EP3999962A1 - Procédé de protection contre les surcharges dans un appareil informatique virtualisé en conteneur - Google Patents

Procédé de protection contre les surcharges dans un appareil informatique virtualisé en conteneur

Info

Publication number
EP3999962A1
EP3999962A1 EP20771467.6A EP20771467A EP3999962A1 EP 3999962 A1 EP3999962 A1 EP 3999962A1 EP 20771467 A EP20771467 A EP 20771467A EP 3999962 A1 EP3999962 A1 EP 3999962A1
Authority
EP
European Patent Office
Prior art keywords
load
container
request message
application
computing device
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
EP20771467.6A
Other languages
German (de)
English (en)
Inventor
Christian Knierim
Angela Schattleitner
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Siemens AG
Original Assignee
Siemens AG
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Siemens AG filed Critical Siemens AG
Publication of EP3999962A1 publication Critical patent/EP3999962A1/fr
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
    • G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
    • G06F9/505—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering the load
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5061—Partitioning or combining of resources
    • G06F9/5077—Logical partitioning of resources; Management or configuration of virtualized resources
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5083—Techniques for rebalancing the load in a distributed system
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network

Definitions

  • the invention relates to a method for overload protection in a container-virtualized computing device, which provides a computer-implemented application through at least one work container, and a corresponding computing device and a corresponding computer program product.
  • Computer-implemented services also referred to as applications in the following, are often provided by a server system.
  • a service user requests the service via a communication network, such as the Internet, from the server system.
  • the application is often run in an execution environment known as a virtual container on the server system.
  • Container virtualization is a method of operating several instances of an operating system in isolation from one another on a so-called host system. In contrast to virtualization using a hypervisor, the containers share a common operating system and only certain operating system parts are duplicated and made available separately to each container. This is why container virtualization is particularly resource-efficient. Containers ensure separate management of the resources allocated by the host system.
  • a container that provides a specific application contains all the resources required to run the application and can easily be transported and installed as a file.
  • Resources are, for example, the program code, a runtime module, system tools, and system libraries.
  • a container also provides a specific application.
  • An application is thus encapsulated in a container.
  • An application is often provided by a container pool comprising a plurality of containers.
  • the number of containers used for an application is usually highly scalable, so that in the event of a heavy load on the existing containers, new containers are dynamically activated that can provide a requested service or application.
  • the application requests for example http (hypertext transfer protocol) requests, are distributed to an increasing number of containers via a load balancer.
  • a maximum number of containers is usually specified, on the one hand to cover the costs for the provision of the application and to prevent damage to the Rechenvor direction or the underlying system.
  • request messages are usually not processed and, for example, an error message is sent to the user as a response to the request returned to.
  • An overload of the computing device generated by a large number of application requests also referred to as a denial of service DoS attack, is often deliberately generated by a malicious third party or so-called bots. The application is then no longer accessible by authorized users.
  • ABDELZAHER T F ET AL "Web content adaptation to improve Server overload behavior", COMPUTER NETWORKS, ELSEVIER, AMSTERDAM, NL, vol. 31, No. 11-16, May 17, 1999 (1999-05-17), pages 1563-1577, XP004304574, ISSN: 1389-1286, DOI:
  • Scalability is the term used to describe the method that, as the load increases, additional containers that provide the same application are generated and request messages for the application are distributed to the further container or containers and processed there.
  • Further measures in an overload situation are what is known as over-provisioning, which means that more containers are kept ready than are necessary for processing the usual number of inquiries. These can process the request messages in the event of unusual loads.
  • Another measure is the automated scaling up of resources with a very high upper limit, the merging of container pools with other pools or redirection to other pools, especially to clusters in a cloud environment.
  • measures are known that attempt an attack, in particular an attack through a lot of inquiries and a provoked rejection of inquiries, and not to process or redirect the inquiry messages classified as defective.
  • some of the request messages can be discarded without any further indication and not processed.
  • the object of the present invention is to provide an improved measure in overload situations in a container-virtualized computing device in which less authorized request messages for an application are rejected and the quality of service is thus improved.
  • the invention relates to a method for preventing overloads in a container-virtualized computing device that provides a computer-implemented application through at least one working container, comprising the following steps:
  • an inquiry message is sent in the event of an overload situation of the computing device, which is caused by the load limit value is specified, not discarded, but forwarded to a special load defense container, which enables the request to be processed in a resource-saving manner.
  • the provision of the application generates fewer resources than when processing by a conventional work container and more new request messages can be processed and the application can be made available at least in a modified form, even in overload situations. If the load limit is not reached, the request is forwarded unchanged to the working container.
  • a working container denotes a container in a container-virtualized computing device which provides an unmodified, normal processing form, that is to say a processing form in normal operation.
  • a load protection container provides the application in a modified form and includes functions that reduce the number of request messages or the resources required to provide the application. Such a function checks, for example, whether the request was sent by an authorized user and only forwards request messages from authorized users.
  • the load distribution device forwards the request message to a working container or to a load defense container depending on the check result.
  • Such a load distribution device can, for example, be a load balancer which is provided by software for orchestrating containers.
  • a load distribution device works very efficiently and can therefore check and forward a large number of request messages in a short time.
  • forwarding is carried out in a work container. This has the advantage that existing load distribution facilities remain unchanged and only the process in the work containers has to be supplemented with the checking and forwarding function.
  • checking and distribution rules can be implemented in different ways in different work containers and can be easily implemented.
  • the checking and forwarding is carried out in a representative unit which is arranged in the working container.
  • checking and forwarding in the container are provided separately from the actual application. Direct influence on the application itself is avoided.
  • the work container can thus easily be supplemented with the checking and forwarding function.
  • the forwarding of the request message to a load protection container is carried out in a state-oriented manner, in particular by a state-oriented package inspection.
  • a request message is stateless, also referred to as stateless, distributed to any of the existing work containers from a pool of work containers. If an overload situation is recognized, i.e. the specified load limit value is exceeded, the request message and all other messages for the provision of the application are status-oriented, for example by a status-oriented package check, to the one existing load protection container or to a load protection container pool that contains several load protection containers, forwarded. This is applied either by a load distribution device or by the forwarding function in a Häcon container or in the representative unit within the Ar beitscontainers. This ensures that the modified form of processing of the application provided by the load defense container is carried out correctly in all steps and by the same load defense container.
  • the load limit value has different load type-specific limit values that are dependent on the type of load currently present on the computing device.
  • Different types of overload are, for example, reaching a maximum number of work containers, a sudden overload due to a high number of request messages per unit of time, or too many applications open in parallel.
  • Different types of overload can also be given by evaluating the error logs of the application that are operated in the container.
  • a load type-specific limit value can also be specified by a specified maximum response time to an incoming request message. It is thus possible to react flexibly to the different types of overload and to the disruption of the computing device associated with the corresponding overload.
  • the request message is forwarded to different load defense containers, which provide different, modified, load-reducing forms of processing of the application, depending on the type of current utilization.
  • load protection containers can advantageously be tailored or configured to suit a load type. If there is a high number of query messages per time unit, which for example indicate automatically generated query messages, a check can be carried out, for example, by switching on a challenge response method, for example by sending a captcha, which usually requires input by a human user. If the query is not answered correctly by the captcha, the query is considered malicious and will not be processed further.
  • a load protection container with a load-reducing processing variant could be used, in which computing power is saved, for example by dispensing with dynamically generated web content that heavily loads the computing device. For example, a response message is sent that shows, for example, a maintenance time with static content. If there are too many parallel requests that are not closed, the overload container could work with shorter timeouts.
  • the forwarding is additionally carried out depending on at least one of the following criteria, namely depending on a type of application requested, an identifier of a sender of the request message or a role of the user of the request message.
  • forwarding can be carried out dynamically and depending on the criteria, in particular to a load defense container.
  • Different load containers with specific load defense functions can thus be created and used depending on the criterion checked.
  • the only disadvantage here is that the distribution of the request messages requires further checks and the forwarding process therefore takes more time.
  • the request message is checked against a load distribution rule in the load defense container and, depending on the test result, at least one modified processing form of the application specified by the load distribution rule is carried out.
  • a check of the request message in a load protection container is only carried out if an overload has already been recognized. These are only a fraction of the query messages that are received in the computing device.
  • the request message is analyzed in the load defense container against an attack pattern, which is queried in particular in an attack database.
  • the attack patterns can be updated by means of data exchange with IT security service providers and attacks that occur worldwide can be reacted to quickly (taking into account the global threat landscape).
  • the request message is analyzed in the load defense container in addition to attack patterns against heuristics that are determined in advance by calibrating the system.
  • the huristics describes the usual behavior of inquirers, e.g. that significantly fewer inquiries are made outside of business hours.
  • attack signatures which are generally determined and made available via attack databases and, for example, continuously updated, can be used in the load defense container to analyze the request message and to identify attack patterns. If a corresponding attack pattern is recognized, the request message can be marked as harmful and countermeasures can be taken, for example the request cannot be processed and / or reported to a higher-level unit. All the indicators described for DoS attacks can be used by the container to calculate a DoS risk assessment (DoS risk score) for an inquiry.
  • DoS risk score can be made available to other services and applications via an interface that process or transmit the request.
  • one or more load protection containers are provided depending on the utilization of the computing device, in particular in addition to the existing working containers or as a replacement for an existing working container.
  • load defense containers that can be activated dynamically depending on the load on the computing device, a larger number of request messages can be processed without simply rejecting or not processing request messages and at least reduced processing forms of the application can be made available to the user.
  • the requested application can be made available to the user more frequently. Thus, the quality of service for the user is improved.
  • a second aspect of the invention relates to a computing device for container-virtualized provision of a computer-implemented application by at least one work container, having a receiving device that is designed to receive a request message for calling up the application, a checking unit that is designed in such a way to check the current utilization of the computing device against a predetermined load limit value and a forwarding unit, which is designed in this way, the request message to a load defense container, which provides at least one form of the application modified, in particular load-reducing, compared to the working container to forward when the current load is higher than the load limit value, the checking being carried out in a load distribution device (23) which is implemented via orchestration software, or the checking is carried out in the working container (24).
  • the computing device provides the application with improved quality of service with the same resources compared to a conventional computing device, since more request messages are processed and fewer request messages are discarded by authorized users.
  • control unit is designed in a load distribution device and the load distribution device forwards the request message to a working container or to a load protection container depending on the check result.
  • control unit is designed in a working container, in particular in the form of a representative unit.
  • the control device can be designed in each of the active work containers.
  • the load distribution device can be kept unchanged and thus optimized in terms of time.
  • the control device in the work container can be flexibly adapted to the application.
  • a third aspect of the invention relates to a computer program product comprising a computer-readable medium that can be loaded directly into a memory of a digital computer, comprising program code parts that are suitable for performing the steps of the method according to one of claims 1 to 11.
  • the terms “receive”, “check”, “forward” and the like preferably refer to actions and / or processes and / or processing steps (also: communication protocols) that change data and / or generate and / or convert the data into other data Ren, the data in particular being represented or being able to be present as physical quantities, for example as electrical impulses.
  • the term computing device denotes electronic devices with data processing properties and can be, for example, one or more servers, which are used in particular shared by several cloud users and are referred to as a cloud system, which can process data with computer support and which include at least one processor for data processing.
  • “computer-implemented application” can be understood to mean, for example, an implementation of the application in which, in particular, a processor executes at least one method step of the application
  • the control unit or the load distribution device can be implemented in terms of hardware and / or software.
  • the respective unit can be designed as a device or as part of a device, for example as a microprocessor.
  • the respective unit can be implemented as a computer program product, as a function, as a routine, as part of a program code or as an executable object.
  • a computer program product such as a computer program means, for example, can be provided or delivered as a storage medium or also in the form of a downloadable file from a server in a network.
  • 1 shows a first embodiment of the method according to the invention as a flow chart
  • 2 shows a second exemplary embodiment of the method according to the invention with a check in a load distribution device in a schematic representation
  • FIG. 3 shows a third embodiment of the method according to the invention with a check in a work container in a schematic representation
  • FIG. 4 shows a first exemplary embodiment of the computing device according to the invention in a block diagram
  • FIG. 5 shows a second exemplary embodiment of the computing device according to the invention in a block diagram.
  • a computer-implemented application is provided by a computing device which has a single or several computing units, such as processors or also server units.
  • the computing device is operated in a container-virtualized mode of operation.
  • An application is encapsulated by containers in which, in addition to a program for processing the application, operating system resources, runtime management, modules or even software libraries that are required to provide the application are encapsulated and executed as a closed system.
  • a work container usually executes a large number of instances of an application in a "normal" processing form. If new request messages can no longer be processed by a work container, a new work container is usually activated and made available for processing these further request messages.
  • Such an additional load-based activation of containers is also referred to as autoscaling. Is a maximum When the number of working containers is reached, further inquiries are usually rejected and answered, for example, with an error message to the sender. In order to reduce such rejection of admissible query messages, a method as shown in FIG. 1 is proposed.
  • step S1 a request message REQ for calling or starting a computer-implemented application is received in the computing device.
  • the currently present utilization L of the computing device is checked against a predetermined load limit value maxL. If this check determines that the current load L is higher than the load limit value maxL, the request message REQ is forwarded to a load defense container LC, which provides at least one load-reducing processing form of the application that is modified compared to the work container WC, see S3. If the load L is less than or equal to the load limit value maxL, the request message REQ is forwarded to a work container WC, see step S4, and made available by a conventional processing form of the application.
  • the load defense container LC answers the request message by, for example, carrying out additional measures to combat frequent request attacks, also referred to as (D) DoS attacks, at the application level.
  • measures are functions to recognize permissible request messages or to reject request messages recognized as inadmissible request messages.
  • measures are carried out, for example, before the actual application is executed and, together with the application itself, represent a modified, load-reducing form of processing of the application.
  • Such a measure is, for example, an upstream application of a question-answer procedure, also called a challenge-response procedure denotes before the actual execution of the application.
  • the computing device sends the inquiring device or user zer a question that the device or the user answers and is checked by the computing device.
  • the question-and-answer method can also be application-oriented, so that the application changes the question in a predetermined manner and sends it back to the application in the computing device. This can be, for example, solving a mathematical problem with the question as an input parameter.
  • the question-and-answer process can be transparent for the user of the application or it can also require user input. This is the case, for example, with a captcha, in which, for example, the number of a certain item on an image that is sent to the user is answered by the user.
  • a dynamically generated page content that represents a particularly high load for the computing device
  • only a static, predetermined page content can be transmitted.
  • a request message is only answered with specific information if this information is already available in the memory of the computing device, for example through a previous request.
  • a page content with a lower resolution can be sent back to the request message for display.
  • Another modified load-reducing form of processing of the application provided by the work container is achieved, for example, by at least partial evaluation of the request message. For example, an IP address of the sender, a role of the user or a priority identifier is determined and only request messages from authorized senders or roles or for prioritized services, which are specified, for example, by a positive list, are processed further. All other inquiries are not processed or rejected.
  • a role of the user can for example be predetermined by differently prioritized user classes, for example “preferred user”, “premium user”.
  • a role of the user can, however, also be “administrator”, “service provider” or the like.
  • a preferred processing of inquiry messages can also take place depending on geographical regions, so that inquiry messages are specifically discarded or not processed by spatially localizable attackers.
  • the request message is analyzed against an attack pattern in a load defense container and manipulated request messages are thus recognized before further processing.
  • a message identified as being manipulated for example a warning signal can be generated and / or the request message can not be processed further.
  • the request message is analyzed in a load protection container in relation to a heuristic, i.e. unusual behavior patterns in the case of requests.
  • manipulated request messages are recognized and classified as manipulated with a high degree of probability before further processing.
  • a warning signal can be generated and / or the request message can not be processed further.
  • the load limit value maxL can have different load type-specific limit values depending on the type of load currently present on the computing device.
  • a load type-specific limit value is, for example, a value for the maximum number of working containers configured for the service, a value for a number of query messages per second, a value for a number of parallel open query connections or applications running in parallel also be a value for a maximum number of error logs. Error logs are determined by evaluating errors in the application that are operated in a working container and are saved as error logs in the working container. Depending on a corresponding type of overload, a measure adapted to the situation can be taken. If, for example, the number of query messages per second is too high, a question-and-answer method is suitable for reducing this number.
  • the computing device could exploit asymmetrical resource consumption by means of restrictions on the backend requests, i.e. by creating static content instead of dynamic and resource-intensive content and sending it to the sender be sent to be a solution.
  • Such an evaluation can also be operated using an artificial intelligence security algorithm.
  • a load container LC provides all the modified, load-reducing embodiments of the application described.
  • load protection containers can also be provided which comprise a single or a subset of the possible modified, load-reducing processing forms of the application.
  • a request message can be used depending on the type of currently available load on a specific load defense container that provides a processing form that is optimized for the load type.
  • the forwarding can also be carried out depending on at least one further criterion.
  • the request message is at least partially analyzed when checking the load or when it is forwarded.
  • information in the header of the request message that contains information about the sender and the requested application is evaluated and, for example, depending on the requested application or depending on the identifier of the sender of the request message, such as its IP address, or depending on the Role of a user of the request message the Forwarding to different load protection containers carried out.
  • the request message can also be forwarded depending on a session identifier, also referred to as a session cookie, which is also contained in the header of the request message.
  • one or more load defense containers LC can be provided. Either a load protection container can be provided in addition to the existing work container or it can be started up as a replacement for an existing work container.
  • Fig. 2 shows an embodiment of the method in which to inquire messages 20, 21, 22 are checked by a Lastverannonsein direction 23 and are forwarded to a working container 24 or to a load defense container 25 depending on the verification result.
  • a work container 24 provides the application in an unmodified processing form 28.
  • several working containers 24 are provided in a working container pool 26.
  • Load defense containers 25 provide the application through the modified load-reduced processing form 29.
  • One or more load defense containers 25, 25.1, 25.2 can also be provided in a load defense container pool 27.
  • the load defense containers 25, 25.1, 25.2 can be in different forms.
  • the Darge presented load defense containers 25.1 and 25.2, for example, provide different modified processing forms for the application.
  • the load defense container 25 can for example comprise a variety of different modified processing forms.
  • the load defense container 25, 25.1, 25.2 can comprise different rules according to load distribution.
  • the distribution of the query messages is implemented by the load distribution device 23, which is typically implemented by so-called orchestration software, such as Kubernetes, for example.
  • orchestration software such as Kubernetes, for example.
  • the load distribution device 23 is aware of the utilization of the computing device in relation to the application.
  • the load distribution device 23 carries out the checking of the load as well as the forwarding depending on the result of the check.
  • the orchestration software must recognize the overload situation and redirect the request message to a container with built-in load protection functions, i.e. the modified, load-reducing processing forms 29 of the application.
  • the load protection containers 25, 25.1, 25.2 are also orchestrated in order to be able to scale them according to the load. If a load protection container 25 is started, a working container 24 is terminated, for example, in order to prevent resource scarcity. Alternatively, 25 special resources can be reserved for the load protection container.
  • FIG. 3 shows an embodiment of the method in which the checking and forwarding of request messages 20, 21, 22 is not carried out in a load distribution device 33, but in each of the load defense containers 34 of the load defense container pool 36.
  • a representative unit 31 is provided in each work container 34.
  • the representative unit 31 which is connected upstream of an unmodified processing form 28, can initiate the forwarding locally on the work container 34 by checking the local utilization of the corresponding work container 34.
  • the currently present utilization L can be made available to the representative device 31 by the orchestration software, that is to say in particular by the load distribution device 33. This available utilization is used to check the request message 20, 21, 22.
  • the forwarding In order to be able to carry out the forwarding of the request messages 20, 21, 22 to a load defense container, the forwarding must be carried out by a condition-based package inspection.
  • a packet is a data packet as defined in the Internet Protocol, for example.
  • request messages 20, 21 and subsequent connection for executing the application are shown. These inquiries messages 20, 21 are forwarded from the load distribution device 23 to any work container 24 without a standstill.
  • the condition-based package inspection is implemented in such a way that a distinction is only made between request messages 20, 21, which are forwarded to a working container 24, and request messages 22 and their subsequent messages, which are forwarded to a load defense container 25, 25.1, 25.2.
  • a condition-based package inspection on the individual work or load defense containers 24, 25 within a work container pool 26 or a load defense container pool 27 is not required. It can therefore be switched back and forth as desired between the individual containers in a pool.
  • the load distribution device 33 distributes the request messages 20, 21, 22 stateless and thus arbitrarily to the work containers 34 in the work container pool 36.
  • a representative unit 31 is installed in each container 34, which decides whether the request message 20, 21, 22 is forwarded to the unmodified processing form 28 of the application or to a load defense container 25.
  • the respective destination of the forwarding i.e. either the unmodified processing form 28 in the work container 34 or the modified processing form 29 in the load defense container 25, to which the request message is to be forwarded, must be communicated to each representative unit 31.
  • the forwarding of the request message 22 to a load defense container 25 is carried out in the representative unit 31 by means of a condition-based package inspection. It must be ensured that the load protection containers 25 are either kept permanently available or can be activated by the representative unit 31 or the orchestration software in the load distribution device 23.
  • FIG. 4 now shows a computing device 40 which provides an application through container virtualization.
  • One or more work containers 44 from a work container pool 46 and one or more load defense containers 45 from a load defense container pool 47 access a common operating system 48 provided by the computing device.
  • the computing device 40 further includes a receiving unit 41 and a load distribution device 43.
  • the load distribution device 43 includes a control unit 42 which is configured to carry out the checking and forwarding of the method according to the invention.
  • the Lastver dividing device 43 is implemented in particular via orchestration software.
  • the control device 42 comprises a checking unit and a forwarding unit, which are integrated, as shown, or are designed separately in two parts.
  • the control device 42 can be implemented in terms of software, for example as part of the orchestration software, or also in terms of hardware, for example one or more processors.
  • Fig. 5 shows a further embodiment of a computing device 50 which, according to the computing device 40, container-virtualized, executes a computer-implemented application using work containers 54 or load containers 55 from a work container pool 56 or a load protection container pool 57.
  • the working container pool 56 and the load protection container pool 57 access a common operating system 58.
  • the computing device 50 further comprises a receiving unit 51 and a load distribution device 53.
  • the load distribution device 53 forwards all received request messages from the receiving unit 51 to the work container pool 56.
  • Each request is rricht one of the work container 54 upstream Stellver treading unit 52 forwarded.
  • the representative unit 52 forwards the request message to a load defense container 55 in the load container pool 57.
  • a load defense by switching in load defense containers starts early when the load increases and can thus already ward off the load before the computing device or the working container become overloaded.

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)
  • Computer And Data Communications (AREA)
  • Information Transfer Between Computers (AREA)

Abstract

L'invention concerne un procédé de protection contre les surcharges dans un appareil informatique virtualisé en conteneur qui fournit une application mise en œuvre par ordinateur au moyen d'au moins un conteneur de travail, comprenant les étapes suivantes : - réception (S1) d'un message de demande (REQ) pour appeler l'application, - vérification (S2) de la charge de travail actuellement existante de l'appareil informatique par rapport à une valeur limite de charge stipulée, et – si la charge de travail actuelle est supérieure à la valeur limite de charge, transfert (S3) du message de demande (REQ) à un conteneur de rejet de charge (LC) qui fournit au moins une forme de manipulation de l'application qui est modifiée par rapport au conteneur de travail (WC).
EP20771467.6A 2019-09-17 2020-08-31 Procédé de protection contre les surcharges dans un appareil informatique virtualisé en conteneur Pending EP3999962A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
EP19197775.0A EP3796164A1 (fr) 2019-09-17 2019-09-17 Procédé de protection contre la surcharge dans un dispositif de calcul à instances virtualisées
PCT/EP2020/074226 WO2021052739A1 (fr) 2019-09-17 2020-08-31 Procédé de protection contre les surcharges dans un appareil informatique virtualisé en conteneur

Publications (1)

Publication Number Publication Date
EP3999962A1 true EP3999962A1 (fr) 2022-05-25

Family

ID=67997355

Family Applications (2)

Application Number Title Priority Date Filing Date
EP19197775.0A Withdrawn EP3796164A1 (fr) 2019-09-17 2019-09-17 Procédé de protection contre la surcharge dans un dispositif de calcul à instances virtualisées
EP20771467.6A Pending EP3999962A1 (fr) 2019-09-17 2020-08-31 Procédé de protection contre les surcharges dans un appareil informatique virtualisé en conteneur

Family Applications Before (1)

Application Number Title Priority Date Filing Date
EP19197775.0A Withdrawn EP3796164A1 (fr) 2019-09-17 2019-09-17 Procédé de protection contre la surcharge dans un dispositif de calcul à instances virtualisées

Country Status (4)

Country Link
US (1) US20220334877A1 (fr)
EP (2) EP3796164A1 (fr)
CN (1) CN114365089A (fr)
WO (1) WO2021052739A1 (fr)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11868287B2 (en) * 2020-12-17 2024-01-09 Micron Technology, Inc. Just-in-time (JIT) scheduler for memory subsystems
US11665106B2 (en) * 2021-09-07 2023-05-30 Hewlett Packard Enterprise Development Lp Network-aware resource allocation

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8392562B2 (en) * 2009-12-23 2013-03-05 Citrix Systems, Inc. Systems and methods for managing preferred client connectivity to servers via multi-core system
US10892942B2 (en) * 2016-01-22 2021-01-12 Equinix, Inc. Container-based cloud exchange disaster recovery
US10439987B2 (en) * 2017-06-12 2019-10-08 Ca, Inc. Systems and methods for securing network traffic flow in a multi-service containerized application

Also Published As

Publication number Publication date
EP3796164A1 (fr) 2021-03-24
WO2021052739A1 (fr) 2021-03-25
US20220334877A1 (en) 2022-10-20
CN114365089A (zh) 2022-04-15

Similar Documents

Publication Publication Date Title
US8539068B2 (en) Methods and systems for providing customized domain messages
CN108416665B (zh) 数据交互方法、装置、计算机设备和存储介质
DE102018204859A1 (de) Dynamischer Lastenausgleich in Netzschnittstellenkarten für eine optimale Leistung auf Systemebene
EP3796167B1 (fr) Gestion de routeur par un gestionnaire de grappe de traitement de flux d'événements
WO2021052739A1 (fr) Procédé de protection contre les surcharges dans un appareil informatique virtualisé en conteneur
CN110661815A (zh) 基于频率限制微服务的请求处理方法、装置、系统及介质
US8141151B2 (en) Non-intrusive monitoring of services in a service-oriented architecture
CN1692617A (zh) 客户机Web服务访问
CN109088909B (zh) 一种基于商户类型的服务灰度发布方法及设备
US20070289014A1 (en) Network security device and method for processing packet data using the same
CN104899482A (zh) 限制批量请求服务的方法和装置
CN114095258A (zh) 攻击防御方法、装置、电子设备及存储介质
CN111385328B (zh) 业务请求的处理方法、系统及电子设备
WO2017189811A1 (fr) Acheminement fondé sur des gtm multiples pour éviter des latences
EP3901807A1 (fr) Création d'une image de conteneur de système de nom de domaine pour créer une instance de conteneur de système de nom de domaine
CN112671945A (zh) 管理ip代理池的方法、装置、计算机设备及存储介质
US20150113655A1 (en) System, Method and Computer Program Product for Using Opinions Relating to Trustworthiness to Block or Allow Access
CN112311728A (zh) 主机攻陷判定方法、装置、计算设备及计算机存储介质
CN115580618A (zh) 一种负载均衡方法、装置、设备及介质
CN114513465A (zh) 负载均衡方法、负载均衡装置、电子设备和存储介质
CN113159614A (zh) 任务分配方法、装置、电子设备及计算机可读存储介质
WO2014128732A1 (fr) Procédé de génération d'identité de corrélation pour environnement en nuage
CN114816723A (zh) 一种负载均衡系统、方法及计算机可读存储介质
CN113377866A (zh) 一种虚拟化数据库代理服务的负载均衡方法及装置
CN114520808A (zh) 请求处理方法、装置、电子设备及计算机可读存储介质

Legal Events

Date Code Title Description
STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: UNKNOWN

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE

PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE

17P Request for examination filed

Effective date: 20220218

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

DAV Request for validation of the european patent (deleted)
DAX Request for extension of the european patent (deleted)