OS Image Selection and Caching for Provisioning Bottlenecks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data processing systems face bottlenecks during provisioning due to the time-consuming transmission of complete disk images over networks, especially when multiple resource requests are processed simultaneously, leading to potential delays.
Innovation Solution
A method is introduced to select and cache operating system (OS) images in a data processing system based on probability calculations, determining the optimal combination of OS images to minimize the expected time needed for transmission, using a combination determination unit and probability calculation unit to identify the most likely used OS images and cache them, thereby reducing the need for transmitting full images and accelerating provisioning by using differential images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If complete disk images are transmitted through the network during provisioning, then the provisioning can be performed with full functionality, but the transmission time becomes excessively long and network bandwidth is consumed
Solution Approach 1:
The disk image is segmented into two parts: a base image that is cached in advance on the client system, and a differential image that contains only the changes needed for the specific provisioning request. This segmentation allows the majority of the image data to be locally available while only transmitting the minimal necessary differences over the network.
Solution Approach 2:
The base image is cached in advance on the client system before the actual provisioning operation occurs. This preliminary action ensures that the bulk of the image data is already present locally, eliminating the need to transmit it during the provisioning process and significantly reducing transmission time.
2Productivity
If numerous resource requests are processed simultaneously on the repository server, then more provisioning can be performed in parallel, but a bottleneck occurs in the repository server and processing is delayed
Solution Approach 1:
The provisioning process is segmented into a preparation phase (caching base images) and an execution phase (applying differential images). This segmentation distributes the computational workload: the repository server only needs to manage and transmit small differential images during parallel provisioning operations, while the client systems perform the heavier processing of applying these differences to their locally cached base images.
3Reliability
If the cache size for OS images is increased to store more image types, then the probability of having the required image available increases, but the storage space required increases
Solution Approach 1:
The image storage is segmented into a base image component that is cached locally and a differential image component that is stored and transmitted separately. This allows the system to maintain a small cache of base images while supporting a much broader variety of OS configurations through the use of differential images that can be generated on-demand from the base image and the desired configuration.
Data Source
AI summary
A computer-implemented method, selection program, device and article of manufacture for selecting images of one or more operating systems (OS) that are cached in a target data processing system. The method can be implemented in a provisioning system where the system includes a first pool having the images of a plurality of different OS and a second pool having a plurality of data processing systems. The method includes: calculating the probability that the respective OS images in the first pool will be used in the next provisioning; and determining a combination of one or more OS images as one or more OS images to be cached.