Dynamic integrity monitoring of container runtime environment executing on host
By creating instance-specific and host-specific integrity rules in the orchestration device, the resource allocation of container instances is dynamically monitored, solving the problem of the inability to identify dynamic changes in container instances in the prior art, and realizing effective integrity protection of container instances in industrial environments.
Patent Information
- Application Number
- CN202380029171.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2022-03-21
- Filing Date
- 2023-03-08
- Publication Date
- 2024-11-08
AI Technical Summary
Existing integrity monitoring systems cannot effectively identify the dynamic startup and shutdown processes of container instances initiated by the orchestrator, especially in industrial settings where container instances run briefly and attackers may have unauthorized access to nodes, making it impossible to identify and prevent illegal operations in a timely manner.
By creating instance-specific integrity rules and host-specific integrity criteria in the orchestrator, the resource allocation of container instances is dynamically monitored, these rules are protected with digital signatures, and these criteria are updated when resource allocation changes, ensuring that the integrity of container instances conforms to the orchestrator's predetermined configuration.
It enables dynamic integrity monitoring of container instances, which can promptly identify and alert on illegal operations, protect the integrity of the host, and prevent attackers from tampering with resource configurations.
Smart Images

Figure CN118922829A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a method for dynamic integrity monitoring of a container runtime environment executed on a host, wherein at least one container is executed in the container runtime environment and managed by an orchestration device. Background Art
[0002] Container virtualization is a method in which multiple instances of an operating system can use the operating system kernel of a host machine in isolation from one another. A software container (hereinafter referred to as container) therefore represents a lightweight virtualization of a runtime environment on a host machine (also called host system) and encapsulates a software application running in the container from the underlying host. By using container technology, software applications are now being implemented in many areas, for example in industrial automation and process control, but also in transportation systems or building automation.
[0003] In order to start a container on a host, a container image is required, which contains the binary programs and libraries required by the application software in addition to the application software. A container, or more precisely, a container instance, is created on the host from the container image and executed in the host's runtime environment.
[0004] The orchestrated runtime environment includes an orchestrator and at least one host (usually multiple hosts, also called nodes) that have been assigned to the orchestrator. The orchestrator starts, manages, and terminates container instances on the assigned hosts. Typical orchestrated runtime environments are Kubernetes-based container environments that either manage cloud-based container-as-a-service environments or manage virtual instances running in the cloud as nodes of the orchestrated runtime environment.
[0005] In addition to the use of orchestrated runtime environments in the IT sector, these runtime environments are also used in the field of the Industrial Internet of Things (IoT). In contrast to the typical Kubernetes scenario described above, it is not uncommon for devices running as nodes of an orchestrated runtime environment in the industrial sector that container instances run on the device briefly, usually for a maximum of 24 hours, but rather run without interruption for a relatively long period of time, i.e. from days to weeks. If an attacker gains unauthorized access to a node, the attacker could install additional software on the node managed by the central orchestration unit or could modify the node. In particular, in such an environment, container instances can be started in particular in parallel with instances managed by the orchestration component, without the orchestrator being aware of this or being able to prevent this process.
[0006] Known integrity monitoring systems are not optimally able to identify the dynamic start and stop processes of individual container instances initiated by the orchestrator, but rather do so based on a static set of rules that can be extended as necessary by querying orchestrator-specific deployment information. To create a static set of rules, all application use cases must be known, or, since such cases are rare, the rule set must be defined on a very large scale. Summary of the invention
[0007] Therefore, the object of the present invention is to create a method which is able to automatically and according to the runtime configuration defined by the orchestrator perform integrity monitoring of nodes and thus is able to perform a comparison on the target node: whether the instance was actually started by the orchestrator and whether the runtime configuration pre-given by the orchestrator of all instances running on the node is identical to the specifications of the orchestrator.
[0008] This object is achieved by the measures described in the independent claim. Advantageous developments of the invention are specified in the dependent claims.
[0009] According to a first aspect, the present invention relates to a method for dynamic integrity monitoring of a container runtime environment executed on a host, wherein at least one container is executed in the container runtime environment and managed by an orchestration device, the method comprising the following steps:
[0010] In the arrangement device,
[0011] - when launching a container instance in a container runtime environment by an orchestration device, creating an instance-specific integrity rule associated with at least one resource of a host machine that has been allocated for executing the container instance,
[0012] - adding the instance-specific integrity rule to the host-specific integrity criteria, the host-specific integrity criteria comprising the instance-specific integrity rule for each container instance executed in the container runtime environment of the host and arranged in the orchestration device,
[0013] - communicate host-specific integrity criteria to the host, where
[0014] Checking resources allocated by container instances on the host against host-specific integrity criteria, and
[0015] If a violation of a host-specific integrity criterion is detected during the check, an alarm message is output.
[0016] The instance-specific integrity rules include one or more integrity rules that are respectively related to one or more resources of the host that are assigned to the container instance. The host-specific integrity criteria relate to the entire host, that is, to all container instances executed in the runtime environment and the resources of the host that are managed for this purpose. The host-specific integrity criteria are created for each host managed by the orchestration device, for example, in the process of assigning the host to the orchestration unit or in the process of starting a container instance on the host for the first time. At each further start of a container instance on the host initiated by the orchestration device, the host-specific integrity criteria are updated and are therefore dynamically updated. By transmitting and checking the host-specific integrity criteria in the host, container instances that are not started by the orchestration device or resources that deviate from the configuration of the started container instance can be identified throughout the runtime of the container instance. By integrating instance-specific integrity rules, integrity monitoring can also be performed on the host itself.
[0017] In an advantageous embodiment, the method comprises the further steps of:
[0018] - receive notification in the orchestration if one of the container instances on the host is terminated, and
[0019] -Communicating updated host-specific integrity criteria regarding the notification to the host.
[0020] Thus, host-specific integrity criteria are updated not only when termination of a container instance is initiated by the orchestration device, but also when the termination is initiated by the host or due to a crash.
[0021] In an advantageous embodiment, if an event is detected in the orchestration device that changes the composition and / or configuration of the container instance on the host, the host-specific integrity criteria are updated and the updated host-specific integrity criteria are transmitted to the host.
[0022] Therefore, the orchestration device adapts the host-specific integrity criteria in the event of a change in the composition (i.e., which container instances and how many container instances), but also in the event of a change in the configuration of the container instances running on the host, such as the allocated resources.
[0023] In a preferred embodiment, the event is at least one of the following actions: failure of a container instance on the host, starting a new container instance on the host, updating a container instance based on a changed container image, identifying an application crash, and restarting the container instance involved on the host.
[0024] In an advantageous embodiment, instance-specific integrity rules are created in a manner that depends on information related to the container image from which the container instance is built.
[0025] By means of the instance-specific integrity rules, integrity protection can be performed for the host itself. An advantage of instance-specific integrity rules that rely on information related to the container image is that the integrity rules can be simply copied and reused for other container instances of the same container image.
[0026] In an advantageous embodiment, the instance-specific integrity rules are created in a manner that depends on information from provisioning information that specifies the configuration of the container instance when the container instance is started in the corresponding container runtime environment.
[0027] This is especially advantageous in cases where the configuration of a service defined within a container image or the target of a write operation is defined outside the container image and in particular via provisioning specific variables or configuration files. A specific example is provisioning an ad server that is used multiple times in different contexts within a micro-segmented application. For example, there are container instances of externally exposed front-end components and non-exposed container instances of the back-end that supply information for the entire application and, for example, provision performance data related to control facilities that are displayed on the front-end.
[0028] In an advantageous embodiment, the host-specific integrity criteria and / or the individual instance-specific integrity rules contained therein are protected by a digital signature of the orchestration device.
[0029] The host knows the relevant public key of the orchestration device and can therefore check the integrity of the criteria or rules and the device that signed, i.e. the orchestration unit. It is therefore possible to identify criteria or rules that have been modified by an attacker. The orchestration unit that signed can be verified.
[0030] In an advantageous embodiment, the checking of the host-specific integrity criteria relates to resources of the file system which are assigned to the container instance and / or to network connection resources and / or process resources and / or executable program files.
[0031] Therefore, check all resource types required to execute the container instance on the host.
[0032] In an advantageous embodiment, the checking of at least part of the host-specific integrity criteria is performed outside the host.
[0033] This enables easy inspection of resources used by container instances that have been moved from the host, such as storage capacity that has been moved.
[0034] In an advantageous embodiment, the alarm message is forwarded to a central alarm issuing device and the alarm action is executed by the alarm issuing device according to the alarm issuing criterion.
[0035] Therefore, the alarm action can be easily configured and assigned uniformly to all connected hosts.
[0036] In an advantageous embodiment, the host-specific integrity criterion has a predefined maximum validity period and / or the host-specific integrity criterion is updated at cyclical time intervals.
[0037] Thus, an attacker can be prevented from obtaining old host-specific integrity criteria and can be prohibited from making further updates.
[0038] According to a second aspect, the present invention relates to a system for dynamic integrity monitoring of a container runtime environment executed on a host, in which at least one container instance is executed and managed by an orchestration device, the system comprising:
[0039] The host includes a container runtime environment, and
[0040] a receiving unit designed to receive host-specific integrity criteria,
[0041] an integrity detection unit, which is designed to check resources assigned to at least one container instance with respect to host-specific integrity criteria, and
[0042] an output unit which is designed to output a warning message if a violation of a host-specific integrity criterion is detected during the checking process; and
[0043] The arrangement device comprises:
[0044] An integrity unit, the integrity unit being designed to,
[0045] - when launching a container instance in a container runtime environment by an orchestration device, creating an instance-specific integrity rule associated with at least one resource of a host machine that has been allocated for executing the container instance, and
[0046] - adding instance-specific integrity rules to host-specific integrity criteria, the host-specific integrity criteria comprising instance-specific integrity rules for each container instance executed in a container runtime environment of the host and arranged in the orchestration device, and
[0047] An output unit is designed to transmit the host-specific integrity criterion to the host.
[0048] According to a third aspect, the invention relates to a host comprising a container runtime environment, and
[0049] a receiving unit designed to receive host-specific integrity criteria,
[0050] an integrity detection unit, which is designed to check resources assigned to at least one container instance with respect to host-specific integrity criteria, and
[0051] An output unit is designed to output a warning message if a violation of a host-specific integrity criterion is detected during the checking process.
[0052] According to a fourth aspect, the present invention relates to an arrangement device, comprising:
[0053] An integrity unit, the integrity unit being designed to,
[0054] - when launching a container instance in a container runtime environment by an orchestration device, creating an instance-specific integrity rule associated with at least one resource of a host machine that has been allocated for executing the container instance, and
[0055] - adding instance-specific integrity rules to host-specific integrity criteria, the host-specific integrity criteria comprising instance-specific integrity rules for each container instance executed in a container runtime environment of the host and arranged in the orchestration device, and
[0056] An orchestration interface is designed to transmit host-specific integrity criteria to the host.
[0057] The system as well as the host and the orchestration device are designed to carry out the described method.
[0058] The system, the orchestration device and the host have the same advantages as the method.
[0059] A fifth aspect of the invention relates to a computer program product comprising a non-volatile computer-readable medium which can be directly loaded into the memory of a digital computer, the non-volatile computer-readable medium comprising a program code portion which, during execution of the program code portion by the digital computer, causes the digital computer to perform the steps of the method.
[0060] Unless otherwise stated in the description below, the terms "start", "create", "add", "transfer", "check", "output" etc. preferably relate to actions and / or processes and / or processing steps of changing and / or generating data and / or converting data into other data, wherein the data can in particular be represented or exist as physical parameters, for example as electrical pulses.
[0061] The system and the components optionally included therein (such as the orchestration device and the host etc.) may include one or more processors. The processor may be in particular a main processor (Central Processing Unit, CPU), a microprocessor or a microcontroller, such as an application specific integrated circuit or a digital signal processor, possibly in combination with a memory unit for storing program instructions etc.
[0062] The computer program product (eg computer program means) can be provided or delivered, for example, as a storage medium, such as a memory card, USB stick, CD-ROM, DVD or also in the form of a downloadable file from a server in a network. BRIEF DESCRIPTION OF THE DRAWINGS
[0063] Exemplary embodiments of the method according to the invention and of the device according to the invention are shown in the drawings by way of example and will be explained in more detail with reference to the following description, in which:
[0064] Figure 1 An embodiment of the method according to the invention is shown as a flow chart;
[0065] Figure 2 An embodiment of a system according to the invention is shown schematically;
[0066] Figure 3 An embodiment of the arrangement device according to the present invention is shown in a schematic diagram; and
[0067] Figure 4 An embodiment of a host according to the present invention is shown in a schematic diagram. DETAILED DESCRIPTION
[0068] In all the Figures, parts which correspond to one another are provided with the same reference symbols.
[0069] Known existing integrity monitoring systems are not able to optimally identify the dynamic start and stop processes of a single container instance initiated by the orchestrator, but rather perform the identification based on a static set of rules, which can be extended if necessary by querying orchestrator-specific deployment information. The method is used to monitor operations or file system changes on orchestrated nodes, but works independently of the predetermined runtime configuration of the orchestrator. Therefore, existing instances that were not started by the orchestrator can only be identified if the anomaly identification component obtains access rights to, for example, orchestrator metadata at runtime. If the connection between the orchestrator and the node is interrupted, or an attacker is able to generate the same deployment information for a container instance without an orchestrator, this cannot be identified using existing methods. The same applies to possible licensed application use cases, in which it should be ensured, for example, that containers for applications should only be installed through the orchestrator and not manually on the device.
[0070] The basic idea of the present invention is that, in addition to orchestrating the container instances to be run on the nodes, the orchestration device or the management system of one or more nodes also dynamically creates device-specific integrity criteria (in addition to the provisioning information) according to the runtime configuration predefined by the orchestration device, and optionally protects the execution integrity of these integrity criteria by means of signatures. In the following, the nodes managed by the orchestration device are referred to as hosts.
[0071] refer to Figure 1 The method for dynamic integrity monitoring of a container runtime environment executed on a host according to the present invention will now be described more precisely, with reference to the flowchart in FIG. 1 , wherein at least one container is executed in the container runtime environment and managed by an orchestration device. To this end, host-specific integrity criteria are created based on a single container instance executed on a device or in the container runtime environment of the host. The host-specific integrity criteria include a static part, which contains integrity rules for the overall operating system of the host and is extended by instance-specific aspects. Integrity monitoring is performed for the host itself or the operating system of the host by means of the static part of the host-specific integrity criteria.
[0072] For each container instance, instance-specific integrity rules are first defined separately. To this end, the following steps are performed by the orchestration device. In a first method step S1, during the process of starting a container instance in a container runtime environment by the orchestration device, an instance-specific integrity rule related to at least one resource of a host allocated for executing the container instance is created.
[0073] Instance-specific integrity rules are created in a manner that depends on information related to the container image from which the container instance is built. For example, the information related to the container image can be determined by meta-information related to the container image. Alternatively or additionally, instance-specific integrity rules can be created in a manner that depends on information from provisioning information, wherein the provisioning information predetermines the configuration of the container instance when the container instance is started in the corresponding container runtime environment. If, for example, the configuration of a service defined in the container image or the target of a write operation is defined outside the container image, a variant based on the provisioning information should be preferred. In both variants, multiple instance-specific integrity rules can be associated with each other by means of one or more digital signatures of the orchestration device. The host-specific integrity criteria are protected by the digital signature of the orchestration device.
[0074] Subsequently, at least one instance-specific integrity rule is added to the host-specific integrity criteria, see step S2. The host-specific integrity criteria are arranged in the orchestration device and include at least one instance-specific integrity rule for each container instance executed in the container runtime environment of the host. The host-specific integrity criteria are generated in the orchestration device, for example, during the process of assigning the host to the orchestration device or during the process of starting the first container instance on the host. The host-specific integrity criteria are transmitted from the orchestration device to the host, see S3 and the container.
[0075] In the process of creating the dynamic part of the host-specific integrity criteria, the orchestration device ensures that the instance-specific integrity rules predefined by the container image and / or by the provisioning information are only relevant to the resources defined in the container instance or in the provisioning information, and not to the conditions of the host itself or other deployed resources, i.e., applications executed by the container instance on the same host. The reason for this is that the creator of the integrity rule set should only create integrity rules or integrity criteria for his own applications. This can be achieved by checking whether the integrity rules associated with the deployment are only relevant to the deployed resources or to the image or the instance generated therefrom itself.
[0076] The host checks the resources allocated to the container instance executed on the host against the host-specific integrity criteria, see step S4. If the allocated resources do not comply with the specifications in the host-specific integrity criteria, an alarm message is output, see step S5. If the allocated resources comply with the specifications in the host-specific integrity criteria, the check is terminated without further action, see step S6. Optionally, the check, the time point of the check and the result of the check may be recorded and stored. This enables a temporal categorization of deviations between the actually allocated resources and the host-specific integrity criteria.
[0077] In order to create host-specific integrity monitoring criteria based on the container instance currently executing for the application, the orchestration device updates the host-specific integrity criteria every time a change is made to the host. To this end, when one of the container instances on the host is terminated, a notification from the host is received in the orchestration device. Subsequently, the host-specific integrity criteria is updated in the orchestration device based on the notification and transmitted to the host.
[0078] If one of the container instances is terminated or fails, the host actively sends a notification to the orchestration device. Alternatively or additionally, the orchestration device sends a query to the host regarding the status. If the orchestration device is managing multiple hosts, the orchestration device performs the steps using each of the managed hosts. The identified failure of one or more of the container instances may result in the orchestration device removing instance-specific integrity rules for the container instances that no longer exist from the host-specific integrity criteria, and generating and transmitting updated new host-specific integrity criteria to the host.
[0079] Once an event is detected in the orchestration device that changes the composition and / or configuration of the container instance on the host, the host-specific integrity criteria are updated and the updated host-specific integrity criteria are transmitted to the host. The relevant events that lead to the update of the host-specific integrity criteria include: starting at least one new container instance, for example, allocating more resources to the application by scaling (Up / Downscaling); importing at least one updated container image by adapting the allocation configuration at the orchestration device, if this allocation leads to changes on the required host; identifying an application crash and restarting it by the orchestration device, if necessary, restarting it on another host according to the scheduling policy.
[0080] In the integrity protection process, the dynamically generated host-specific integrity criteria should also have a maximum validity period and the host-specific integrity criteria should not only be updated in an event-related manner. The host-specific integrity criteria have a predetermined maximum validity period. Alternatively and / or additionally, the host-specific integrity criteria are updated at cyclic time intervals.
[0081] Figure 2An example of application of the method in a system 10 according to the present invention is shown, the system comprising an orchestration device 11 and a host 12. The system 10 may include other hosts managed by the orchestration device 11, here a host 13. The orchestration device 11 creates and updates host-specific integrity criteria for each individual host of these hosts 12, 13, such as a host-specific integrity criteria 15 for the host 12 and a host-specific integrity criteria 17 for the host 13. The host-specific integrity criteria 15 include instance-specific integrity rules 16 for each container instance 18 executed on the host 12.
[0082] The import of a container instance by a user 19 is described based on the system 10. After the orchestration device 11 selects a host 12 for executing the container instance, the orchestration device 11 obtains the instance identifier of the container instance 18 to be monitored from the container runtime environment on the host 12, and creates instance-specific integrity rules 16 associated with the container instance 18 according to the provisioning information of the container instance 18, and / or creates instance-specific integrity rules 16 associated with the container instance 18 according to the container image signature as the dynamic part of the host-specific integrity criteria 15. The orchestration device can also supplement the instance identifier by marking the identifier of the host. The dynamic part and the static part with the integrity rules for the host operating system constitute the host-specific integrity criteria, which are signed using the encryption key of the orchestration device 11. Subsequently, the host-specific integrity criteria 15 are transmitted to the host 12. In the next step, for example, the container instance 18 is started.
[0083] By means of the instance-specific assignment, that is to say the associated instance-specific integrity rules 16 , it is achieved that as soon as a container instance started outside of the orchestration device 11 is identified, an alarm can be issued.
[0084] The host-specific integrity criteria 15 are checked in relation to: resources of the file system assigned to the container instance 18, and / or network connection resources and / or process resources related to the host 12. For example, this method is performed by a "Catch-All" rule, in which all processes of unknown container instances trigger an alarm by default.
[0085] Thus, the integrity of the file system of the host 12 can be checked based on the device-specific integrity criteria 15. If the integrity of the file system is determined periodically, a reference database is usually determined for the files and directories to be monitored and their metadata. Alternatively, integrity violations can be determined directly with the help of corresponding dynamically generated audit rules or eBPF programs (extended Berkeley PacketFilter). If identification is performed by means of a reference database, integrity monitoring must be deactivated for each non-exclusive file system resource inside and outside the container instance 18. If the resource is no longer used because the container instance 18 is deleted, the container instance 18 fails or is terminated as planned, the status of the container instance is either notified by the host 12 directly through active notification or by periodic monitoring of the host 12 by the orchestration device 11. The orchestration device 11 can recalculate the defined records of the integrity database for the terminated container instance, and does not have to fully calculate the reference database.
[0086] The check of integrity rules for processes can monitor the integrity of processes started in the process namespace of each container instance and on the underlying host. In order to be able to perform integrity monitoring in a container instance-dependent and host-dependent manner, the integrity monitoring must identify whether a process belongs to a container instance. This is done with the help of at least one eBPF program or at least one kernel module, wherein the eBPF program or kernel module determines the generated system calls and their passed parameters.
[0087] Based on the integrity rules for network connections, it is possible to monitor for each instance: which network connections are expected and trigger an alarm when the guidelines are violated. Here, detection can also be performed through the corresponding eBPF program or kernel module, which detects specific system calls and analyzes their passed parameters.
[0088] Once an integrity violation is determined on a host 12, an alarm message is generated, which is transmitted to an alarm issuing device 14, which is preferably centrally arranged and processes alarm messages from multiple hosts 12, 13 or even for multiple orchestration devices 11. In the alarm issuing device, an alarm action is executed according to the alarm issuing criteria and forwarded to the orchestration device 11. According to the alarm action in the alarm issuing criteria, an operation is finally performed on the orchestration device 11. For example, for a container instance of a pre-given container type, the host 12 is excluded from scheduling.
[0089] Figure 3An embodiment of an orchestration device 20 according to the present invention is shown. The orchestration device 20 includes a user interface 21, through which a container instance from a user is received. The orchestration device 20 includes an orchestration interface 22, through which host-specific integrity criteria are transmitted to hosts managed by the orchestration device 20. The orchestration device 20 also includes a storage unit 24, in which the host-specific integrity criteria of the managed and integrity-monitored hosts are stored. The orchestration device 20 includes an integrity unit, in which instance-specific integrity rules are created for each container instance and inserted into the host-specific integrity criteria.
[0090] Figure 4 A host 30 is shown. The host 33 comprises an input unit designed to receive device-specific integrity criteria from the orchestration device. The host comprises an integrity identification unit 33 for checking and identifying integrity violations and for generating alarm messages. For various aspects, such as file system integrity monitoring, the reference database may be formed on a device that is separated from the host and that monitors the protected host from there by issuing commands.
[0091] The integrity detection unit 32 is designed to generate an alarm message and forward it via the output unit 21 to a central alarm output device.
[0092] If the device-specific integrity criteria are integrity-protected by signature, the integrity identification unit 33 running on the host can check whether the device-specific integrity criteria have been compromised. If the integrity identification unit 33 identifies a compromised device-specific integrity criterion, an alarm is also triggered. From this point in time, the entire host 30 is also considered to be compromised.
[0093] Therefore, for a managed host, the device-specific integrity criteria are dynamically adapted by the orchestration device each time a configuration change is made regarding a running container instance, and the device-specific integrity criteria relate to the entire host. By dynamically creating integrity criteria in a host-specific manner, the orchestration device is able to identify situations where, in addition to the container instances defined by it, container instances managed outside the orchestration device are also running. By means of the alert issuance criteria, the orchestration device can exclude hosts identified as compromised from the scheduling of defined container instances or all container instances.
[0094] All method steps can be implemented by corresponding devices suitable for performing the corresponding method steps. All functions that can be performed by specific features can be method steps of the method. Within the scope of the present invention, all described and / or depicted features can be advantageously combined with each other. The present invention is not limited to the described embodiments.
Claims
1. A method for dynamic integrity monitoring of a container runtime environment executed on a host (12), wherein: In the container runtime environment, at least one container instance (18) is executed and managed by an orchestration device (11), and the method comprises the following steps: In the arrangement device (11), - when launching a container instance (18) in a container runtime environment by the orchestration device (11), creating (S1) an instance-specific integrity rule (16) associated with at least one resource of the host (12), wherein the resource has been allocated for executing the container instance (18), - adding (S2) instance-specific integrity rules (16) to host-specific integrity criteria (15), the host-specific integrity criteria comprising instance-specific integrity rules (16) for each container instance executed in the container runtime environment of the host (12) and arranged in the orchestration device (11), - transmitting (S3) said host-specific integrity criteria (15) to said host (12), in which, checking (S4) resources allocated on the host (12) by a container instance (18) against the host-specific integrity criteria (15), and If a violation of the host-specific integrity criterion (15) is detected during the checking process, a warning message is output (S5).
2. The method according to claim 1, comprising the additional method step of: If one of the container instances (18) on the host (12) is terminated, a notification is received in the orchestration device (11), and - transmitting host-specific integrity criteria (15) updated in relation to said notification to the host (12).
3. The method according to claim 1 or 2, wherein: If an event is identified in the orchestration device (11) that causes a change in the composition and / or configuration of the container instance (18) on the host (12), the host-specific integrity criterion (15) is updated and the updated host-specific integrity criterion (15) is transmitted to the host (12).
4. The method according to claim 3, wherein the event is at least one of the following actions: failure of a container instance on the host (12), starting a new container instance (18) on the host (12), updating the container instance (18) according to a changed container image, identifying an application crash and restarting the container instance (18) involved on the host (12).
5. The method according to any of the preceding claims, wherein the instance-specific integrity rules (16) are created in a manner that depends on information related to a container image from which the container instance (18) is built.
6. The method according to claim 5, wherein: The instance-specific integrity rules (16) are created in a manner dependent on information from provisioning information, wherein the provisioning information predetermines the configuration of the container instance (18) when the container instance (18) is started in a corresponding container runtime environment.
7. A method according to any of the preceding claims, wherein the host-specific integrity criteria (15) and / or the individual instance-specific integrity rules (16) contained therein are protected by a digital signature of the orchestration device (11).
8. The method according to any of the preceding claims, wherein the checking of the host-specific integrity criteria (15) relates to resources of the file system assigned to the container instance (18), and / or to network connection resources and / or process resources.
9. The method according to any of the preceding claims, wherein the checking of at least a part of the host-specific integrity criteria (15) is performed outside the host (12).
10. A method according to any one of the preceding claims, wherein: The alarm message is forwarded to a central alarm issuing device (14), and an alarm action is executed by the alarm issuing device (14) according to an alarm issuing criterion.
11. The method according to claim 1, wherein the host-specific integrity criterion (15) has a predetermined maximum validity period and / or the host-specific integrity criterion (15) is updated at cyclical time intervals.
12. A system (10) for dynamic integrity monitoring of a container runtime environment executed on a host (12, 30), wherein at least one container instance (18) is executed and managed by an orchestration device (11), the system comprising: The arrangement device (11, 20), the arrangement device comprises: An integrity unit (23) designed to: - when launching a container instance (18) in a container runtime environment by the orchestration device (11, 20), creating an instance-specific integrity rule (15) associated with at least one resource of the host (12, 30) that has been allocated for executing the container instance (18), and - adding instance-specific integrity rules (16) to host-specific integrity criteria (15), the host-specific integrity criteria (15) comprising the instance-specific integrity rules (16) for each container instance (18) executed in the container runtime environment of the host (12, 30) and arranged in the orchestration device (11, 20), and an orchestration interface (22) designed to transmit the host-specific integrity criterion (15) to the host (12, 20); and The host (12, 20), the host including a container runtime environment, and a receiving unit (31) designed to receive host-specific integrity criteria (15), an integrity detection unit (33) designed to check resources assigned to at least one container instance (18) with respect to the host-specific integrity criterion (15), and An output unit (32) is designed to output a warning message if a violation of the host-specific integrity criterion is detected during the checking process.
13. A host (12, 30), the host comprising a container runtime environment, and a receiving unit (31) designed to receive host-specific integrity criteria (15), an integrity detection unit (33) designed to check resources assigned to at least one container instance (18) with respect to the host-specific integrity criterion (15), and An output unit (32) is designed to output a warning message if a violation of the host-specific integrity criterion (15) is detected during the checking process.
14. A layout device (11, 20), comprising: Integrity unit (23), the integrity unit is designed to - when launching a container instance (18) in a container runtime environment by the orchestration device (11, 20), creating an instance-specific integrity rule (16) associated with at least one resource of the host (12), wherein the resource has been allocated for executing the container instance (18), and - adding the instance-specific integrity rule (16) to a host-specific integrity criterion (15), the host-specific integrity criterion (15) comprising the instance-specific integrity rule (16) for each container instance (18) executed in the container runtime environment of the host (12, 30) and arranged in the orchestration device (11, 20), and An orchestration interface (22) is designed to transmit the host-specific integrity criterion (15) to the host (12, 30).
15. A computer program product comprising a non-volatile computer-readable medium, which can be directly loaded into the memory of at least one digital computer, and the non-volatile computer-readable medium includes program code portions, which, during the execution of the program code portions by the at least one digital computer, cause the at least one digital computer to perform the steps of the method according to any one of claims 1 to 11.
