Detecting resource redundancy and conflicts in a heterogeneous computing environment

    公开(公告)号:US11593192B2

    公开(公告)日:2023-02-28

    申请号:US17302943

    申请日:2021-05-17

    Abstract: Detecting resource errors in a heterogeneous computing environment is provided. A plurality of individual resource dependency graphs corresponding to a plurality of computing systems that comprise the heterogeneous computing environment is consolidated to form a consolidated resource dependency graph. An analysis of respective nodes representing respective resources of the heterogeneous computing environment in the consolidated resource dependency graph is performed to identify a resource error caused by a new resource being added to a computing system of the plurality of computing systems based on defined rule sets. It is determined whether the new resource causes an error to sibling resources at a same level under a parent resource in the consolidated resource dependency graph based on the analysis. In response to determining that the new resource does not cause an error to the sibling resources, the new resource is deployed in the computing system of the heterogeneous computing environment.

    Automated port configuration management in a service mesh

    公开(公告)号:US11570279B1

    公开(公告)日:2023-01-31

    申请号:US17681046

    申请日:2022-02-25

    Abstract: Systems, methods and/or computer program products for managing and dynamically automating service mesh communications between microservices, eliminating unnecessary exposure of microservice ports and increasing security between microservices of the service mesh. The control plane collects data describing communications between microservices and tracks the frequency at which microservices communicate. Collected data is fed to machine learning models which outputs a forecast predicting future communication interactions between microservices. Using the predicted requirements for facilitating communications between microservices of the service mesh, an allowed list of communications can be generated describing the microservices allowed to send and receive communications, duration of communications allowed, when such communications are allowed, and the ports that will be used for facilitating the communication between microservices. Administrators of the service mesh may manually override the one or more approved aspects of the dynamically generated allowed list configured automatically by the service mesh.

    Dynamic management of network policies between microservices within a service mesh

    公开(公告)号:US11563636B1

    公开(公告)日:2023-01-24

    申请号:US17671949

    申请日:2022-02-15

    Abstract: Systems, methods and/or computer program products optimizing network policies between microservices of a service mesh. The service mesh tracks incoming API calls of applications and based on the historical transactions, the context of API calls, and the microservices in the microservice chain being invoked, network controls and policy configurations are set to optimize the transactions performed by the service mesh. Dimensions of the communications between microservices of the service mesh are dynamically optimized via the service mesh control plane using a policy optimizer. Optimized dimensions of service mesh transactions includes automated policy adjustments to retries between microservices, circuit breaking between microservices, automated timeout adjustments between microservices and intelligent rate limiting between microservices and/or rate limiting applied to user profiles.

    MICROSERVICE SYSTEM WITH GLOBAL CONTEXT CACHE

    公开(公告)号:US20220292032A1

    公开(公告)日:2022-09-15

    申请号:US17198458

    申请日:2021-03-11

    Abstract: Disclosed embodiments provide features for the architecture of microservices. A global context cache is created for a microservice environment that is accessible from multiple deployed microservices. Data from various customers/applications can be aggregated to establish a determination of when a read or write access would fail due to permissions, and/or other condition such as existence or non-existence of certain data. In such situations, an error can be returned from the global context cache in much less time than if the access request propagated throughout the computer network to the persistent storage. In this way, disclosed embodiments reduce downtime and save money for organizations, and increase the efficiency of utilization of computer resources

    Optimizing Insight Generation in Heterogeneous Datasets

    公开(公告)号:US20210263951A1

    公开(公告)日:2021-08-26

    申请号:US16796996

    申请日:2020-02-21

    Abstract: Embodiments relate to a system, computer program product, and method to merge two or more heterogeneous datasets. Seed attributes of each dataset that is the subject of the merge are identified. The seed attributes are derived from candidate attributes of the respective datasets. A correlation is assessed to create a set of mergeable attributes and a set of non-mergeable attributes. A cohesiveness characteristic is leveraged to iteratively identify one or more attributes from the set of non-mergeable attributes, and to amend the set of mergeable attributes with the one or more attributes identified in the set of non-mergeable attributes. A merged dataset based on the amended set of mergeable attributes and representing non-trivial similarities between the first and second dataset is formed as output.

    Management of container host clusters

    公开(公告)号:US10684872B2

    公开(公告)日:2020-06-16

    申请号:US15969929

    申请日:2018-05-03

    Abstract: A method, executed by a computer, includes receiving, for a cluster, a request to create a new container host, wherein the cluster comprises a plurality of container hosts having a corresponding plurality of local repositories, creating the new container host, selecting an image for the new container host according to one or more host preferences, wherein the image is stored in an image location, and copying the image from the image location to the local repository of the new container host. A computer system and computer program product corresponding to the above method are also disclosed herein.

    OPTIMIZING TIMEOUTS AND POLLING INTERVALS
    109.
    发明申请

    公开(公告)号:US20180196696A1

    公开(公告)日:2018-07-12

    申请号:US15399827

    申请日:2017-01-06

    Abstract: An approach is provided for managing a timeout and polling interval of an operation of an application. A type of the application is received. Based on the type, it is determined that the application requires optimizations of the timeout and polling interval. Whether a data repository includes recommendations of timeout(s) and polling interval(s) for operation(s) of applications of the received type is determined. If the data repository includes the recommendations, a recommendation that specifies the timeout and polling interval is selected from the recommendations and the timeout and polling interval are applied to a deployed image, or if the data repository does not include the recommendations, a user entry of values for an optimal timeout and an optimal polling interval is received and a workload is deployed. The workload has a configuration that includes the values for the optimal timeout and the optimal polling interval.

    SPLITTING AND MERGING MICROSERVICES
    110.
    发明申请

    公开(公告)号:US20170364434A1

    公开(公告)日:2017-12-21

    申请号:US15182718

    申请日:2016-06-15

    CPC classification number: G06F11/3668 G06F11/3692

    Abstract: A method, executed by a computer, includes receiving test data corresponding to a first microservice and a second microservice, determining, using the test data, whether the first microservice can be split, determining, using the test data, whether the first and second microservices can be merged, responsive to determining that the first microservice can be split, issuing a recommendation to split the first microservice, and responsive to determining that the first and second microservices can be merged, issuing a recommendation to merge the first microservice and the second microservice. A computer system and computer program product corresponding to the above method are also disclosed herein.

Patent Agency Ranking