16 results about "Storage resource management" patented technology

The invention provides a multi-type storage cluster dynamic nano-management method based on a cloud platform, and belongs to the technical field of cloud platform storage resource management, a storage pool is used as a minimum management unit, a configuration center provides a unified nano-management entrance and is associated with a platform initialization process, and initialization of standardized parameters such as a storage cluster IP, an account and the like is automatically completed; after management, block storage, file storage or object storage types and exclusive parameters thereof are configured for the storage pool according to business requirements, and available storage pools are screened according to binding relations when storage services are created, so that resource mismatching is avoided. Bidirectional synchronization is carried out between the operation and maintenance resource management module and the storage module through a full life cycle by using an equipment SN number, so that state consistency and traceability are realized; expansion of newly added storage pools, unbinding of the storage pools and resource recovery are supported, and the resource utilization rate and the operation and maintenance efficiency are improved.

The invention discloses a storage resource management method for cloud edge collaboration and a program product. The method comprises the following steps: acquiring multiple target storage resource parameters of a target edge node at multiple moments, and determining a first capacity quota of the target edge node according to the multiple target storage resource parameters at the multiple moments; generating a first token according to the first capacity quota, generating a first signature according to the first token, sending the first token to the target edge node, and receiving a second signature and a first storage resource sent by the target edge node, the first storage resource being determined according to the first token, the generation mode of the second signature is determined according to the generation mode of the first signature; and writing the first storage resource of the target edge node into a cloud resource scheduling center based on the first signature and the second signature. By analyzing the storage resource data of the edge nodes, more reasonable and effective storage resource allocation is realized, and the stability of the cloud edge collaborative system is effectively guaranteed.

The invention discloses a checkpoint management method and device, a storage medium and a product, and relates to the technical field of storage resource management, and the method comprises the steps: obtaining a checkpoint generated in a model training process; based on the time attribute information of the check points, the check points are distributed to the corresponding check point retention intervals, and the time span of the check point retention interval of any later stage is larger than the time span of the check point retention interval of the adjacent previous stage; and after the check point reservation interval is filled with the check points, at least one check point is selected from the check points of the check point reservation interval for reservation processing, and unselected check points in the check point reservation interval are deleted. Through the cooperation of the characteristics of the rapid growth sequence and the management among the partitions, the bottleneck of the linear growth of the memory space is broken through, the automatic and ordered management of the check points is realized, manual intervention is not needed, the management stability and reliability are remarkably improved, and the method is adaptive to a long-period large-scale deep learning model training scene.

The invention relates to the technical field of transformer substation intelligent inspection, and provides a transformer substation inspection visual analysis system storage resource management method and system, and the method comprises the steps: classifying and organizing inspection data, and enabling the inspection data to be classified and stored according to a time sequence relation and a source; performing intelligent analysis and compression on ultra-high-definition inspection data in the inspection data; processing data collected by the silent patrol point positions in the patrol data, and eliminating low-value data which does not trigger alarm; and automatically cleaning low-value historical data through real-time calculation according to the actual use condition of the storage space. Through comprehensive application of hierarchical storage, intelligent compression and automatic cleaning, efficient management of the whole life cycle of the inspection data can be realized.

The present disclosure provides a graphics card storage resource management method and device, electronic equipment and readable medium, and relates to the technical field of graphics cards. The graphics card storage resource management method comprises: obtaining an access history of a virtual address block; and according to the access history, performing step-by-step migration of the virtual address block between a graphics card, a host memory and a solid state disk in a UVM framework. Through the embodiment of the present disclosure, the solid state disk is organically integrated into the UVM system as the third level of storage, the addressable space of the graphics card is expanded, the access frequency and the characteristics of the storage medium are balanced, and the probability of hot data residing in the high-speed layer is improved. While optimizing the utilization rate of the graphics card storage resource, the graphics card access efficiency and reliability are ensured.

This application discloses a storage resource management system and a storage system, relating to the storage field. The storage resource management system includes: a first control module, used to determine the current access mode and generate a switching instruction based on the current access mode; and a first switching module, which includes multiple first ports and multiple second ports. The first ports are used to connect to computing resources, and the second ports are used to connect to storage resources. Upon receiving a switching instruction, the first switching module controls the connection between the first port and the second port corresponding to the switching instruction, so that the computing resources access the storage resources according to the current access mode. This application offers flexible configuration and ensures that if one data link fails or loses power, the remaining data links can still operate normally without affecting business operations, thus improving the reliability of the storage system.

The application provides a storage resource management method and device, a storage medium and an electronic device. When a storage resource management request is received, device information of a storage device corresponding to the storage resource management request and a keyword of an API to be called are determined. The device information and the keyword are input into an IoC container to obtain an object instance of the API to be called. The IoC container obtains registration information and a dependency relationship of the API to be called based on the device information and the keyword, and reflects the registration information and the dependency relationship to obtain the object instance of the API to be called. Based on the object instance, the API to be called is called to manage a storage resource corresponding to the storage resource management request in the storage device. It can be seen that, according to the application, the IoC container is pre-constructed in a test environment, the object instance is generated by the IoC container, the API is called, and thus the coupling degree and the difficulty of code maintenance are reduced.

Examples described herein provide a computer-implemented method for identity provider agnostic departmental multi-tenancy management of storage resources that includes providing a software-defined storage-as-a-service (SDSaaS) workspace that defines a tenancy circle, wherein the SDSaaS workspace interconnects an open-source container orchestration system namespace and the storage resources. The method further includes assigning the storage resources to departments within the SDSaaS workspace. The method further includes enabling departments to manage their own storage resources independently from one another. The method further includes implementing role-based access control at a SDSaaS workspace level to control user access to the storage resources of the SDSaaS workspace.

The application provides a storage resource management method and device of a data processor and electronic equipment. The method comprises the following steps: determining a target bus link in which the data processor is located from a plurality of peripheral component interconnect standard bus links by enumerating a device supporting the peripheral component interconnect standard; configuring the target bus link to adopt a target link function configuration; the supported storage resource read-write bandwidth of the target link function configuration is higher than that of a default link function configuration; and managing the storage resource of the data processor based on the target bus link. By configuring the target bus link to adopt the target link function configuration, the target bus link after the configuration can support high-speed read-write of the storage resource of the DPU, and efficient use of the storage resource of the DPU under the operating system can be realized.

The application belongs to the technical field of storage, and discloses a storage resource management method and system, which comprises the following steps: defining a cluster storage resource controller, storing bandwidth information of each resource, and performing full-dimension event monitoring on cluster resources; deploying an adjusting unit in each node through the cluster storage resource controller, and allocating a local resource space by using the adjusting unit; according to the bandwidth information in the cluster storage resource controller, calculating a global reference weight of each adjusting unit according to a weight priority in each resource, and performing normalization processing; adding metadata information of a storage resource of each adjusting unit to a ring queue, constructing a storage service quality scheduling virtual ring queue, and dynamically allocating the storage resource of the adjusting unit by using pointer scheduling and time period segmentation. The application improves the flexibility of storage resource allocation among different applications.

The application provides a storage resource management method and device, electronic equipment and storage medium. The method comprises the following steps: determining a storage resource allocation request and a local pooled hard disk corresponding to the storage resource allocation request; if the storage capacity in the local pooled hard disk is less than the storage demand capacity in the storage resource allocation request, determining an expansion capacity according to the storage capacity in the local pooled hard disk and the storage demand capacity in the storage resource allocation request; if the expansion capacity is greater than the available capacity of a global local hard disk, determining a first other local pooled hard disk according to the hard disk type information in the storage resource allocation request; and expanding the local pooled hard disk according to the first target local hard disk and the first other local pooled hard disk to obtain a target local pooled hard disk, wherein the first target local hard disk is a hard disk in an available state in the global local hard disk. The method improves the allocation effect of the local storage resource.

The application discloses a kind of service object dynamic configuration method, it is related to storage resource management technical field, including steps: by the characteristic parameters such as data scale, attribute number, change frequency, query frequency and response time requirement of service object acquisition, calculate configuration flexibility index CFI and performance decay index PDM in application server memory, according to the comparison of preset threshold with the calculation result Adaptive selection semantic storage mode SSM or structured storage mode CSM;When selecting SSM, adopt two-stage batch query to obtain entity ID after batch query attribute value, when selecting CSM, according to metadata dynamically generated and executed table structure change statement, simultaneously, corresponding entity class bytecode is generated in memory and is loaded, if execution fails, then trigger rollback operation.The application can adaptively select optimal storage mode, realizes the non-stop change of service object structure, balances the flexibility and high performance demand of system.

The present application discloses a resource management method and apparatus and a device. The method comprises: in response to a first request, using a first service to generate a second request for a first application, the first request representing an image encoding or decoding request, the first service representing an image encoding and decoding service, and the second request representing a hardware storage resource allocation request; on the basis of the first application, using the first service to generate corresponding identification information for the second request, the identification information being used for indicating whether hardware storage resources allocated to the second request are allowed to be passively recycled; and on the basis of the second request, using a second service to allocate the hardware storage resources to the second request, the second service representing a hardware storage resource management service, and the hardware storage resources representing storage resources used for performing image encoding or decoding.

The invention relates to the technical field of computers, and provides a storage resource management method and device. The method comprises the following steps: acquiring a user login request; redirecting the user login request to a target server operating system through a bastion host; collecting the storage resource use information of each current user account in real time by using the storage resource attribute identifier corresponding to the target server operating system; according to the storage resource use information of each current user account, performing freezing operation on a target user account by using the bastion host; and performing storage resource directional management and control processing based on the freezing operation. Through the embodiment of the invention, the problem of uneven distribution of the storage resources can be improved by automatically managing and controlling the storage resources of the server operating system supporting simultaneous login of multiple users.

The invention provides a display card storage resource management method and device, electronic equipment and a readable medium, and relates to the technical field of display cards, and the display card storage resource management method comprises the following steps: obtaining an access history of a virtual address block; according to the access history, in the UVM framework, the virtual address blocks are migrated step by step among a graphics card, a host memory and a solid state disk. According to the embodiment of the invention, the solid state disk is organically integrated into the UVM system as the third-level storage, the addressable space of the graphics card is expanded, the access frequency and the storage medium characteristics are balanced, the probability that the hotspot data resides in the high-speed layer is improved, and the access efficiency and reliability of the graphics card are guaranteed while the utilization rate of the storage resources of the graphics card is optimized.

A data processing method, apparatus, chip, and cluster are disclosed, relating to the field of cloud computing. The method includes, when multiple tenants' task data are stored in the memory of a first computing node, a storage resource management module stores the task data of the first tenant in the memory of the first computing node according to the virtual storage resources of the first tenant, and stores the task data of the second tenant in the memory of the first computing node according to the virtual storage resources of the second tenant; the virtual storage resources of the first tenant and the second tenant are isolated. The data processing method provided in this application ensures that each tenant has independent virtual storage resources, and the virtual storage resources of multiple tenants are isolated. Even when multiple tenants share memory, storing the task data of multiple tenants in memory according to their different virtual storage resources achieves data isolation among multiple tenants and improves data security.