Container Acceleration Method Based on Image Block Granularity Optimization

In the DevOps development mode, a significant amount of redundant data often accumulates in the image file layer. The repeated loading of this redundant data consumes valuable storage and network resources in the cluster, impacting the efficiency of cluster scheduling and Pod creation. To address the data redundancy caused by container images, this paper proposes a block granularity optimization method for image storage. The image file layer is divided into multiple equally sized data blocks, with a weak check code calculated for each block to validate its content. Additionally, a strong check code is used to verify the integrity of each data block and determine if the internal data of the file layer can be reused, reducing data redundancy across different versions. Experimental results demonstrate that the algorithm effectively reuses identical data within the file layer, resulting in a more than 4.2 times increase in Pod creation speed.