User environment for exchanging edge computing system and high performance computing system

Abstract

The invention relates to a user environment for exchanging an edge computing system and a high performance computing system. A process includes hosting a user interface instance on an edge computing system external to a high performance computing system. The process includes providing, by a user interface instance, an encrypted authentication token to a high performance computing system to form a connection between the high performance computing system and an endpoint instance of the high performance computing system; in response to providing the encrypted authentication token, forming a connection between the user interface instance and the endpoint instance; mounting a first file system of the edge computing system to a second file system of the high-performance computing system to provide an expanded first file system; exchanging a user environment of the edge computing system with a user environment of the high performance computing system; in response to a request received by the user interface instance from the connection, the extension file system is accessed by the user interface instance to transfer data between the edge computing system and the high performance computing system.

Description

Background Art

[0001] A high performance computing (HPC) cluster consists of a collection of computing nodes that coordinate their processing activities to achieve a common goal. The computing nodes of an HPC cluster can perform parallel processing on large multidimensional data sets to achieve any of a number of purposes, such as scientific research, financial services fraud detection, animation, hydrocarbon reservoir analysis, aircraft design, autonomous car design, genomics, or any other use case that relies on solutions to problems of computational complexity. BRIEF DESCRIPTION OF THE DRAWINGS

[0002] Figure 1 is a block diagram of a computer network having an HPC system and an edge computing system according to an example implementation.

[0003] Figure 2 is a sequence flow diagram illustrating configuration of an HPC system to prepare the HPC system to exchange user contexts with an edge computing system, according to an example implementation.

[0004] Figure 3 is a sequence flow diagram illustrating configuration of an edge computing system to prepare the edge computing system to exchange user contexts with an HPC system, according to an example implementation.

[0005] Figure 4 is a sequence flow diagram illustrating operations for transferring data between an HPC system and an edge computing system, according to an example implementation.

[0006] Figure 5 is a flow diagram depicting a process for exchanging user contexts of an HPC system and an edge computing system, according to an example implementation.

[0007] Figure 6 is an illustration of machine-readable instructions stored on a non-transitory storage medium that, when executed by a machine, cause the machine to provide a user interface instance on an edge computing system to extend a data plane of an HPC system to the edge computing system according to an example implementation.

[0008] Figure 7 is a block diagram of an HPC system according to an example implementation, the HPC system including: a plurality of computing nodes that perform parallel processing jobs; and an edge computing system manager endpoint that communicates with an interface instance of the edge computing system to extend a data plane of the HPC system to the edge computing system. DETAILED DESCRIPTION

[0009] The HPC system can perform parallel processing to solve computational complexity problems that are beneficial to the edge computing system. In an example, the edge computing system can be an experimental or instrument edge facility, which has a large amount of computing, storage, and networking resources compared to, for example, a resource-constrained edge computing system (e.g., an edge computing system of an Internet of Things (IoT) device). Therefore, a large amount of data can be exchanged between the edge computing system and the HPC system. In addition, the computing activities between the systems can be interrelated. In an example, the edge computing system can perform data reasoning at the edge using a machine learning model, and the HPC system can perform related computing tasks, such as machine learning model training and model parameter tuning. Since the edge system and the HPC system can be independent systems belonging to different management domains, setting up two systems to exchange data can be a daunting challenge. In addition, the data exchange between the two systems may involve manual intervention, which may be impractical for supporting the interrelated computing activities of the two systems. According to an example embodiment, by deploying a user access interface instance (referred to herein as a "UAIe instance") on the edge computing system, the data plane of the HPC system can be extended to the edge computing system. The UAIe instance provides a powerful connection between edge computing systems and HPC systems, which, among other benefits, simplifies connecting edge computing systems to HPC systems and allows HPC systems to exchange data with edge computing systems in an automated manner.

[0010] UAIe instances allow user environments to be exchanged between edge computing systems and HPC systems. In this context, "user environment" refers to a framework for accessing the resources of a computing system. In an example, a user environment may correspond to a specific operating system, a file system, (multiple) applications, certain resources (e.g., input devices, files, data capture devices, mass storage devices, or any other device whose access can be controlled by an operating system) and access rights. In the context used herein, "exchanging" a first user environment with a second user environment refers to creating a user environment with elements from both the first user environment and the second user environment. In an example, a UAIe instance may provide a user environment having a file system familiar to users of a software stack (e.g., an operating system, middleware, virtualization components, applications, and other features) and an edge computing system. In addition, continuing with this example, a UAIe instance may provide elements of a user environment from an HPC system, such as, for example, a shared file system and access rights for an HPC system. Since the UAIe instance is hosted on an edge computing system, a laboratory workflow on an edge computing system may be executed within the context of the UAIe instance. In addition, an experimenter may manage instant control of an experiment in a consistent user environment familiar to both developers and scientists.

[0011] According to an example implementation, setting up an HPC system and an edge computing system to use a UAIe instance involves actions taken in two management domains of the two systems. On the HPC system side, these actions include creating a UAIe container image, associating the UAIe container image with a cryptographic authentication token, initiating an edge manager endpoint instance on the HPC system, and configuring the edge manager endpoint instance based on the cryptographic authentication token to allow external access to the entry service. On the edge computing system side, these actions include deploying a UAIe instance, and using the deployed UAIe instance to mount the resources of the edge computing system to the file system of the edge computing system and to mount the shared file system of the HPC system to the file system of the edge computing system.

[0012] More specifically, according to an example embodiment, to set up an HPC system, a system administrator with appropriate credentials for the HPC system deploys an edge manager endpoint instance on the HPC system. The edge manager endpoint instance provides an ingress service that provides communication with a UAIe instance for the purpose of exchanging data between the edge system and the HPC system. According to an example embodiment, the edge manager endpoint configures a service grid gateway of the HPC system to provide external access to the ingress service. According to an example embodiment, a system administrator may configure the service grid gateway to expose an address (e.g., an Internet Protocol (IP) address or a uniform resource identifier (URI)) of the system for external access. The address (referred to as an "external address") may include a specific port number.

[0013] A system administrator of an HPC system can configure a service grid gateway with an ingress authorization policy that sets forth one or more rules for controlling access to an external address of an ingress service by an external requestor. According to an example implementation, the ingress authorization policy includes a rule that specifies that access requests involving an external address of an ingress service are allowed for a requestor that presents a cryptographic authentication token but is otherwise denied.

[0014] The system administrator of the HPC system can further create a container image for the UAIe instance and store the container image in an image repository or registry of the HPC system. In addition, according to an example embodiment, the system administrator associates an encrypted authentication token with the UAIe container image so that an authorized user of the UAIe image can use the authentication token.

[0015] To set up the edge computing system, an edge computing system user with appropriate credentials deploys an instance of the UAIe instance on the edge computing system using the UAIe container image. The user can then access the UAIe instance to remotely mount the file system of the HPC system to the file system of the edge computing system, thereby creating what is referred to herein as an "extended file system". In an example, the remote mount can use a file system client of the edge computing system. In an example, the file system client can be a secure shell file system (SSHFS) client that uses a file transfer protocol (e.g., a secure file transfer protocol (SFTP) or a secure copy protocol (SCP) or other secure file transfer protocol) to transfer files between the edge computing system and the HPC system. In an example, the edge computing system may have a LINUX operating system, and the SSHFS client may be a tool of the LINUX operating system package. In an example, the extended file system may be a user space file system (FUSE). The configuration of the edge computing system may further include a user of the UAIe instance providing input to the UAIe instance to enable the resources of the edge computing system (e.g., input devices and / or storage devices) to be mounted to the extended file system.

[0016] After the UAIe instance is deployed on the edge computing system, the UAIe instance is connected to the entry service of the HPC system. The connection process includes the UAIe instance providing an encrypted authentication token, and the process includes the service grid gateway of the HPC system authenticating the UAIe instance based on the encrypted authentication token and registering the UAIe instance. After successfully authenticating the UAIe instance, the HPC system initiates and maintains a tunnel connection with the UAIe instance. Through the tunnel connection, the platform service of the HPC system can provide an application programming interface (API) request to the UAIe instance to achieve purposes such as exchanging data between the HPC system and the edge computing system using an extended file system.

[0017] In an example, data exchange may include transferring a file containing input data from an edge computing system, which is consumed by a job of an HPC system. In an example, the input data may represent inference data generated by a machine learning-based inference operation performed on the edge computing system. The computing nodes of the HPC system may process input data for several reasons, such as machine learning model simulation, machine learning model training, machine learning model parameter tuning, or other parallel processing operations involving solving computational complexity problems. In another example, data exchange may include transferring a file containing data generated as output by a job of the HPC system from an HPC system. In an example, the output data may represent a machine learning model, machine learning tuning parameters, or other data consumed by processing operations on the edge computing system.

[0018] refer to Figure 1, computer network 100 includes edge computing system 104 and HPC system 164. According to an example embodiment, edge computing system 104 and HPC system 164 are located in different management domains. In an example, HPC system 164 may be cloud-based and correspond to one or more data centers. In an example, the high-performance computing system may be an exascale computing system that includes hundreds to thousands of computing nodes. In an example, computer network 100 may be a wide area network (WAN) such as the Internet, and edge computing system 104 may correspond to a local branch network. Figure 1 , the network fabric 150 can connect the edge computing system 104 and the HPC system 164. The network fabric 150 can be associated with one or more types of communication networks, such as, by way of example, a Fibre Channel network, a Compute Express Link (CXL) fabric, a dedicated management network, a local area network (LAN), a WAN, a global network, a wireless network, or any combination thereof.

[0019] In the context used herein, an "edge computing system" refers to a system associated with an entry point to a network, such as computer network 100. In an example, edge computing system 104 may be an instrument or experimental facility, and a user of edge computing system 104 may be a scientist or researcher. Among its resources 106, edge computing system 104 may include storage 114, computing resources 109, input devices, data capture devices, operating system 113, and extended file system 112, which includes directories on both edge computing system 104 and HPC system 164, as further described herein. In an example, computing resource 109 may be a node that processes jobs related to reasoning based on a machine learning model based on data captured by edge computing system 104.

[0020] In an example, the HPC system 164 may provide elastic and scalable HPC services for the edge computing system 104. In an example, the HPC system 164 may include clusters of computing nodes 179 that perform parallel processing jobs, such as machine learning model simulation, training, and parameter tuning, for the edge computing system 104. These jobs consume input data provided by the edge computing system 104. The edge computing system 104 may consume the output results of the jobs, such as, for example, data related to the machine learning models used for inference, and tuning adjustments to these models.

[0021] A user associated with the edge computing system 104 may develop, manage, and initiate one or more jobs for the HPC system 164 from a user node 154 (e.g., a computer platform such as a laptop, desktop computer, tablet computing device, or other processor-based electronic device). In an example, the user node 154 may be a dedicated user access node (UAN) configured for general multi-user access to the HPC system 164. In another example and as Figure 1 As depicted in , user node 154 may be a general-purpose processing platform that is configured by a user access interface (UAI) instance 156 hosted on HPC system 164. UAI instance 156 may appear on demand and may run short-term or long-term tasks related to HPC system 164 without dedicating user node 154 to that purpose. According to some embodiments, UAI instance 156 may connect to external storage by mounting the storage within its container, but does not contain its own persistent state. As a result, UAI instance 156 may be considered disposable in that UAI instance 156 may be started and stopped without losing user data. As shown in FIG. Figure 1 As depicted in , a user node 154 may provide a graphical user interface (GUI) 158 (eg, a GUI 158 provided by a web browser) to interact with a UAI instance 156 .

[0022] Figure 1 A dedicated UAI instance, namely a UAIe instance 130, is depicted, which can be deployed on the edge computing system 104 to achieve the purpose of extending the data plane of the HPC system 164 to the edge computing system 104. In addition, the UAIe instance 130 provides a mechanism to exchange user environments of the edge computing system 104 and the HPC system 164. This extension of the data plane automates the exchange of data between the HPC system 164 and the edge computing system 104, as described herein. According to an example embodiment, the UAIe instance 130 provides a user environment that has features acquired from both the edge computing system 104 and the HPC system 164. In an example, the user environment provided by the UAIe instance 130 may have a certain "look and feel" familiar to users of the edge computing system 104, such as a certain software stack, a specific file system, and certain access permissions. In an example, the user environment provided by the UAIe instance 130 may further have features familiar to users of the HPC system 164, such as a shared file system and certain access permissions of the HPC system.

[0023] According to an example embodiment, the UAIe instance 130 provides a command and control interface 133 and a user interface 131. The command and control interface 133 receives and processes requests from the HPC system 164 that are related to transferring data (e.g., files) between the edge computing system 104 and the HPC system 164. Among other possible uses, the user interface 131 allows a user associated with the edge computing system 104 to set up or configure the UAIe instance 130, as further described herein. In this context, "transferring data" between the edge computing system 104 and the HPC system 164 refers to a process that includes moving data from the edge computing system 104 to the HPC system 164, or vice versa. In an example, the file can be transferred from the edge computing system 104 to the HPC system 164 by communicating data representing the file to the HPC system 164, and the transfer of data may or may not include removing or deleting the source file stored on the edge computing system 104. In another example, a file may be transferred from HPC system 164 to edge computing system 104 by communicating data representing the file to edge computing system 104 , and the transfer of the data may or may not include removing or deleting the source file stored on HPC system 164 .

[0024] According to an example embodiment, the command and control interface 133 communicates with the HPC system 164 via the network connection 160. According to an example embodiment, the HPC system 164 may initiate and maintain a secure tunnel via the network connection 160. The secure tunnel maintains the privacy of the data communicated between the UAIe instance 130 and the HPC system 164 even though the data may be communicated over a public network. In an example, a tunneling protocol may be used to form the secure tunnel. In an example, the tunneling protocol may be an Internet Protocol Security (IPSec) protocol, a Secure Socket Tunneling Protocol (SSTP), or another tunneling protocol that uses a payload portion of a data packet corresponding to a first protocol to carry a data packet corresponding to a different second protocol. In another example, the network connection 160 may be a secure shell (SSH) connection, and the secure tunnel may be an SSH tunnel.

[0025] Using the network connection 160, the platform service 170 of the HPC system 164 may communicate requests to the command and control interface 133 of the UAIe instance 130 for various purposes, including communicating requests with the UAIe instance 130 for the purpose of exchanging data (e.g., files) between the edge computing system 104 and the HPC system 164. In an example, the platform service 170 may send a request to the UAIe instance 130, the request including a command initiated for the UAIe instance 130 to copy or move a file from the edge computing system 104 to the HPC system 164. In another example, the platform service 170 may send a request to the UAIe instance 130, the request including a command initiated for the UAIe instance 130 to copy or move a file from the HPC system 164 to the edge computing system 104. In another example, the UAIe instance 130 may send a confirmation of the request sent by the platform service 170 to the platform service 170. According to an example implementation, files are exchanged between the edge computing system 104 and the HPC system 164 using a file system client 140 hosted on the edge computing system 104 .

[0026] like Figure 1 , the file system client 140 communicates with the HPC system 164 via a network connection 162. In an example, the connection 162 may be an SSH connection. In this context, a "file system client" refers to a component of a computing system (e.g., the edge computing system 104) that is constructed to mount a file system of a remote computing system (e.g., the HPC system 164). According to an example implementation, the file system client 140 manages an extended file system 112 that includes file system directories corresponding to storage locations in both the edge computing system 104 and the HPC system 164.

[0027] As used herein, "file" refers to an information container, and "file system" is a method and data structure for organizing a collection of files. In an example, files in a file system can be organized using directories, which are a hierarchical structure or tree; and the location of a file within the file system can be identified by a corresponding file path. In an example, a file system can be a physical file system in which files conform to a specific data storage format. In another example, a file system can be a virtual file system, which can be an upper file system of a composite file system that includes a lower physical file system accessed through a virtual file system. In an example, the edge computing system 104 can have a root file system that contains files for booting the edge computing system 104, and other file systems can be mounted at mount points corresponding to subdirectories of the root file system.

[0028] In an example, for a LINUX operating system, the extended file system 112 may be a user space file system (FUSE), which is a file system that can be created in user space without modifying the operating system kernel. According to an example implementation, the file system client 140 uses a network file transfer protocol to mount the file system of the HPC system 164 locally on the edge computing system 104. In this context, "mounting" of the file system of the HPC system 164 refers to binding one or more directories of the file system to one or more locations or mount points of the file system that is otherwise local to the edge computing system 104. In an example, the file system client 140 may mount a directory or directory tree of a shared file system 177 of the HPC system 164 to a directory ("mount point" or "volume mount point") of the edge computing system 104. The shared file system 177 may be shared by the computing nodes 179 of the HPC system 164.

[0029] In an example, the file system client 140 may be a secure shell file system (SSHFS) client that uses a network file transfer protocol (e.g., secure file transfer protocol (SFTP) or secure copy protocol (SCP) or other secure network file transfer protocols) to transfer files between the edge computing system 104 and the HPC system 164. In an example, the UAIe instance 130 may correspond to a LINUX operating system, and the SSHFS client (file system client 140) may be a tool of the LINUX operating system package. In another example, the file system client 140 may correspond to an independent software entity that is not part of the operating system package.

[0030] According to an example embodiment, in response to a request received by the command and control interface 133 from the HPC system 164, the UAIe instance 130 may initiate an operation to manage file exchange between the edge computing system 104 and the HPC system 164. In this context, the "start" operation refers to initiating or starting an operation. In an example, in response to a specific request from the HPC system 164, the UAIe instance 130 may generate one or more operating system commands to satisfy the request. In a more specific example, the UAIe instance 130 may be associated with a LINUX operating system, and the UAIe instance 130 may receive a request to copy a file from the edge computing system 104 to the HPC system 164. The file may be stored, for example, in a storage device 114 of the edge system 104, which corresponds to a specific directory of the extended file system 112, and the destination may correspond to such a directory of the extended file system 112, that is, the directory is associated with the storage device 116 of the HPC system 164. In response to the request, the UAIe instance 130 may issue a cp command with appropriate arguments to the operating system. According to an example embodiment, the UAIe instance 130 may respond to any of a variety of different file-related requests and file system-related requests by issuing corresponding operating system commands. The user interface 131 of the UAIe instance 130 allows a user (e.g., a user of the user node 154 with appropriate credentials) to access the UAIe instance 130 to configure the extended file system 112. In an example, when the UAIe instance 130 is first started, the user may access the UAIe instance 130 through the user interface 131 to execute operating system commands to mount the resources of the edge system 104 to what will become the extended file system 112. In an example, the user may enter one or more operating system commands (e.g., a create directory command or mkdir command of a LINUX operating system) to add one or more directories to the extended file system and mount an input device or other local resource of the edge computing system 104 to the (multiple) directories. Then, the user may enter a command to remotely mount the shared file system 177 of the HPC system 164. In an example, for a LINUX operating system, a user may first create a directory of the extended file system 112 (which will be a mount point (e.g., using the mkdir command to create the directory)), and then enter an sshfs command to cause the file system client 140 to remotely mount the shared file system 177 of the HPC system 164. The sshfs command may specify a specified directory or directory tree of the shared file system 177 to be mounted to the mount point of the extended file system 112. According to further embodiments, instead of the user entering an operating system command, the user interface 131 of the UAIe instance 130 may receive other inputs from the user (e.g., text input or input via manipulation of an input device (such as a touch pad or a mouse)) and generate corresponding operating system commands to perform the operation indicated by the input.

[0031] When started, the UAIe instance 130 provides a connection request to the HPC system 104 to achieve the purpose of establishing a network connection 160. According to an example embodiment, the UAIe instance 130 provides an encrypted authentication token to the HPC system 104, and as further described herein, the HPC system 104 uses the encrypted authentication token to authenticate the UAIe instance 130. The encrypted authentication token can be provided to the UAIe instance 130 in any of several different ways. In an example, a user can log in to the user interface of the UAIe instance 130 using a set of credentials (e.g., a login ID and a password), and the UAIe instance 130 can find the encrypted authentication token based on the credentials. In another example, the user can provide the encrypted authentication token to the UAIe instance 130 through the user interface of the UAIe instance 130 (e.g., provided by key input). In another example, the user can identify or upload a file containing the encrypted authentication token through the user interface of the UAIe instance 130.

[0032] According to an example implementation, the HPC system 104 includes an edge manager endpoint instance 166 that controls ingress access to platform services 170 of the HPC system 164. According to some implementations, the HPC system 164 includes a service mesh gateway instance 168 that controls whether to allow a network connection request (e.g., an incoming Transmission Control Protocol (TCP) connection request) received from the network fabric 150 based on an authorization policy. If the authorization policy allows the request, the service mesh gateway instance 168 forms the requested network connection, otherwise, the service mesh gateway instance 168 denies the connection.

[0033] In an example, the service mesh gateway instance 168 is an ISITIO gateway that controls whether a network connection can be established with a service 167 of the edge manager endpoint instance 166 based on an ingress authorization policy. In an example, the service 167 may have an external address (e.g., a specific IP address or domain name together with a port number), and the service 167 forms a communication link with the platform service 170. In an example, the authorization policy may allow external connections to the service 167 if the corresponding incoming request targets the external address of the service and provides appropriate real credentials. In an example, the authorization policy may authorize a request made by the UAIe instance 130 to connect to the service 167 in response to a request to provide an encrypted authentication token that is generated by a system administrator of the HPC system 104 and is associated with a container image of the UAIe.

[0034] According to an example embodiment, in response to the service mesh gateway instance 168 successfully authenticating the UAIe instance 130, the edge manager endpoint instance 166 registers the UAIe instance 130. In addition, in response to the successful authentication, the edge manager endpoint instance 166 can initiate and maintain a tunnel with the UAIe instance 130 to enable communication of messages (e.g., messages including file system-related requests and file transfer-related requests) between the platform service 170 and the UAIe instance 130.

[0035] The system administrator may access the HPC system 164 through the management node 152. In this manner, the management node 152 may include a GUI 153 (e.g., a GUI provided by a web browser) for the purpose of allowing the system administrator to perform actions such as creating and configuring a container image 175 corresponding to the UAIe image 130 and storing the container image 175 in an image repository or registry 174 of the HPC system 164. The system administrator of the HPC system 164 may further generate an encrypted authentication token for the UAIe image 130 and store the data in the image registry 174 or another data store, the data representing the encrypted authentication token and accessible to authorized users of the UAIe instance 130. In addition, the system administrator of the HPC system 104 may configure a specific ingress authorization policy for the service mesh gateway instance 168 to allow a connection for a network request from the UAIe instance 130 when the encrypted authentication token is provided. Additionally, to complete the setup of the HPC system 104 for use with the UAIe instance 130, the system administrator may launch an edge manager endpoint instance 166 configured by a specific ingress authorization policy. According to some embodiments, the UAIe instance 130 is a container environment 132. In the context, a "container environment" refers to a collection of one or more instantiated containers (also referred to herein as "containers"). For a container environment that includes multiple containers, these containers can collaborate to achieve a specific purpose (e.g., providing one or more microservices). According to some embodiments, the container environment can be orchestrated. The orchestrated container environment has an orchestrator that manages the lifecycle and workload of the containers of the environment. In an example, the orchestrator can manage the provisioning and resource allocation of containers. In other examples, the orchestrator can manage container replication, when containers start and stop, container scaling, workload distribution among containers, or other lifecycle stages or workload aspects of the container environment. In an example, the orchestrated container environment may have a Kubernetes orchestrator or a Docker Swarm orchestrator.

[0036] According to some embodiments, the UAIe instance 130 may be an orchestrated container environment that includes a cluster of worker nodes 134 (virtual or physical), and each worker node 134 of the cluster may host one or more groups of containers, referred to as "container pods" 136. The lifecycle of the worker nodes 134 may be managed by a control plane of an orchestrator (not shown). The container environment 132 may provide one or more services 138. In an example, a service 138 may correspond to a logical abstraction of a group of container pods 136 that perform the same function. In an example, one or more services 138 may provide a user interface 131 of the UAIe instance 130. In an example, one or more services 138 may provide a command and control interface 133. In an example, one or more services 138 may provide an interface for communicating with a file system client 140.

[0037] In general, a container is a virtual runtime environment for one or more applications and / or application modules, and the virtual runtime environment is built to be docked to the operating system kernel. A container for a given application may, for example, contain the executable code of the application and its dependencies, such as system tools, libraries, configuration files, executable files, and binaries of the application. The container includes an operating system kernel mount interface, but does not include the operating system kernel. Thus, a UAIe instance 130 may, for example, include multiple containers that share a kernel through corresponding operating system kernel mount interfaces. Docker containers and rkt containers are examples of containers.

[0038] According to some embodiments, edge manager endpoint instance 166 may be an orchestrated container environment. The orchestrated container environment may provide an ingress service 167 and one or more services corresponding to a service mesh gateway instance 168.

[0039] According to further example embodiments, one or both of the UAIe instance 130 and / or the edge manager endpoint instance 166 may not be a container environment. In an example, according to further embodiments, the UAIe instance 130 may execute on bare metal or within a virtual machine (and outside a container).

[0040] Among other features of the edge computing system 104, the computing resources 109 may be associated with one or more hardware processors 108 (e.g., central processing units (CPUs), processing cores, graphics processing unit (GPU) cores, or other processors) and memory 110. In an example, the memory 110 is a non-transitory storage medium that may be formed by semiconductor memory devices, memristor-based memory devices, magnetic memory devices, phase change memory devices, a combination of devices of one or more of these storage technologies, and the like. The memory 110 may represent a memory collection of both volatile memory devices and non-volatile memory devices. According to some embodiments, the memory 110 may store machine-readable instructions that, when executed by the one or more hardware processors 108, cause the (multiple) hardware processors 108 to form software components of the edge computing system 104. As further described herein, these components may include a file system client 140, a UAIe instance 130, and one or more application operating environments 120 (e.g., a bare metal environment, a virtual machine, or other virtualized environment).

[0041] In an example, a given storage device 114 of the edge computing system 104 may be a hard drive, an optical drive, a solid-state drive (SSD), a locally attached storage device, a storage area network (SAN), or any other system or device that stores data.

[0042] Among its other features, the HPC system 164 may include a network interface 169. The network interface 169 may include one or more actual or physical network devices (e.g., a network interface controller and a gateway) and one or more virtual network devices. The hardware 180 of the HPC system 164 may include a hardware processing core 182 (e.g., a CPU core and / or a GPU core), a memory 184, and one or more storage devices 186. Similar to the storage device 114 of the edge computer system 104, a given storage device 186 may be a hard drive, an optical drive, a solid-state drive (SSD), a local attached storage device, a storage area network (SAN), or any other system or device that stores data. In general, the hardware 180 supports one or more platform services 170 that may be provided by the HPC system 164. In addition, the platform services 170 may be supported by the infrastructure support 176 of the HPC system 164. In an example, the infrastructure support 176 may include a computing node 179, a shared file system 177 shared by the computing nodes 179, an operating system, a database layer, a hypervisor, or other infrastructure components. In an example, the platform services 170 may include services related to HPC as a Service (HPCaaS).

[0043] Figure 2is a sequence flow diagram 200 illustrating communications between a management node 252 and an HPC system 264 to set up the HPC system 264 to access a UAIe instance hosted by an edge computing system, according to an example implementation. Figure 2 , the sequence flow diagram 200 includes a system administrator of the HPC system 264 creating and configuring a UAIe image, as depicted at 204. Additionally, as also depicted at 204, the system administrator stores the UAIe image in an image repository of the HPC system 264.

[0044] Next, the setup of the HPC system 264 further includes that a system administrator of the HPC system 264 may generate an encrypted authentication token (e.g., an encryption key, a digital certificate, or other encryption artifact containing an alphanumeric sequence) and associate the encrypted authentication token with the UAIe image, as depicted at 208. In an example, the system administrator may associate the HPC system access credentials of one or more edge system users with the encrypted authentication token so that the user can access the token when logging into the HPC system 264.

[0045] Next, the system administrator may create and configure an edge manager endpoint for the HPC system 264, as depicted at 212. In an example, the system administrator may launch an edge manager endpoint instance on the HPC system 264 and configure the instance with an ingress authorization policy. In an example, the ingress authorization policy may specify rules for a requestor requesting a connection to a specific address (e.g., a domain name or IP address along with a port number) corresponding to an ingress service. In an example, the rule may require the requestor to provide a cryptographic authentication token so that the connection is not dropped and require the requestor (e.g., an authorized UAIe instance) to register with the edge manager instance.

[0046] Figure 3 is a sequence flow diagram 300 illustrating actions taken by an authorized user of an edge computing system 304 using a user node 354 to set up the edge computing system to communicate with an edge manager endpoint instance of an HPC system 364, according to an example implementation. Figure 3 , the sequence 300 first includes a user deploying a UAIe instance on an edge computing system 304, as depicted at 320. In an example, the user may deploy the UAIe instance on the edge computing system 304 using Podman, CRI-O, DOCKER, or another container engine.

[0047] After the UAIe instance is deployed and started on the edge computing system, the user next logs into the user interface of the UAIe instance and provides the cryptographic authentication token, as depicted at 324. The cryptographic authentication token may be provided to the UAIe instance in a number of different ways, such as by virtue of a login (e.g., login credentials are mapped to the cryptographic authentication token), the user providing the cryptographic authentication token via input, or user identification a file or location that stores data representing the cryptographic authentication token.

[0048] Next, as depicted at 332, the user may provide input to the UAIe instance to configure a file system (e.g., FUSE) of the edge computing system 304, which includes one or more directories and / or directory trees of the HPC system 364. More specifically, according to an example embodiment, the user may provide input to the UAIe instance to create one or more mount points for the file system of the edge computing system and then mount these mount points to certain directories and / or directory trees of the shared file system of the HPC system. In an example, creating a mount point and remotely mounting the shared file system of the HPC may involve the user providing operating system commands that call operations performed by a file system client of the remote computing system. In another example, creating a mount point and remotely mounting the shared file system of the HPC may involve the user providing input that causes the UAIe system to generate operating system commands that call operations performed by the file system client.

[0049] According to an example embodiment, in response to being provided with the encrypted authentication token, the UAIe instance may send a network connection request (with the encrypted authentication token) to the edge manager endpoint instance of the HPC system 364 and connect to the edge manager endpoint instance, as depicted at 336. Figure 3 This is depicted as occurring after the remote mount related operations, but in general, the UAIe instance can connect to the edge manager endpoint instance at any time after gaining access to the encrypted authentication token. As also depicted at 336, the UAIe instance connects to the edge manager endpoint instance causing the edge manager endpoint instance to initiate and maintain a tunnel with the UAIe instance. At this point, both the HPC system 364 and the edge computing system 304 are configured to extend the data plane of the HPC system 364 to the edge computing system 304.

[0050] Figure 4is a sequence flow diagram 400 illustrating communications to exchange data between an HPC system 480 and an edge computing system 454, according to an example implementation. Figure 4 , as depicted at 404, a platform service 488 of an HPC system 480 generates an API request involving an exchange of data between the HPC system 480 and the edge computing system 454. In an example, the platform service 488 is associated with a job scheduler or manager. In an example, the API request may involve copying or moving a file stored on a storage device 464 of the edge computing system 454 to a storage device 486 of the HPC system 480. In an example, the movement or copying of the file may be associated with moving a file containing data that is captured and processed by the edge computing system 454 and used as an input to one or more parallel processing jobs executed by the computing nodes of the HPC system 480. In an example, the movement or copying of the file may be associated with moving a file containing data that is generated by machine learning-based reasoning (the machine learning-based reasoning is performed by the edge computing system 454) and used as an input to one or more parallel processing jobs executed by the computing nodes of the HPC system 480.

[0051] In another example, the API request may involve copying or moving a file stored on the file storage 486 of the HPC system 480 to the file storage 464 of the edge computing system 454. In an example, the movement or copying of the file may be associated with moving a file containing data representing a machine learning model generated by parallel processing performed by the compute nodes of the HPC system 480. In an example, the movement or copying of the file may be associated with moving a file containing data representing tuning parameters of a machine learning model generated by parallel processing performed by the compute nodes of the HPC system 480.

[0052] As depicted at 408 and 412, the service 484 of the HPC system 480 forwards the API request to the UAIe instance 470 hosted on the edge computing system 454. In an example, the service 484 can be a service of an edge manager endpoint instance of the HPC system 480, and the forwarding of the API request can involve communicating through a tunnel between the edge manager endpoint instance and the UAIe instance 470.

[0053] As depicted at 416, in response to receiving the forwarded API request, the UAIe instance 470 initiates one or more data exchange operations in response to the forwarded API request. In an example, the initiation of the data exchange operation may include the UAIe instance 470 generating one or more operating system commands. The file system client 478 of the edge computing system 454 may then respond to the initiation to perform the requested data exchange operation(s). In an example, the data exchange operation may correspond to an operation to copy or move a file from a directory corresponding to a storage location of the edge computing system 454 to a directory corresponding to a storage location of the HPC system 480, or vice versa. In an example, the file system client 478 may use a network file transfer protocol (such as SFTP) to move or copy data between systems 454 and 480.

[0054] Although Figure 4 Not depicted in the illustration, but according to some embodiments, the UAIe instance 470 may communicate a confirmation message back to the platform service 488 via a network connection with the service 484. In an example, the confirmation may confirm receipt of the API request. In an example, the confirmation may confirm completion of (multiple) data exchange operations.

[0055] refer to Figure 5 According to an example embodiment, process 400 includes hosting (block 504) a user interface instance on an edge computing system external to the high performance computing system. In an example, the edge computing system may be an experimental or instrument edge facility that has a large amount of computing, storage, and network resources compared to a resource-constrained edge computing system (e.g., an edge computing system of an IoT device). In an example, the edge computing system may perform reasoning based on a machine learning model. In an example, the high performance computing system may perform simulations based on a machine learning model. In an example, the high performance computing system may be an exascale computing system comprising hundreds to thousands of computing nodes. In an example, the high performance computing system may tune parameters of the machine learning model. In an example, the user interface instance may be a container environment. In an example, the container environment may include one or more container pods.

[0056] According to block 508, process 500 includes: providing, by the user interface instance, an encrypted authentication token to an endpoint instance of the high performance computing system. In an example, the encrypted authentication token may be an encryption key, a digital certificate, or other cryptographic artifact comprising an alphanumeric sequence. In an example, the endpoint instance may be a container environment. In an example, the endpoint instance may correspond to an edge manager of the high performance computing system. In an example, the endpoint instance may configure a service grid gateway of the high performance computing system. In an example, the service grid gateway may be an ISITIO service grid gateway.

[0057] Process 500 further includes: in response to providing the encrypted authentication token, forming a connection between the user interface instance and the endpoint instance in accordance with block 512. In an example, the endpoint instance can initiate and maintain a tunnel with the user interface instance. In an example, a tunnel protocol can be used to form the tunnel. In an example, the tunnel protocol can be an IPSec protocol, SSTP, or another tunnel protocol that uses a payload portion of a data packet corresponding to a first protocol to carry a data packet corresponding to a different second protocol. In another example, the tunnel can be an SSH tunnel.

[0058] Process 500 includes mounting (block 516) a directory of a first file system of a high performance computing system to a second file system of an edge computing system to provide an extended file system. In an example, mounting may include mounting one or more directories of the first file system to a volume mount point of the second file system. In an example, the second file system of the edge computing system may have one or more locations corresponding to resources of the edge computing system. In an example, mounting includes using a file system client, such as an SSHFS client. In an example, mounting includes using a network file transfer protocol, such as SFTP.

[0059] According to block 520, process 500 includes: exchanging a user environment of the edge computing system with a user environment of the high performance computing system. The exchange includes: in response to receiving a request from the high performance computing system via a connection by the user interface instance, an operation for transmitting data between the edge computing system and the high performance computing system is initiated by the user interface instance. In an example, the transmission of data may include transmitting one or more files between the high performance computing system and the edge computing system. In an example, the transmission of data may include transmitting from the edge computing system a file containing inputs of a job processed by a computing node of the high performance computing system. In an example, the transmission of data may include transmitting from the high performance computing system a file containing outputs of a job processed by a computing node of the high performance computing system.

[0060] refer to Figure 6According to an example embodiment, a non-transitory storage medium 600 stores machine-readable instructions 610. Instructions 610, when executed by a machine, cause the machine to provide a user interface instance on an edge computing system. In an example, the edge computing system may be an experimental or instrument edge facility that has a large amount of computing, storage, and network resources compared to a resource-constrained edge computing system (e.g., an edge computing system of an IoT device). In an example, an edge computing system may perform reasoning based on a machine learning model. In an example, a high-performance computing system may perform simulations based on machine learning models. In an example, a high-performance computing system may be an exascale computing system comprising hundreds to thousands of computing nodes. In an example, a high-performance computing system may tune the parameters of a machine learning model. In an example, a user interface instance may be a container environment. In an example, a container environment may include one or more container pods. In an example, a user interface instance may provide one or more microservices. In an example, a user interface instance may be provided by an orchestrated cluster of worker nodes.

[0061] Instruction 610, when executed by the machine, further causes the machine to perform the following operations: cause the user interface instance to provide a first request to a service grid gateway of the high-performance computing system to register the user interface instance with an endpoint manager instance of the high-performance computing system. In an example, the endpoint manager instance can be a container environment. In an example, the endpoint manager instance can configure the service grid gateway of the high-performance computing system to control external access to the endpoint manager instance service. In an example, the service grid gateway can be an ISITIO service grid gateway.

[0062] Instruction 610, when executed by the machine, further causes the machine to perform the following operations: cause the user interface instance to mount the file system of the edge computing system and the file system of the high-performance computing system to provide an extended file system accessible to the edge computing system. In an example, mounting may include mounting one or more directories of the file system of the high-performance system to a volume mount point of the file system of the edge computing system. In an example, the file system of the edge computing system may have one or more locations corresponding to resources of the edge computing system. In an example, mounting includes using a file system client, such as an SSHFS client. In an example, mounting includes using a network file transfer protocol, such as SFTP.

[0063] Instruction 610, when executed by the machine, further causes the machine to perform the following operations: cause the user interface instance to transfer data between locations in the extended file system in response to a second request sent via the service grid gateway to exchange data between the edge computing system and the high-performance computing system. In an example, the transmission of data may include transferring one or more files between the high-performance computing system and the edge computing system. In an example, the transmission of data may include transferring a file containing data from the edge computing system, the data representing an input to a job processed by a computing node of the high-performance computing system. In an example, the transmission of data may include transferring a file containing data from the high-performance computing system, the data representing an output of a job processed by a computing node of the high-performance computing system.

[0064] refer to Figure 7 According to an example embodiment, the first computing system 700 includes a plurality of computing nodes 704 that perform parallel processing jobs associated with an edge computing system. In an example, the first computing system 700 may be a high performance computing system. In an example, the computing nodes 704 may perform jobs associated with simulations based on machine learning models. In an example, the computing nodes 704 may perform jobs associated with tuning parameters of a model based on machine learning. In an example, the first computing system 700 may be an exascale system having hundreds to thousands of computing nodes 704.

[0065] The first computing system 700 further includes an edge manager endpoint 708. In an example, the edge manager endpoint 708 can be an instance corresponding to a container environment. The edge computing system manager endpoint 708 authenticates the interface instance based on an encrypted authentication token provided by the interface instance of the edge computing system. In an example, the edge computing system can be an instrument or an experimental facility. In an example, the edge computing system can perform reasoning based on a machine learning model. In an example, the encrypted authentication token can be an encryption key. In an example, the interface instance of the edge computing system can be provided by a container environment. In an example, the interface instance can be provided by one or more container pods.

[0066] In response to successfully authenticating the interface instance, the edge manager endpoint 708 forms a connection between the edge manager endpoint and the interface instance. In an example, the endpoint manager endpoint 708 can initiate and maintain a tunnel with the user interface instance. In an example, the tunnel can be formed using a tunneling protocol. In an example, the tunneling protocol can be an IPSec protocol, SSTP, or another tunneling protocol that uses a payload portion of a data packet corresponding to a first protocol to carry a data packet corresponding to a different second protocol. In another example, the tunnel can be an SSH tunnel.

[0067] The edge manager endpoint 708 sends a request to the interface instance over the connection to transfer a file between a first location of the extended file system associated with the edge computing system and a second location of the extended file system associated with the high-performance computing system. In an example, the request may cause a plurality of files to be transferred between the high-performance computing system and the edge computing system. In an example, the file may be transferred from the edge computing system, and the file may contain data representing an input to a job processed by a computing node of the first computing system 700. In an example, the file may be transferred from the first computing system 700, and the file may contain data representing an output of a job processed by a computing node of the first computing system 700.

[0068] According to an example implementation, forming a connection between the user interface instance and the endpoint instance includes forming a connection with a service mesh gateway of the high performance computing system. A particular advantage is that the data plane of the high performance computing system can be extended to the edge computing system.

[0069] According to an example implementation, hosting the user interface includes hosting the container environment on the edge computing system. A particular advantage is that the data plane of the high performance computing system can be extended to the edge computing system.

[0070] According to an example implementation, mounting the directory of the first file system includes: communicating between the edge computing system and the high performance computing system using a network file transfer protocol. A particular advantage is that the data plane of the high performance computing system can be extended to the edge computing system.

[0071] According to an example implementation, computing resources of the edge computing system are controlled in response to a second request received by the first user interface instance from the connection. A particular advantage is that a data plane of a high performance computing system can be extended to an edge computing system.

[0072] According to an example implementation, the user interface instance reports the status of operations performed by computing resources of the edge computing system through the connection.A particular advantage is that the data plane of the high performance computing system can be extended to the edge computing system.

[0073] According to an example implementation, the high performance computing system includes a plurality of computing nodes, and the second file system includes a file system shared by the plurality of computing nodes. A particular advantage is that the data plane of the high performance computing system can be extended to the edge computing system.

[0074] According to an example implementation, resources of the edge computing system are mounted to the second file system through the user interface instance. A particular advantage is that the data plane of the high performance computing system can be extended to the edge computing system.

[0075] According to an example implementation, mounting includes: mounting a subdirectory of the first file system to a mount point of the second file system. A particular advantage is that the data plane of the high performance computing system can be extended to the edge computing system.

[0076] According to an example implementation, machine learning inference is performed by an edge computing system. The machine learning inference is associated with a machine learning model. The high performance computing system trains the machine learning model. A particular advantage is that the data plane of the high performance computing system can be extended to the edge computing system.

[0077] According to an example implementation, transmitting data includes transmitting data associated with the machine learning inference from the edge computing system to the high performance computing system. A particular advantage is that the data plane of the high performance computing system can be extended to the edge computing system.

[0078] According to an example implementation, transmitting data includes transmitting data associated with parameters of the machine learning model from the high performance computing system to the edge computing system. A particular advantage is that the data plane of the high performance computing system can be extended to the edge computing system.

[0079] According to an example implementation, the initiating operation includes initiating an operation to transfer data associated with the machine learning reasoning from the edge computing system to the high performance computing system. A particular advantage is that the data plane of the high performance computing system can be extended to the edge computing system.

[0080] According to an example implementation, the initiating operation includes initiating an operation to transfer data associated with parameters of the machine learning model from the high performance computing system to the edge computing system. A particular advantage is that the data plane of the high performance computing system can be extended to the edge computing system.

[0081] According to an example implementation, a user interface instance receives a request to transfer a file associated with machine learning reasoning to a high performance computing system. The file is transferred to the high performance computing by the edge computing system in response to the request. A particular advantage is that the data plane of the high performance computing system can be extended to the edge computing system.

[0082] The detailed description set forth herein refers to the accompanying drawings. Wherever possible, the same reference numerals are used in the drawings and the foregoing description to refer to the same or similar parts. However, it will be clearly understood that the drawings are for illustration and description purposes only. Although several examples are described in this document, modifications, adaptations and other implementations are possible. Therefore, the detailed description does not limit the disclosed examples. On the contrary, the proper scope of the disclosed examples may be defined by the appended claims.

[0083] The terms used herein are only used for the purpose of describing specific examples and are not intended to be restrictive. As used herein, the singular forms "one", "a kind of" and "the" are intended to also include plural forms, unless the context clearly indicates otherwise. As used herein, the term "multiple" is defined as two or more. As used herein, the term "another" is defined as at least the second or more. Unless otherwise indicated, the term "connection" as used herein is defined as connection, whether it is a direct connection without any intermediate element or an indirect connection with at least one intermediate element. Two elements can be mechanically, electrically or communicated through a communication channel, path, network or system. As used herein, the term "and / or" refers to and covers any and all possible combinations of the associated listed items. It will also be understood that although the terms first, second, third, etc. can be used herein to describe various elements, these elements should not be limited by these terms, because unless otherwise stated or the context indicates otherwise, these terms are only used to distinguish one element from another element. As used herein, the term "includes" means including but not limited to, and the term "including" means including but not limited to. The term "based on" means based at least in part on.

[0084] While the present disclosure has been described with respect to a limited number of embodiments, numerous modifications and variations will become apparent to those skilled in the art having the benefit of this disclosure. The appended claims are intended to cover all such modifications and variations.

Claims

1. A method comprising: Hosting a user interface instance on an edge computing system; Providing, by the user interface instance, an encrypted authentication token to an endpoint instance of a high performance computing system external to the edge computing system; In response to providing the encrypted authentication token, forming a connection between the user interface instance and the endpoint instance; Mounting a directory of a first file system of the high performance computing system to a second file system of the edge computing system to provide an extended file system; as well as In response to receiving a request by the user interface instance from the high performance computing system via the connection, operations are initiated by the user interface instance to exchange data between the edge computing system and the high performance computing system.

2. The method according to claim 1, wherein: Forming the connection between the user interface instance and the endpoint instance includes forming a connection with a service grid gateway of the high performance computing system.

3. The method according to claim 1, wherein: Hosting the user interface instance includes hosting a container environment on the edge computing system.

4. The method according to claim 1, wherein: Mounting the first file system includes: using a network file transfer protocol to communicate between the edge computing system and the high-performance computing system.

5. The method according to claim 1, further comprising: In response to a second request received by the user interface instance from the connection, computing resources of the edge computing system are controlled.

6. The method according to claim 1, further comprising: In response to a computing resource of the edge computing system performing an operation, a status of the operation is reported by the user interface instance and through the connection.

7. The method according to claim 1, wherein: The high performance computing system includes a plurality of computing nodes, and the second file system includes a file system shared by the plurality of computing nodes.

8. The method according to claim 1, further comprising: The user interface instance mounts the resources of the edge computing system to the first file system.

9. The method according to claim 1, wherein: The mounting includes: mounting the subdirectory of the first file system to the mount point of the second file system.

10. The method according to claim 1, further comprising: performing, by the edge computing system, machine learning inference associated with the machine learning model; as well as The machine learning model is trained by the high performance computing system.

11. The method according to claim 10, wherein: Initializing the operations includes: Initiate operations to transfer data associated with the machine learning inference from the edge computing system to the high performance computing system.

12. The method according to claim 10, wherein: Initializing the operations includes: Initiate operations to transfer data associated with parameters of the machine learning model from the high performance computing system to the edge computing system.

13. The method according to claim 10, further comprising: receiving, by the user interface instance, a request to transfer a file associated with the machine learning inference to the high performance computing system; as well as The edge computing system transmits the file to the high performance computing in response to the request.

14. A non-transitory storage medium storing machine-readable instructions that, when executed by a machine, cause the machine to: Providing a user interface instance on the edge computing system; Causes the user interface instance to perform the following operations: providing a first request to a service grid gateway of a high performance computing system to register the user interface instance with an endpoint manager instance of the high performance computing system; Mounting the file system of the edge computing system and the file system of the high performance computing system to provide an extended file system accessible to the edge computing system; as well as In response to a second request sent via the service mesh gateway, data is moved between locations of the extended file system to exchange data between the edge computing system and the high performance computing system.

15. The storage medium according to claim 14, wherein: The instructions, when executed by the machine, further cause the machine to perform the following operations: in response to the communication, provide data representing a file directory of the first file system to the high performance computing system.

16. The storage medium according to claim 14, wherein: The instructions, when executed by the machine, further cause the machine to perform the following operations: transfer a file stored in a storage resource associated with the edge computing system to the high performance computing system.

17. The storage medium according to claim 14, wherein: The instructions, when executed by the machine, further cause the machine to perform the following operations: transfer a file stored in a storage resource associated with the high performance computing system to the edge computing system.

18. A first computing system comprising: a plurality of computing nodes that perform parallel processing jobs associated with the edge computing system; as well as The edge manager endpoint, which does the following: authenticating the interface instance based on an encrypted authentication token provided by the interface instance of the edge computing system; In response to successfully authenticating the interface instance, forming a connection between the edge manager endpoint and the interface instance; as well as A request is sent to the interface instance over the connection to move a file between a first location of an extended file system associated with the edge computing system and a second location of the extended file system associated with the first computing system.

19. The first computing system of claim 18, wherein: The parallel processing job processes input data provided by the edge computing system; as well as In response to the access, the edge computing system manager transmits the input data from the edge computing system to the first computing system.

20. The first computing system of claim 18, wherein: The edge computing system performs machine learning-based reasoning based on the machine learning model; The parallel processing determines parameters of the machine learning model; as well as In response to the access, the edge computing system manager transmits the data representing the parameter from the first computing system to the edge computing system.

Images