the application brick are not respected (e.g. resource consumption threshold not respected or processing times not respected).

For example, in an IT infrastructure comprising a CPU and 8 Gb of RAM where under usual conditions there is consumption of 80% of CPU and 2 Gb of RAM 10 then the critical resource will be the CPU resource.

The term “footprint” within the meaning of the invention corresponds to data on the resource consumption of a batch process or the resource availability of a machine (for the empty footprint) during its execution period.

The “step of a resource” can be on the one hand the minimum quantity of resource that can be added or subtracted during a modification step and on the other hand a multiple of the modification that can be subjected to the level of said resource.

Some resources have a high particle size (e.g. servers) while others have a low particle size (e.g. available RAM).

The expression “initial quantity of resources” within the meaning of the invention corresponds to the values of quantity of resources of an IT infrastructure before the establishment of a controlled consumption of resources.

The expression “quantity of available resources” within the meaning of the invention corresponds to the values of quantity of resources which will be accessible to batch processing after the implementation of controlled consumption.

The expression "quantity of resources allocated" within the meaning of the invention corresponds to the values of quantity of resources calculated, according to the method of the invention, as being sufficient for the IT infrastructure to implement a plan. scheduling executing over a period of less than a predetermined maximum overall duration.

The expression "controlled consumption of at least one resource" within the meaning of the invention corresponds to a reduction in the quantity of resource that can be used by a batch processing running on an IT infrastructure, for example via the execution of codes on said computer infrastructure calling on the resources concerned. 15

The term "process", "calculate", "determine", "display", "extract" "compare" or more broadly "executable operation", within the meaning of the invention, means an action performed by a device or a processor unless the context indicates otherwise.

In this regard, operations refer to actions and / or processes of a data processing system, for example a computer system or an electronic computing device, which manipulates and transforms the data represented as physical (electronic ) in computer system memories or other information storage, transmission or display devices.

These operations can be based on applications or software. 10

The terms or expressions "application", "software", "program code", and "executable code" mean any expression, code or notation, of a set of instructions intended to cause data processing to perform a particular function directly or indirectly (eg after a conversion operation to another code).

Examples of program code may include, but are not limited to, a routine, function, executable application, source code, object code, library, and / or any other sequence of instructions designed for the purpose. execution on a computer system.

By "processor" is meant, within the meaning of the invention, at least one hardware circuit configured to perform operations according to instructions contained in a code.

The hardware circuit can be an integrated circuit.

Examples of a processor include, but are not limited to, a central processing unit, a graphics processor, an application specific integrated circuit (ASIC), and a programmable logic circuit.

By "coupled" is meant, within the meaning of the invention, connected, directly or indirectly with one or more intermediate elements.

Two elements can be mechanically coupled, electrically coupled, or linked by a communication channel.

The expression "man-machine interface" within the meaning of the invention corresponds to any element allowing a human being to communicate with a particular computer and without this list being exhaustive, a keyboard and means allowing in response to the commands entered at the keyboard to carry out displays and possibly to select, using the mouse or a touchpad, elements displayed on the screen.

Another exemplary embodiment is a touch screen making it possible to select directly on the screen the elements touched by the finger or an object and possibly with the possibility of displaying a virtual keyboard. 16

In the remainder of the description, the same references are used to designate the same elements.

As has been mentioned, most of the systems and methods of the prior art are concerned with optimizing the scheduling of batch processing operations or else organizing the dimensioning of the infrastructures supporting these batch processing operations. in order to ensure a high level of service.

On the contrary, in the context of the present invention, the inventors have developed a method and a device making it possible to prioritize the use of resources with maximum efficiency.

Thus, the batch processes can be executed in a given time by mobilizing a given amount of resources so as to have the lowest possible impact on the other processes to be executed on the IT infrastructure.

For example, running batch processing will be prioritized over resources that are not essential to other applications or that are widely available within the IT infrastructure. 15

For this, the inventors propose in particular a method and a device capable of calculating the dimensioning of the IT infrastructure necessary for the efficient execution of a scheduling plan.

The objective is in particular to determine the dimensioning (s) making it possible to maximize the efficiency of the use of resources and also to ensure that the execution of the scheduling plan respects a predetermined maximum duration of execution of each processing. or more broadly a predetermined maximum overall duration of execution of the scheduling plan, as well as preferably rules for executing said scheduling plan.

As illustrated in FIG. 1, according to a first aspect, the invention relates to a method for generating a dimensioning plan optimized for improving the efficiency of use of the resources of a. IT infrastructure involved in the execution of a plurality of batch processing operations according to a scheduling plan, said method comprising a data acquisition step 100 comprising the acquisition of a first dimensioning plan, a calculation step 200 of a first value of resource utilization efficiency, a step of determining 300 extended execution times of the plurality of batch processes as a function of values of quantities of resources available, and a step 400 of generating d '' at least one optimized dimensioning plan. 17

In particular, the method according to the invention comprises a step 100 of data acquisition that can be implemented by an acquisition module 10.

This step may include in particular the acquisition of a first dimensioning plan comprising first values of quantity of resources allocated allowing the execution of a plurality of batch processes.

This first dimensioning plan and these first values of quantity of resources allocated can correspond to values of maximum quantity of resources which should be reduced within the framework of the method according to the invention.

Conventionally, the execution of the plurality of batch processes is carried out according to a scheduling plan and takes place over a duration corresponding to a first overall duration of execution of the scheduling plan.

As will be detailed later and unlike many innovations in this field, the object of the invention is not to reduce this first overall execution time of the scheduling plan but on the contrary to determine a dimensioning an IT infrastructure with quantities of resources having the least impact on the operation of the other 15 applications or more generally on the management of the IT infrastructure.

In addition, so as to allow the choice of the least impacting resources, this data acquisition step 100 comprises in particular the acquisition of a predetermined weighting coefficient for each resource of a plurality of resources, preferably a plurality of critical resources, and predetermined execution parameters of a scheduling plan.

The predetermined weighting coefficient for each resource may be a function of factors capable of modeling the characteristics of the resources.

For example, a resource activity factor can be taken into account in the predetermined weighting coefficient and correspond to the current situation of the resource and its trend of evolution.

This activity factor can also relate to the governance of the resource and be representative of the influence of the resource on the performance of the batch processing operations and of the temporality of this resource.

The predetermined weighting coefficient can also take into account a factor specific to the management of the resource, the regulation of its structure comprising, for example, the influence of the resource within the IT structure, its organization and its hosting. within the IT structure.

A resource safety factor taking into account the safety of each resource and its risk of malfunction or anomaly as well as its level of reliability can also be modeled by the predetermined weighting coefficient.

Finally, the predetermined weighting coefficient can also be a function of a performance factor of the resource 35 taking into account its level of use and / or its storage, as well as its yield. 18

As illustrated in Figure 2, the method according to the invention may include beforehand a measurement 101, by a metrics collection module 40, of an imprint of each of the batch processing operations and the identification 102, by a resource management module, critical resources for each of the batch processes.

Indeed, the method according to the invention aims to optimize the dimensioning of an IT infrastructure and more particularly to reduce to a minimum the quantity of resources used within the framework of a scheduling plan and to improve the efficiency of use of the resources of a dimensioning plan.

It may be advantageous to focus, in the context of this sizing optimization, in only part of the resources referred to as critical resources.

Thus, the values of quantities of resources studied within the framework of the method according to the invention can represent only a part of the resources of an IT infrastructure and are preferably the critical resources.

Thus, preferably, the first scheduling plan can consist of a list of the values of quantity of resources, for example critical, not to be exceeded. 15

The method according to the invention can also include a step of generating a dimensioning plan.

Indeed, while the method can preferably be applied from a sizing plan already known or well estimated, it is possible that the sizing plan allowing the execution of the scheduling plan is not known.

The step of generating a dimensioning plan according to one embodiment of the invention is illustrated in FIG. 3.

It may in particular include a step 110 of data acquisition, which can be implemented by the acquisition module 10, comprising the acquisition of predetermined execution rules, and the acquisition of a resource consumption footprint on a plurality of resources from each batch process. 25

The step of generating a sizing plan also comprises a calculation step 120 comprising, as a function of the footprint of each batch process and according to a scheduling plan respecting the predetermined execution rules, the calculation of maximum values of resource consumption per period by batch processes.

This calculation step 120 can for example be covered by a dimensioning module 70.

The step of generating a sizing plan may also include a step of defining 130 a first sizing plan of an IT infrastructure 2 as a function of the maximum values of resource consumption per 19 calculated periods, said first dimensioning plan comprising first values of quantity of allocated resources.

This step of defining 130 a first dimensioning plan can also be implemented by a dimensioning module 70. 5

The method of generating a dimensioning plan for improving the efficiency of use of the resources of a dimensioning plan of an IT infrastructure also comprises a step of calculating 200 of a first value d efficiency of use of resources, by an optimization module 50, on the basis of the first values of quantity of resources allocated, of the predetermined weighting coefficients and of the first overall execution time of the scheduling plan.

This calculation step can be performed for each resource.

In addition, this step makes it possible to obtain an efficiency value for the use of resources as a function of the quantity of resource allocated and of the predetermined weighting coefficients by taking into account the overall duration of execution of the scheduling plan.

Thus, the first resource efficiency value makes it possible to determine a first dimensioning for the execution of the batch processes according to a duration corresponding to a first overall duration of execution of the scheduling plan.

In addition, some resources are more available than others in terms of IT infrastructures due to their usefulness, accessibility, etc. and are therefore subject to different constraints.

It is therefore necessary for the proper execution of batch processes and for their execution optimization, that the use of resources be done with maximum efficiency.

The first efficiency value makes it possible to obtain an efficiency value for all the resources used for the execution of the batch processes according to a scheduling plan.

This constitutes a starting value which should be improved by means of the method according to the invention.

The optimization method according to the invention also comprises a step of determining 300 of extended execution times of the plurality of batch processes as a function of values of quantities of available resources.

An embodiment of this determination step 300 is illustrated in FIG. 4.

This step 300 can be initiated by the selection of a batch process then, as illustrated, this step comprises the selection 305 of at least one resource, preferably critical, by which the process of calculating the execution times will be initiated. extended.

Then, there is generation 310 on a computer infrastructure 2 of a controlled consumption of at least one resource according to rarefaction rules previously stored so as to leave available to the selected batch processing a quantity of available resources less than one. initial amount of resources.

As will be detailed below, this controlled consumption, or rarefaction, can be implemented by a module for rarefaction of resources 20.

This controlled consumption generally takes into account the granulometry of the resources and therefore takes into account the pitch of each of the resources.

Preferably, following the establishment of the rarefaction or before the establishment of the rarefaction (ie the main thing is that the steps are at least concomitant), this step comprises the execution 320 on the. IT infrastructure 2 of the batch processing selected from among the plurality of batch processing operations.

The execution can conventionally be carried out by a batch processing execution module 30.

Then, there is a measure 330 of extended execution times as a function of values of quantity of available resources.

The measurement can in particular be carried out by a metrics collection module 40 which will be described below.

There can be as many measures of lengthened execution times as there have been “steps” of scarce resources.

There may for example be three measures as presented in Table 1 (the first measure corresponds to the absence of controlled consumption).

Since the amount of resource available for the batch processing is reduced (i.e. due to scarcity), the execution time of the batch processing is lengthened.

Indeed, in the presence of a reduction in available resources, the durations of the batch processing tend to lengthen.

These measured data can be recorded in one or more backup files such as flat files of spreadsheet type, or database making it possible to record a correspondence between the execution time of a batch process. depending on a value of available resources.

Table 1 represents an example of recording the durations (in seconds - s) of execution of the batch processes T1, T2, and T3 according to the values of resources R1 (CPU) and R2 (random access memory), namely 100 %, 75%, 50% and 25% (percentage of the initial resource quantity).

Further, a first resource efficiency value can be calculated for each resource independently.

Table 1 Treatment Ri - Ri - R2 - R2 - in batches Quantity Duration Quantity Duration T1 100% 9256 s 100% 9256 s 21 T1 75% 9865s 75% 12320s T1 50% 12327 s 50% 18635 s T1 25% 16256 s 25% 22845 s T2 100% 1820 s 100% 1820 s T2 75% 1981 s 75% 1956 s T2 50% 2406s 50% 2142s T2 25% 3561 s 25% 2365 s T3 100% 4120s 100% 4120s T3 75% 6245 s 75% 5641 s T3 50% 8456 s 50% 6742 s T3 25% 10578 s 25% 8102 s

This determination step can also include the recording of the imprint of the resource consumption of the batch processing during its execution in the presence of a rarefaction of a resource, for example critical.

Indeed, the execution of a batch process in the presence of a scarcity of resources can have an impact on other resources given the interdependence of certain resources.

For example, running a batch process in the presence of a reduction in RAM may result in an unwanted increase in I / O accesses.

As shown in Figure 4, once the lengthened runtime values have been measured and recorded, the determining step 300 may include a comparison 350 of the extended runtimes with a maximum runtime of the. batch processing.

For example, in the case presented in Table 1, it has been predetermined that the processing T1 must be carried out in a maximum of 15,000 seconds.

Thus, in Fig. 4, the reduction to 75% of the resource R1 is acceptable (OK) and therefore there may be an increase 355 in the controlled consumption of the resource until a reduction to 25% of the resource R1 is reached. which causes the maximum execution time (NOK) to be exceeded.

The consumption controlled on the resource R1 is then stopped and the method then comprises a step of storing the data generated.

[0059] Then, it can comprise a comparison 360 of the resources, preferably critical, which have been tested in relation to the resources to be tested and a change 365 of the resource to be tested.

Still according to Table 1, during the comparison 350, the resource R2 leads from 75% to exceeding the maximum execution time (NOK) and being the 22 last resource, preferably critical, to test 360 (0K), the process may include a change 370 in the batch process to be investigated.

Thus, as the batch processes involved in the scheduling plan are tested, it is possible to collect for each of them an extended execution time as a function of the available resources.

The method according to the invention also includes a step 400 for generating at least one optimized dimensioning plan.

As will be described below, this generation step 400 can be implemented by an optimization module 50.

As illustrated in FIG. 5A, the step 400 of generating at least one optimized sizing plan may include a step of defining 410 values of quantity of resources, less than the first values of quantity of resources allocated.

Thus, it is possible to generate for example randomly, or preferably from data on the execution times, a set of resource quantity values within which the resource quantity values of the optimized sizing plan will be selected. .

The definition 410 of resource quantity values less than the first allocated resource quantity values can be followed by a step of calculating 420, for each resource quantity value, of a new efficiency value of. resource use.

In addition, this calculation is a function of the resource quantity values, the predetermined weighting coefficients and the measured extended execution times of the batch processes of the scheduling plan.

Thus, it is possible to obtain new values of efficiency of use of resources in relation to the execution parameters of the scheduling plan and in particular with measured extended execution times of the batch processing operations.

25 Table 2 Values of Ri - R2 - R3 - Value Duration quantities of Quantity Quantity Quantity of overall resource execution efficiency V1 100% 100% 100% 100 16256 s V2 75% 75% 100% 150 20 248 s V3 75% 50 % 100% 190 27748 s V4 50% 75% 90% 180 24651 s 23 V5 50% 50% 90% 210 32157 s

Table 2 presents for example a fictitious example of first values of quantity of allocated resources (V1) and of several values of quantity of resources defined during a step 410 (V2, V3, V4, V5).

For each of these values, Table 1 shows the new resource use efficiency values, for example 5, calculated during step 420.

The step of selecting new values of the quantity of resources allocated may further comprise a calculation of the execution parameter of the scheduling plan as a function of new values of resource use efficiency and a comparison of the values. run parameters calculated at predetermined run parameters.

Thus, the generation of an optimized sizing plan may include the selection of resource quantity values for which calculated or measured execution parameter values are less than predetermined execution parameter values.

For example in table 2, the last column comprises the overall calculated execution times of the scheduling plan.

It can be seen that the V5 values exceed a limit of 28800 s corresponding to a predetermined maximum overall duration of execution of the scheduling plan.

Thus these values are discarded in the context of the method according to the invention.

The generation step 400 then comprises the selection 430 of new values of quantity of resources allocated to the at least one optimized sizing plan.

These new values of quantity of allocated resources are selected from among the values defined during step 410 and correspond to the values for which the new value of efficiency of use of resources is improved compared to the first value of efficiency d. use of resources.

Alternatively, the IS administrator can select the appropriate resource quantity values.

The improvement of the new resource efficiency value corresponds to a maximized value relative to the first resource efficiency value.

With reference to Table 2, it is the values of quantity of resources V3 which present the maximum efficiency value.

Indeed, although the values of quantity of resources V4 are generally lower, they are not those which will make it possible to maximize the efficiency of use of the resources within the framework of this scheduling plan.

According to another embodiment of the generation step 400, illustrated in FIG. 5B, this step can also include a calculation step 450 for each value of the generated quantity of resources, of a new overall duration d '' execution of the scheduling plan according to the measured extended execution times of the batch processes.

This calculation can be carried out for each value of quantity of resources generated during step 410.

In order to generate the optimized sizing plan, a selection step 460 is implemented.

This step comprises the selection of new values of quantity of resources allocated for the optimized sizing plan, for which the new overall execution time of the corresponding scheduling plan is less than a predetermined maximum overall execution time of the plan of scheduling. 10

In addition, as presented in FIG. 5B, the method according to the invention can comprise the generation 440 of a plurality of optimized dimensioning plans.

Generation 440 of a plurality of optimized sizing plans allows an IS administrator to select the one whose available resources match the required resources and the improved resource utilization efficiency value.

Thus, in the case of a generation 440 of a plurality of optimized sizing plans the selection of a sizing plan to be applied may be left to a user.

Nevertheless, advantageously, the method according to the invention can comprise a step of automated selection 500 of an optimized sizing plan.

Following the step of generating 400 at least one optimized sizing plan 20 or following the possible selection 500 of an optimized sizing plan, the method according to the invention may further comprise a step of verification 600 of compliance with the scheduling plan, for example during its implementation on an IT infrastructure corresponding to the new optimized sizing plan.

As will be described below, this verification step 600 can be implemented by a verification module 60.

Although the sizing plan was generated from data measured over the plurality of batch processes involved in the scheduling plan, there may be deviations between the calculated values and the actual runtime values.

In fact, there may be edge effects between the resources and thus, the decrease in one resource may have an impact on another resource, in turn impacting the duration of execution of the batch processes.

FIG. 6 illustrates such a verification step 600 according to a first embodiment of the invention.

This step first requires having an IT infrastructure corresponding to the optimized sizing plan.

The user can set up a virtual server having characteristics in accordance with the new optimized dimensioning plan, on the one hand or he can alternatively have an IT infrastructure of server type having identical or similar dimensioning to the design plan. optimized sizing, on the other hand.

Alternatively, for verification purposes, it can use an IT infrastructure having a dimensioning greater than the new optimized dimensioning plan and implement a step 610 of generating a controlled consumption of one or more resources so as to match sizing to the new optimized sizing plan.

This generation of controlled consumption can be implemented by a resource scarcity module 20.

Verification step 600 comprises the execution 620 on the IT infrastructure, for example by a batch processing execution module 30, of the scheduling plan for the plurality of batch processing operations.

Advantageously, the IT infrastructure has quantities of resources corresponding to the dimensioning plan.

During this execution, the method comprises the measurement 630, for example by a metrics collection module 40, of execution parameters of the scheduling plan.

The execution parameters of the measured scheduling plan are for example selected from: the execution times of the batch processes and the overall execution time of the scheduling plan. 20

Once the execution of the scheduling plan is finalized, the verification step 600 comprises the comparison 640, for example by a verification module 60, of the measured execution parameters of the scheduling plan with parameters d 'predetermined execution of the scheduling plan.

The predetermined execution parameters of the scheduling plan can for example be selected from: execution rules, predetermined maximum execution times for batch processing or a predetermined maximum overall execution time of the scheduling plan .

As has been described, during the verification step 600, the scheduling plan is executed on an IT infrastructure whose available resources correspond to the new optimized sizing plan whether or not this is due to the presence controlled consumption of resources, preferably critical, on the IT infrastructure.

According to the embodiment described in FIG. 7, the step 600 of verifying compliance with the scheduling plan may comprise the setting up 611 of a controlled consumption of resources so that the quantity of resource available for the scheduling plan corresponds to the new optimized dimensioning plan.

Alternatively, it is an IT infrastructure (e.g. virtual machine) whose quantity of resources corresponds to the optimized sizing plan 5 which is used.

Following the execution 620 of the scheduling plan and the measurement 631 of the overall execution time of the scheduling plan and / or the measurement of the execution times of each of the batch processes, there is a comparison 641 of the overall measured execution time of the scheduling plan to a predetermined maximum overall time of execution of the scheduling plan and / or to the comparison of the measured execution times of each of the batch processes to predetermined maximum times for executing these batch processes.

In the event of an overrun (NOK), the verification step can be followed by a step 900 of increasing the quantity of allocated resources which will be detailed below.

In the absence of an overrun (OK), the verification step can be followed by a step 642 15 for calculating the difference (ie the margin) between the overall execution time of the scheduling plan and the maximum overall time predetermined execution of the scheduling plan, then a step of comparing this difference to a predetermined performance threshold so as to determine whether the new sizing plan can be validated 650 or whether a step 800 of reducing the quantity of allocated resources 20 must be executed in the event that the predetermined performance threshold is exceeded (NOK).

The validation 650 can for example comprise the generation of a report comprising the resource quantity values of the new optimized and validated sizing plan.

For example, the step of comparing this difference may consist in determining whether the difference is not significant (NOK), that is to say that it is for example less than 10% of the value of the predetermined maximum overall duration of execution of the scheduling plan.

Then there can be a 650 validation of the optimized sizing plan.

When the difference is significant (OK), that is to say that it is greater than a predetermined threshold such as for example greater than, for example, 10% of the value of the predetermined maximum overall duration of execution of the plan scheduling, preferably greater than 20% and more preferably greater than 30% then the verification step can be followed by a step 700 of modifying the optimized sizing plan so as to reduce the amount of resources allocated . 27

Thus, advantageously, the method according to the invention can include a step 700 of additional modification of the quantity of resources allocated.

Such a step is illustrated in Figure 8.

This additional modification 700 can advantageously occur in two situations: when the quantities of resources allocated were insufficient (eg the maximum overall execution time of the scheduling plan has been exceeded) and therefore it is necessary to increase the quantity of resources. allocated resources, and when the quantities of allocated resources were too high and this leads to a risk of wasting resources (eg between the overall duration of execution of the scheduling plan and the predetermined maximum overall duration of execution of the scheduling plan. scheduling has exceeded a predetermined performance threshold) and therefore it is necessary to reduce the amount of resources allocated.

This step includes in particular the generation 710 of new allocated resource quantity values which may be lower or higher than the previously allocated resource quantity values.

This generation 710 preferably takes into account the “steps” of each of the resources concerned.

This generation 710 is then followed by a step 720 for calculating new values of efficiency in the use of resources.

This calculation step 720 is a function of the extended execution times of the batch processing operations, of the weighting coefficients and in particular of the margin calculated during step 642 (ie the difference between the overall execution time of the plan d scheduling and the predetermined maximum overall duration of execution of the scheduling plan).

This is because there may be side effects when reducing the amount of resources that can be used by batch processes and this should be taken into account when making the additional modification.

Because of this side effect, the measured (ie actual) execution times of the batch processes and the scheduling plan may differ from the calculated execution times and therefore the calculation of new use efficiency values. resources should preferably take this difference (ie margin) into account in order to provide more precise results.

There is then selection 730 of the new values of quantity of resources allocated 30 and then generation 740 of a new optimized sizing plan.

These new values of allocated resource quantities are those for which the overall execution time value of the scheduling plan is less than the predetermined maximum overall duration 28 of the scheduling plan.

Finally, this step 700 can end with the generation 740 of a new optimized sizing plan.

According to another aspect, the invention relates to a device 1 for optimizing the dimensioning of an IT infrastructure 2 for the execution of a scheduling plan.

This device is in particular capable of implementing a sizing optimization method according to the invention described above.

One embodiment of such a device is illustrated in Figure 9.

As illustrated, this device comprises in particular an acquisition module 10, a resource rarefaction module 20, a batch processing execution module 30, a metric collection module 40, and an optimization module 50.

It can also include a verification module 60, a sizing module 70, a communication module 80, and a storage module 90.

The acquisition module 10 according to the invention is in particular configured to acquire a first dimensioning plane.

The first dimensioning plan 15 comprises first values of quantity of allocated resources allowing the execution of a plurality of batch processes according to a scheduling plan.

In addition, it can be configured to acquire a first overall execution time of the scheduling plan, predetermined execution parameters of a scheduling plan, a predetermined weighting coefficient for each resource of a plurality of resources , preferably a plurality of critical resources.

The resource scarcity module 20 according to the invention is in particular configured to generate a controlled consumption of at least one resource on an IT infrastructure 2.

This consumption is preferably a function of rarefaction rules previously stored so as to leave available for a batch processing a quantity of available resources less than an initial quantity of resources.

Advantageously, the resource scarcity module 20 has the function of gradually reducing the availability of a critical resource.

Thus, it is possible to progressively increase the scarcity of the critical resource until a predetermined maximum execution time is exceeded for the batch processing studied.

The rarefaction module 20 can therefore be considered as creating a degraded environment for the studied batch process (s) because the rarefaction rules can control to what quantity of resources of the IT infrastructure 2 the studied batch processes access.

Preferably, the resource rarefaction module 20 is configured to generate a controlled consumption of at least one critical resource according to the rarefaction rules previously stored on a rarefaction repository 21.

There may be a consumption of a critical resource or a simultaneous consumption of several critical resources.

For example, the rarefaction of a resource can be based on all or part of the following data: an identifier of the component or of the resource to be disturbed: for example a network address; a value for the start date of the rarefaction, it may for example be an absolute date evaluated relative to an internal clock of the device producing the rarefaction; - a rarefaction duration value (e.g. number of seconds) or a value of 15 the rarefaction end date; a level of consumption of the starting resource, corresponding for example to the smallest quantity of resource to be consumed; - a final level of resource consumption, corresponding for example to the highest quantity of resource to be consumed 20 - a step of increase in consumption, corresponding for example to the difference between a first quantity of resource to be consumed and a following quantity , and - a predetermined duration between each increment step; and - an order of the different rarefactions to be carried out. 25

These values could for example be acquired by the resource rarefaction module 20 via a call to one or more resource rarefaction configuration files of the rarefaction repository 21 or else to instructions transmitted from an interface ( eg graphic).

Thus, the rarefaction module 20 can be configured to receive and take into account rarefaction rules.

These rarefaction rules are for example entered via a graphical interface and stored in a memory, for example a random access memory.

Certain critical resource scarcity rule data are advantageously entered during the implementation of the dimensioning optimization method according to the invention.

This is for example the case of the start date of the rarefaction and the identifier of the critical resource to be rarefied. 30

When this data is recorded, it is for example in a resource scarcity configuration file being, for example, an XML file, a CSV file, an INI file, or a JSON file.

A configuration file has one or more configuration records.

Alternatively, the rarefaction configuration file is saved in a database.

The rarefaction repository 21 coupled to the rarefaction module 20 allows the generation of codes, the execution of which leads to the implementation of the rarefaction (i.e. the consumption of resources).

These codes are either object codes, for example for the depletion of memory, or so-called "shell" scripts or shell scripts.

The codes can also be compilable codes, the compilation being carried out at the time of the setting up of the rarefaction.

As will be detailed below, the device is configured to carry out the scarcity of resource (s) in parallel with the execution of a batch process to be studied (or of several).

The rarefaction repository 21 is configured to store rules for the rarefaction of critical resources that can be used for a progressive rarefaction of critical resources.

Thus, this rarefaction repository 21 is configured to store instructions allowing the consumption and therefore the rarefaction of the resources available for the batch process (s).

Such resource consumption can be achieved according to the noise generation method described in document EP2975526. 20

The critical resources consumed via the implementation of the consumption rules can include (but are not limited to) a type of network, conditions associated with the network, an available transmission and / or reception bandwidth, a available power / energy, available processor power (eg, clock speed, number of cores, cache, etc.) and available amount of memory.

As mentioned, the scarcity is generated via a consumption of resources whose consumed part (i.e. used or occupied) is then no longer available for the batch processing whose execution time is measured.

The rarefaction repository 21 also makes it possible to define consumption rules for a progressive modification of the rarefaction of the resources of different virtual machines. 30

The batch processing execution module 30 according to the invention is in particular configured to execute each of the batch processing operations of the plurality of batch processing operations on the IT infrastructure 2.

It can be any hardware and software arrangement capable of allowing the execution of batch processing on the IT infrastructure 31 2.

The batch processing execution module 30 is in particular configured to execute the batch processing operations one after the other, for example during the calculation of the extended execution times or according to a scheduling plan (eg with batch processing operations). simultaneous) for example during the verification step. 5

The metrics collection module 40 according to the invention is in particular configured to measure extended execution times as a function of values of quantity of resources available for each of the batch processes so as to determine execution times. lengthened for the plurality of batch processes, based on values of quantities of available resources.

Preferably, the metrics collection module 40 is further configured to measure footprints of the consumption of resources by the batch processing operations.

For this, the module 40 for collecting metrics can use a probe 41 or a plurality of probes 41.

The probe (s) 41 make it possible to collect data on the execution times of the batch processing operations and / or the consumption of the resources 15 of the IT infrastructure 2.

The collection module 40 can call on a monitoring file comprising metrics collection rules.

These metrics collection rules can specify data that must be recorded during the execution of the application brick.

These measurements can for example be carried out using a “Nigel's Monitor” (Nmon) or “Performance Monitor” (Perfmon) type probe.

Nmon probes can be used, for example, to display CPU, memory, swap, network data, information on users, groups, storage media, kernel use, or the most consuming processes.

Perfmon type probes are used to measure the performance of an IT infrastructure.

The information collected 25 may for example correspond to the percentages of resource use, response times, processing times but also the status of the ports, the number of JDBC or JMS message queues, the occupancy rate of the system. of files, the operating rate of the garbage collector (for "garbage collector" in English) for J2EE applications (for "Java Enterprise Edition" in English). 30

These probes can be associated with each parameter (e.g. execution time, resource or functionalities) to return measurement or metric information, representing for example the execution of the batch processing and the operation of the IT infrastructure.

For each parameter, the probe (s) define a parameter identifier and a parameter value.

Parameter values can be monitored continuously or at configurable intervals to obtain information for each parameter over time.

This information can be stored in a memory.

In certain embodiments, the device comprises a man-machine interface making it possible to define the probes on each machine which feed back the metrics originating from the use of the resources of the machine.

The optimization module 50 according to the invention is in particular configured to select new values of quantity of resources allocated for at least one optimized sizing plan, for which the new value of use efficiency resource efficiency is improved relative to the first resource use efficiency value.

It can be any hardware and software arrangement capable of allowing the selection of new values of quantity of resources allocated for the new optimized sizing plan.

These values can be integrated into a file which can then be sent to a client.

Further, this optimization module 50 is also configured for calculating a first resource utilization efficiency value from the first allocated resource quantity values, predetermined weighting coefficients and the first overall duration. execution of the scheduling plan.

The optimization module 50 is also configured to define resource quantity values less than the first allocated resource quantity values.

Moreover, this optimization module 50 is configured to calculate, for each value of quantity of resources, a new value of efficiency of use of resources, said calculation being a function of the values of quantity of resources, predetermined weighting coefficients. and measured extended run times of batch jobs.

Finally, this optimization module 50 is also configured to select new values of quantity of resources allocated to at least one optimized dimensioning plan, said new values of quantity of allocated resources being selected from among the values of quantity of resources of which the new resource use efficiency value is improved over the first resource use efficiency value.

The verification module 60 according to the invention is in particular configured to compare measured execution parameters of the scheduling plan with predetermined execution parameters of the scheduling plan.

This verification module 60 thus makes it possible to ensure that the predetermined maximum durations 33 of execution of the batch processing operations or a predetermined maximum overall execution time of the scheduling plan are respected.

This module also makes it possible to respect the good progress of the execution of the batch processing operations on an infrastructure whose resources correspond to the dimensioning plan optimized for the realization of the scheduling plan (e.g. compliance with the execution rules).

In addition, in order to implement the comparison of the execution times, the verification module may include any physical hardware or software capable of enabling the production or the calculation of a superiority, inferiority or equality ratio or even any differences between the execution parameters, between a predetermined maximum overall execution time of the scheduling plan and a measured global execution time of the scheduling plan.

To this end, the verification module can also be configured to calculate a difference or a ratio with a predetermined performance threshold.

Furthermore, according to certain embodiments, a user can interact with the verification module in order to configure and indicate to the verification module execution rules and maximum execution times or a predetermined performance threshold.

This verification is also possible without generating a controlled consumption and by directly taking an infrastructure of the desired size (this is easy when using a VML for example)

The sizing module 70 according to the invention is in particular configured to generate a sizing plan.

The sizing plan defined by the sizing module is a function of the footprint of each batch process, the predetermined execution rules and the assigned resource quantity values.

It may be any hardware and software arrangement capable of allowing the generation of such a dimensioning plan.

The communication module 80 according to the invention is in particular configured to exchange data with third party devices.

The device 1 communicates with other computer devices thanks to this communication module 80.

The communication module also allows data to be transmitted over at least one communication network and may include wired or wireless communication.

Preferably, the communication is operated by means of a wireless protocol such as wifi, 3G, 4G, 30 and / or Bluetooth.

These data exchanges can take the form of sending and receiving files.

For example, the communication module can be configured to transmit a report comprising for a period a list of the resource quantity values of the new optimized and validated sizing plan. 34

The communication module 80 is further configured to allow communication with a remote terminal, including a client 3.

The client is generally any hardware and / or software.

The communication module 80 can also be configured to communicate with a man-machine interface. 5

[00100] In addition, the device according to the invention may include a storage module 90 capable of, preferably configured for, storing a plurality of frames of reference or of measured data.

For example, the storage module 90 is preferably configured to store the weighting coefficients of each of the resources considered.

For this, the storage module 90 can comprise any computer readable medium known in the art comprising, for example, a volatile memory, such as a static random access memory (SRAM) and a dynamic random access memory (DRAM). ), and / or non-volatile memory, such as read only memory, flash memories, hard disks, optical disks and magnetic tapes.

The storage module 90 may include a plurality of instructions or modules or applications to perform various functionalities.

Thus, the storage module 90 can implement routines, programs, or matrix type data structures.

Preferably, the storage module 90 can comprise a medium readable by a computer system in the form of a volatile memory, such as a random access memory (RAM) and / or a cache memory.

The storage module 90, like the other modules, can for example be connected with the other components of the device 1 via a bus and one or more data medium interfaces.

[00101] The storage module 90 can in particular be configured to store one of the identification data of critical resources used by the batch processing operations.

In this context, the storage module 90 can for example be configured to store identification data of the different virtual machines or partitions supporting the studied batch process (s).

The critical resource can for example be a random access memory but many other resources can be considered such as the reading / writing of a disk, the reading / writing of a network, or the rate of use of a memory.

Additionally, the resource repository can be configured to store the available amount of critical resources (e.g.

CPU number and power, amount of memory).

More particularly, a critical resource can be a resource whose average consumption is greater than 60% or a resource whose consumption has one or more peaks of consumption greater than 70% or a resource whose loss of 20% or more. availability causes a failure on the execution of the studied scheduling plan.

The device 1 according to the invention may also be able to identify which resources are of importance for the execution of the scheduling plan.

The storage module 90 can also be configured to store a repository of execution rules.

The execution rules repository is configured to store data corresponding to the rules applicable during the execution of the scheduling of batch processes.

Thus, a user has the possibility of recording in a memory execution rules that the scheduling plan must comply with.

The rules can for example relate to sequencing orders to be respected (eg processing A must be executed before processing B), schedules to be respected (eg processing C must start before 11 p.m. or processing D must be terminated. before 2h), resource consumption thresholds to be respected (eg the consumption of resource 1.1.1 cannot exceed 80%), etc.

Such rules can be stored in at least one stored matrix table corresponding to each processing Ti.

For example, in order to avoid the simultaneous execution, or overlap, of two batch processes, a row can be added to each processing matrix, said row no longer containing the consumption of a resource for the time slots but a row. identical value (for example 99) in the first column for the two or more processes concerned.

This makes it possible to indicate to the verification module 60 that, for example, the overlap 20 of the processing T1 and T2 is prohibited.

[00103] The different modules or frames of reference are distinct in FIG. 9 but the invention can provide various types of arrangement such as for example a single module combining all the functions described here.

Likewise, these means can be divided into several electronic cards or else grouped together on a single electronic card. 25

The device 1 according to the invention can be integrated into a computer system and thus be able to communicate with one or more external devices such as a keyboard, a pointing device, a display, or any device allowing a user interact with the device 1.

It should be understood that although not shown, other hardware and / or software components could be used in conjunction with a device 1.

Thus, in one embodiment of the present invention, the device 1 can be coupled to a human machine interface (HMI).

The HMI, as already discussed, can be used to allow the transmission of parameters to the devices or, conversely, to provide the user with the data values measured or calculated by the device.

Generally, the HMI is communicatively coupled with a processor and includes a user output interface and a user input interface.

The user output interface may include an audio display and output interface and various indicators such as visual indicators, audible indicators, and haptic indicators.

The user input interface may include a keyboard, mouse, or other cursor navigation module such as a touchscreen, a touchpad, a stylus input interface, and a microphone for inputting audible signals such as as user speech, data and commands that can be recognized by the processor.

[00105] In one embodiment of the present invention, the device 1 can be coupled to a communication interface, for example a network interface of the Ethernet, FiberChannel, Int iniBand type or any other device allowing the device 1 to communicate with one or more other computing devices.

[00106] Further, as will be appreciated by those skilled in the art, aspects of the present invention can be realized as a system, method or computer program product.

Accordingly, aspects of the present invention may take the form of an all-hardware embodiment, an all-software embodiment (including firmware, resident software, microcode, etc.), or a mode. of particular embodiment such as a "circuit", "module" or "system".

Further, aspects of the present invention may take the form of a computer program product embedded in one or more computer readable media having computer readable program code embedded thereon.

Any combination of one or more computer readable media can be used.

In the context of this document, a computer readable medium can be any tangible medium which can contain, or store a program for use by or in connection with an instruction execution system, apparatus or device.

A computer readable medium can be, for example, but not limited to an electronic, magnetic, optical, electromagnetic, infrared or semiconductor system, apparatus or device, or any suitable combination of the foregoing.

More specific examples (a non-exhaustive list) of the computer readable storage medium 30 would include: hard disk, random access memory (RAM).

Computer program code for performing operations for aspects of the present invention can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C ++ or the like, the "C" programming language or similar programming languages, a scripting language such as Perl, or similar languages, and / or functional languages such as Meta Language.

The program code can run entirely on a user's computer, partially on a user's computer, and partially on a remote computer, or entirely on the remote computer or server.

In the latter scenario, the remote computer can be connected to a user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN).

These computer program instructions can be stored on a computer readable medium 10 capable of controlling a computing device (eg computer, server ...), so that the instructions stored in the computer readable medium produce a device. computer configured to implement the invention.

Thus, the present invention makes it possible to automatically generate an optimized sizing plan allowing the execution of a scheduling plan with improvement in the efficiency of use of resources and while preferably respecting a maximum execution time.

Thus, resources that may be essential for other applications are only monopolized in small quantities if other more available resources can compensate for this reduction.

This, as well as all these advantages previously described therefore contribute to facilitate the management of an IT infrastructure 20 involved in the execution.

Claims (11)

CLAIMS 1. A method of generating an optimized sizing plan for improving the efficiency of use of the resources of an IT infrastructure involved in the execution of a plurality of batch processes according to a scheduling plan, said method comprising: - a step (100) of data acquisition, by an acquisition module (10), comprising the acquisition of a first dimensioning plane, of a predetermined weighting coefficient for each resource of a plurality resources and predetermined execution parameters of a scheduling plan, said first dimensioning plan comprising first values of quantity of allocated resources allowing the execution of a plurality of batch processes according to the scheduling plan on a duration corresponding to a first overall duration of execution of the scheduling plan; - a step of calculating (200) a first value of efficiency of use of resources, by an optimization module (50), from the first values of quantity of resources allocated, from the predetermined weighting coefficients and from the first overall duration of execution of the scheduling plan; a step of determining (300) extended execution times of the plurality of batch processes, as a function of values of quantities of available resources, comprising: the generation (310) on an IT infrastructure (2), by a resource rarefaction module (20), of a controlled consumption of at least one resource according to previously stored rarefaction rules so as to leave available for a batch processing a quantity of available resources less than a quantity of resources initial, o the execution (320) on the IT infrastructure (2), by a batch processing execution module (30), of each of the batch processes of the plurality of batch processes, and a step measuring (330), by a metrics collection module (40), extended execution times as a function of values of quantity of resources available for each of the batch processing operations; and - a step of generation (400), by the optimization module (50), of at least one optimized sizing plan, said step of generation comprising: 39 o the definition (410) of resource quantity values, less than the first allocated resource quantity values, the calculation (420), for each resource quantity value, of a new resource use efficiency value, said calculation being a function of the resource quantity values, predetermined weighting coefficients and measured extended execution times of the batch processes, and selecting (430) new resource quantity values assigned to the at least one optimized sizing plan, said new quantity IO values of allocated resources being selected from among the resource quantity values whose new resource utilization efficiency value is improved relative to the first efficiency value of resource use. 15 2. A method of generating a sizing plan optimized for improving the efficiency of use of resources according to claim 1, characterized in that the step of generating (400) at least one sizing plan optimized further comprises a calculation (425) of execution parameters of the scheduling plan as a function of resource utilization efficiency values and in that the new allocated resource quantity values are also selected from among the values. quantity of resources whose calculated execution parameters comply with predetermined execution parameters of the scheduling plan. 3. Method for generating an optimized sizing plan for improving the efficiency of use of resources according to one of claims 1 or 2, characterized in that the first sizing plan is obtained according to a step. for generating a dimensioning plan comprising: - a step (110) of data acquisition, by an acquisition module (10), comprising the acquisition of predetermined execution rules, and of a footprint 30 on a plurality resources of each batch process; - a calculation step (120), by a sizing module (70), comprising, as a function of the footprint on a plurality of resources of each batch process and according to a scheduling plan respecting the predetermined execution rules , the calculation of maximum values of resource consumption per period by the batch processing, and a step of defining (130) a first dimensioning plan of an IT infrastructure (2), by the dimensioning module (70) , as a function of the maximum values of resource consumption per calculated periods, said first dimensioning plan comprising first values of quantity of allocated resources. 4. A method of generating an optimized sizing plan for improving the efficiency of use of resources according to any one of claims 1 to 3, characterized in that it comprises the measurement (101), by the module (40) for collecting 10 metrics, an imprint of each of the batch processes and the identification (102), by the optimization module (50), of the critical resources for each of the batch processes. 5. A method of generating an optimized sizing plan for improving the efficiency of use of resources according to any one of claims 1 to 4, characterized in that the step of generating (400) comprises furthermore: calculating (415), for each generated quantity values of resources, a new overall execution time of the scheduling plan as a function of the measured extended execution times of the batch processes, and deleting (416) resource quantity values, for which the new overall execution time of the corresponding scheduling plan is greater than a predetermined maximum overall execution time of the scheduling plan. 25 6. A method of generating an optimized sizing plan for improving the efficiency of use of resources according to any one of claims 1 to 5, characterized in that during the generation step (400) , the method comprises generating (440) a plurality of optimized dimensioning plans. 30 7. A method of generating an optimized sizing plan for improving the efficiency of use of resources according to any one of claims 1 to 6, characterized in that it further comprises a verification step ( 600) of compliance with the scheduling plan, said step comprising: - the execution (620), by the batch processing execution module (30) and on an IT infrastructure whose available resources correspond to the 41 optimized sizing, of the plurality of batch processes according to the scheduling plan, the measurement (630), by the metrics collection module (40), of execution parameters of the scheduling plan, and the comparison (640 ), by a verification module (60), measured parameters of execution of the scheduling plan with predetermined parameters of execution of the scheduling plan. 8. A method of generating an optimized sizing plan for improving the efficiency of use of resources according to claim 7, characterized in that the step of verifying (600) compliance with the scheduling plan comprises beforehand the generation (610) on the IT infrastructure, by the resource scarcity module (20), of a controlled consumption of several resources according to the optimized sizing plan. 15 9. A method of generating an optimized sizing plan for improving the efficiency of use of resources according to one of claims 7 or 8, characterized in that the verification step (600) further comprises : - the calculation, by the verification module (60), of a difference between a predetermined maximum overall duration of execution of the scheduling plan and a measured overall duration of execution of the scheduling plan, - the comparison of the difference calculated, by the verification module (60), at a predetermined performance threshold, and - the validation, by the verification module (60), of the optimized sizing plan 25 when the predetermined performance threshold is not not exceeded. 10. A method of generating an optimized sizing plan for improving the efficiency of use of resources according to any one of claims 1 to 9, characterized in that it further comprises a modification step ( 700) of the optimized sizing plan if at least one of the following predetermined execution parameters of the scheduling plan is not respected: execution rule, predetermined maximum duration of execution of a batch process and overall duration predetermined maximum of the scheduling plan. 42 11. Device (1) for generating an optimized sizing plan for improving the efficiency of use of the resources of an IT infrastructure involved in the execution of a plurality of batch processing according to a plan of. 'scheduling according to one of claims 1 to 10, characterized in that it comprises: - an acquisition module (10) configured to acquire a first dimensioning plane, a predetermined weighting coefficient for each resource of a plurality of resource and predetermined execution parameters of a scheduling plan, said first sizing plan comprising first allocated resource quantity values allowing the execution of a plurality of batch processes according to a scheduling plan , said execution taking place over a period corresponding to a first overall execution period of the scheduling plan; - a resource rarefaction module (20) configured to generate, on a computer infrastructure (2), a controlled consumption of at least one resource according to the rarefaction rules previously stored so as to make it available for processing by batches a quantity of available resources less than an initial quantity of resources, - a batch processing execution module (30) configured to execute, on the IT infrastructure (2), each of the batch processes of the plurality of 20 batch processes, and - a metrics collection module (40) configured to measure extended execution times as a function of values of the quantity of resources available for each of the batch processes, - an optimization module (50) configured to calculate a first resource utilization efficiency value from the first allocated resource amount values, predetermined weight coefficients and of the first overall execution time of the scheduling plan, and configured to define resource quantity values, less than the first allocated resource quantity values; Calculate, for each quantity value of 30 resources, a new resource use efficiency value, said calculation being a function of the resource quantity values, the predetermined weighting coefficients and the measured extended execution times of the treatments by lots; and selecting new values of quantity of resources allocated to the at least one optimized sizing plan, said new values of quantity of allocated resources being selected from the quantity values of 43 resources of which the new value of utilization efficiency. resource efficiency is improved relative to the first resource use efficiency value.