Harmonium: Elastic Cloud Storage via File Motifs

Modern applications expand to fill the space available to them, exploiting local storage to improve performance by caching, prefetching and precomputing data. In virtual-ized settings, this behavior compromises storage elastic-ity owing to a rigid contract between the hypervisor and the guest OS: once space is allocated to a virtual disk and used by an application, it cannot be reclaimed by the hy-pervisor. In this paper, we propose a new guest filesystem called Harmonium that exploits the ephemeral or deriva-tive nature of application data. Each file in Harmonium optionally has a motif that describes how the file can be reconstructed via computation, network accesses, or oper-ations on other files. Harmonium expands files from their motifs when space is available, and contracts them back to their motifs when it is scarce. Given a target size, the sys-tem selects files to expand or contract based on the load on the CPU, network, and storage, as well as expected access patterns. As a result, Harmonium enables elastic cloud storage, allowing the hypervisor to dynamically balance storage across multiple VMs.