Jun Nemoto

File System Backup to Object Storage for On-Demand Restore

A new backup method, which achieves on-demand restore as part of file system backup to object storage, is proposed. On-demand restore is a function that restores only a certain directory or file according to the request from an end user. The proposed method backs up relationships between a file and an object on a directory basis in order to efficiently handle the relationships in the restore. It is experimentally shown that the proposed method reduces the response time of file access involving restore by over an order of magnitude.

On IO Latency Prediction Accuracy and Automated Load Balancing in Consolidated VM Environments

Manually managing IO workloads and performance in consolidated VM environments is often difficult and error prone. Thus, automated IO workload (re) placement using virtual disk migration is a key functionality of large scale VM infrastructure. A promising approach is to place IO workloads based on predicted IO latencies, but previous prediction models are often inaccurate due to dependence on there being only a linear relationship between workload parameters and IO latency. This paper presents a new accurate IO latency prediction model for use in automated load balancing. Our experiments show that our model improves relative error ratio of IO latency prediction by 67% for SSDs and 43% for HDDs on average. We also evaluate how the improvement of IO latency prediction affects actual load balancing and overall IO performance. Contrary to our expectation, we find that the significant improvement of IO latency prediction accuracy does not translate into significant overall performance improvement.