In this paper, we describe a whole-system live migration scheme, which transfers the whole system run-time state, including CPU memory data, and local disk storage, of virtual machine (VM). To minimize downtime caused by migrating large storage data keep integrity consistency, propose three-phase (TPM) algorithm. facilitate back to initial source machine, use an incremental (IM) algorithm reduce amount be migrated. Block-bitmap is used track all write accesses during migration....

Virtualization essentially enables multiple operating systems and applications to run on one physical computer by multiplexing hardware resources. A key motivation for applying virtualization is improve resource utilization while maintaining reasonable quality of service. However, such a goal cannot be achieved without efficient management. Though most resources, as processor cores I/O devices, are shared among virtual machines using time slicing can scheduled flexibly based priority,...

When a cache is shared by multiple cores, its space may be allocated either sharing, partitioning, or both. We call the last case partition-sharing. This paper studies partition-sharing as general solution, and presents theory an technique for optimizing present to optimize partition sharing. The shows that problem of reducible partitioning. uses dynamic programming partitioning overall miss ratio, two different kinds fairness. Finally, evaluates effect optimal sharing compares it with...

In a multicore system, effective management of shared last level cache (LLC), such as hardware/software partitioning, has attracted significant research attention. Some eminent progress is that Intel introduced Cache Allocation Technology (CAT) to its commodity processors recently. CAT implements way partitioning and provides software interface control allocation. Unfortunately, can only allocate at level, which does not scale well for large thread or program count serve their various...

As virtualization becomes a key technique for supporting cloud computing, much effort has been made to reduce overhead, so virtualized system can match its native performance. One major overhead is due memory or page table virtualization. Conventional virtual machines rely on shadow mechanism manage tables, where maintained by the VMM (Virtual Machine Monitor) maps addresses machine while guest maintains own physical table. This will result in expensive VM exits whenever there fault that...

Virtualization is often used in systems for the purpose of offering isolation among applications running separate virtual machines (VM). Current machine monitors (VMMs) have done a decent job resource memory, CPU and I/O devices. However, when looking further into usage lower-level shared cache, we notice that one machine's cache behavior may interfere with another's due to uncontrolled sharing. In this situation, performance cannot be guaranteed. This paper presents partitioning approach...

Deep learning (DL) shows its prosperity in a wide variety of fields. The development DL model is time-consuming and resource-intensive procedure. Hence, dedicated GPU accelerators have been collectively constructed into datacenter. An efficient scheduler design for such datacenter crucially important to reduce the operational cost improve resource utilization. However, traditional approaches designed big data or high performance computing workloads can not support fully utilize resources....

The reuse distance (least recently used (LRU) stack distance) is an essential metric for performance prediction and optimization of storage cache. Over the past four decades, there have been steady improvements in algorithmic efficiency measurement. This progress accelerating recent years, both theory practical implementation. In this article, we present a kinetic model LRU cache memory, based on average eviction time (AET) cached data. AET enables fast measurement use low-cost sampling. It...

The tiered-memory system can effectively expand the memory capacity for virtual machines (VMs). However, virtualization introduces new challenges specifically in enforcing performance isolation, minimizing context switching, and providing resource overcommit. None of state-of-the-art designs consider address these challenges; we observe that a VM with tiered incurs up to 2× slowdown compared DRAM-only VM. We propose vTMM , hardware-software collaborative management framework virtualization....

Nowadays, there are many similar services available on the internet, making Quality of Service (QoS) a key concern for users. Since collecting QoS values all through user invocations is impractical, predicting more feasible approach. Matrix factorization considered an effective prediction method. However, most existing matrix algorithms focus capturing global similarities between users and services, overlooking local their neighbors, as well non-interactive effects services. This paper...

GPU memory systems adopt a multi-dimensional hardware structure to provide the bandwidth necessary support 100s 1000s of concurrent threads. On software side, GPU-compute workloads also use structures organize We observe that these can combine unfavorably and create significant resource imbalance in subsystem - causing low performance poor power-efficiency. The key issue is it highly application-dependent which address bits exhibit high variability. To solve this problem, we first an entropy...

On multi-core processors, contention on shared resources such as the last level cache (LLC) and memory bandwidth may cause serious performance degradation, which makes efficient resource allocation a critical issue in data centers. Intel recently introduces Memory Bandwidth Allocation (MBA) technology its Xeon scalable it possible to allocate real system. However, how make most of MBA improve system remains an open question. In this work, (1) we formulate quantitative relationship between...

The memory demand of virtual machines (VMs) is increasing, while the traditional DRAM-only system has limited capacity and high power consumption. tiered can effectively expand increase cost efficiency. Virtualization introduces new challenges for tiering, specifically enforcing performance isolation, minimizing context switching, providing resource overcommit. However, none state-of-the-art designs consider virtualization thus address these challenges; we observe that a VM with incurs up to...

This paper surveys virtualization of I/O devices, which is one the most difficult parts in system virtualization. Current technologies virtualizing devices include full virtualization, paravirtualization, software emulation and VMM-bypass direct I/O. Optimizations are also done to improve performance each technology. Most optimizations used paravirtualization technology for reference. performs best performance, but VM direct-access be migrated, it hard capture states without VMM...

Virtual Machine (VM) cloning is to create a replica of source virtual machine (parent machine); the replica, also called child machine, owns exactly same executing status as parent machine. Fast live guarantees that, during period cloning, services running on observe no performance degradation. There are three important goals for fast cloning: reducing total time, minimizing suspension time and maximizing resource sharing between This paper exploits Copy-on-Write (CoW) mechanism fully...