A Block-level Deduplication-based Container Deployment Framework

Container technology has gained great popularity in cloud environment since containers provide near-native performance and are lighter and less expensive than traditional virtual machines.Although the startup of containers is faster than traditional virtual machines,starting up a non-local container,whose image is unavailable locally,is time-consuming.The reason is that pulling an image from a remote registry requires a long time.Experiments show that a lot of redundant data exists among image layers when pulling images from the registry.This redundant data causes additional pull time and makes the startup of a non-local container slower.Container technology uses layer-level sharing and data compression to reduce the data download when pulling images,which speeds up the start of containers.However,layer-level sharing and data compression cannot avoid pulling of the redundant data between different images layers.Therefore,reducing the data download during image pull as well as providing an intact image to accelerate the startup of containers is essential.To solve problems above,a container deployment framework,BED,based on blocklevel deduplication is proposed.To be specific,BED stores an image layer as numerous data blocks,corresponding fingerprints,and a fingerprint list which is generated based on these data blocks.When BED needs to pull an image,it intercepts information of the corresponding image,pulls the corresponding fingerprint lists,deduplicates these fingerprint lists,and pulls the data blocks non-existing locally from the registry.Based on the local and pulled blocks,BED reconstructs the image layers for a container.Block decompression,block pull,and image reconstruction are overlapped to cover up overhead caused by block-level deduplication.Experiments show that compared with original Docker,BED reduces the time of pulling images by 35% on average and saves about 48% data transmission in the network and 43% data in the storage.