Cloud computing is a new computing technique for massive data centers that keeps computational resources online rather than on local machines. As cloud computing grows in popularity, so does the need for cloud resources. Container placements on physical hosts in Infrastructure-as-a-Service data centers are constantly tuned in response to the usage of host resources. When a container is migrated, a huge amount of data is transferred between hosts, and in some cases when it migrates back then the same amount of data is transmitted again. In this paper, the proposed approach for container migration to migrate back to the same host is described. Container migration enables load balancing, system maintenance, and fault tolerance, among other things. In some cases, the container will migrate back to the same host. The original image kept on the source host can be reused in such cases. The memory pages similar to the source image will not be sent back; only the updated pages will be transferred. This approach helps in reducing the amount of data transmission over the network. Furthermore, if the container image is kept on the source host, it will provide demand paging and help recover from failure at the destination host. The result shows the average rate of reduction in the data transfer over the network by 60.68% compared to standard precopy and 52.30% compared to advanced pre-copy.
No takes yet. Share an insight, caveat, or question.
Singh et al. (2022) studied this question.