Youngjin Kwon

Virtualizing performance asymmetric multi-core systems

Performance-asymmetric multi-cores consist of heterogeneous cores, which support the same ISA, but have different computing capabilities. To maximize the throughput of asymmetric multi-core systems, operating systems are responsible for scheduling threads to different types of cores. However, system virtualization poses a challenge for such asymmetric multi-cores, since virtualization hides the physical heterogeneity from guest operating systems. In this paper, we explore the design space of hypervisor schedulers for asymmetric multi-cores, which do not require asymmetry-awareness from guest operating systems. The proposed scheduler characterizes the efficiency of each virtual core, and map the virtual core to the most area-efficient physical core. In addition to the overall system throughput, we consider two important aspects of virtualizing asymmetric multi-cores: performance fairness among virtual machines and performance scalability for changing availability of fast and slow cores. We have implemented an asymmetry-aware scheduler in the open-source Xen hypervisor. Using applications with various characteristics, we evaluate how effectively the proposed scheduler can improve system throughput without asymmetry-aware operating systems. The modified scheduler improves the performance of the Xen credit scheduler by as much as 40% on a 12-core system with four fast and eight slow cores. The results show that even the VMs scheduled to slow cores have relatively low performance degradations, and the scheduler provides scalable performance with increasing fast core counts.

SSD-HDD-hybrid virtual disk in consolidated environments

With the prevalence of multi-core processors and cloud computing, the server consolidation using virtualization has increasingly expanded its territory, and the degree of consolidation has also become higher. As a large number of virtual machines individually require their own disks, the storage capacity of a data center could be exceeded. To address this problem, copy-on-write storage systems allow virtual machines to initially share a template disk image. This paper proposes a hybrid copy-on-write storage system that combines solid-state disks and hard disk drives for consolidated environments. In order to take advantage of both devices, the proposed scheme places a read-only template disk image on a solid-state disk, while write operations are isolated to the hard disk drive. In this hybrid architecture, the disk I/O performance benefits from the fast read access of the solid-state disk, especially for random reads, precluding write operations from the degrading flash memory performance. We show that the hybrid virtual disk, in terms of performance and cost, is more effective than the pure copy-on-write disks for a highly consolidated system.

The effect of multi-core on HPC applications in virtualized systems

In this paper, we evaluate the overheads of virtualization in commercial multicore architectures with shared memory and MPI-based applications. We find that the non-uniformity of memory latencies affects the performance of virtualized systems significantly. Due to the lack of support for non-uniform memory access (NUMA) in the Xen hypervisor, shared memory applications suffer from a significant performance degradation by virtualization. MPI-based applications show more resilience on sub-optimal NUMA memory allocation and virtual machine (VM) scheduling. However, using multiple VMs on a physical system for the same instance of MPI applications may adversely affect the overall performance, by increasing I/O operations through the domain 0 VM. As the number of cores increases on a chip, the cache hierarchy and external memory will become more asymmetric. As such non-uniformity in memory systems increases, NUMA and cache awareness in VM scheduling will be critical for shared memory applications.

Strata: A Cross Media File System

Current hardware and application storage trends put immense pressure on the operating systems storage subsystem. On the hardware side, the market for storage devices has diversified to a multi-layer storage topology spanning multiple orders of magnitude in cost and performance. Above the file system, applications increasingly need to process small, random IO on vast data sets with low latency, high throughput, and simple crash consistency. File systems designed for a single storage layer cannot support all of these demands together. We present Strata, a cross-media file system that leverages the strengths of one storage media to compensate for weaknesses of another. In doing so, Strata provides performance, capacity, and a simple, synchronous IO model all at once, while having a simpler design than that of file systems constrained by a single storage device. At its heart, Strata uses a log-structured approach with a novel split of responsibilities among user mode, kernel, and storage layers that separates the concerns of scalable, high-performance persistence from storage layer management. We quantify the performance benefits of Strata using a 3-layer storage hierarchy of emulated NVM, a flash-based SSD, and a high-density HDD. Strata has 20-30% better latency and throughput, across several unmodified applications, compared to file systems purpose-built for each layer, while providing synchronous and unified access to the entire storage hierarchy. Finally, Strata achieves up to 2.8x better throughput than a block-based 2-layer cache provided by Linuxs logical volume manager.

From Crash Consistency to Transactions

Modern applications use multiple storage abstractions such as the file system, key-value stores, and embedded databases such as SQLite. Maintaining consistency of data spread across multiple abstractions is complex and error-prone. Applications are forced to copy data unnecessarily and use long sequences of system calls to update state in a consistent manner. Not only does this create implementation complexity, it also introduces potential performance problems from redundant IO and fsync() calls, which fragment disk writes into small, random IOs. In this paper, we propose that the operating system should provide transactions across multiple storage abstractions; we can build such transactions with low development cost by taking advantage of a well-tested piece of software: the file-system journal. We present the design of our cross-abstraction transactions and some preliminary results, showing such transactions can increase performance by 31% in certain cases.

Ingens: Huge Page Support for the OS and Hypervisor

Memory capacity and demand have grown hand in hand in recent years. However, overheads for memory virtualization, in particular for address translation, grow with memory capacity as well, motivating hardware manufacturers to provide TLBs with thousands of entries for larger pages, or huge pages. Current OSes and hypervisors support huge pages with a hodge-podge of best-effort algorithms and spot fixes that make less and less sense as architectural support for huge pages matures. The time has come for a more fundamental redesign. Ingens is a framework for providing transparent huge page support in a coordinated way. Ingens manages contiguity as a first-class resource, and tracks utilization and access frequency of memory pages, enabling it to eliminate pathologies that plague current systems. Experiments with a Linux/KVM-based prototype show improved fairness and performance, and reduced tail latency and memory bloat for important applications such as Web services and Redis. We report early experiences with our in-progress port of Ingens to the ESX Hypervisor.