Application level ballooning for efficient server consolidation
15 Apr 2013-pp 337-350
TL;DR: This work extends ballooning to applications so that memory can be efficiently and effectively moved between virtualized instances as the demands of each change over time, with significantly lower memory requirements.
Abstract: Systems software like databases and language runtimes typically manage memory themselves to exploit application knowledge unavailable to the OS. Traditionally deployed on dedicated machines, they are designed to be statically configured with memory sufficient for peak load. In virtualization scenarios (cloud computing, server consolidation), however, static peak provisioning of RAM to applications dramatically reduces the efficiency and cost-saving benefits of virtualization. Unfortunately, existing memory "ballooning" techniques used to dynamically reallocate physical memory between VMs badly impact the performance of applications which manage their own memory. We address this problem by extending ballooning to applications (here, a database engine and Java runtime) so that memory can be efficiently and effectively moved between virtualized instances as the demands of each change over time. The results are significantly lower memory requirements to provide the same performance guarantees to a collocated set of VM running such applications, with minimal overhead or intrusive changes to application code.
Citations
More filters
01 Mar 2014
TL;DR: Ginseng is presented, the first market-driven cloud system that allocates memory efficiently to selfish cloud clients and achieves a 6.2×--15.8x improvement in aggregate client satisfaction when compared with state-of-the-art approaches for cloud memory allocation.
Abstract: Physical memory is the scarcest resource in today's cloud computing platforms. Cloud providers would like to maximize their clients' satisfaction by renting precious physical memory to those clients who value it the most. But real-world cloud clients are selfish: they will only tell their providers the truth about how much they value memory when it is in their own best interest to do so. How can real-world cloud providers allocate memory efficiently to those (selfish) clients who value it the most?We present Ginseng, the first market-driven cloud system that allocates memory efficiently to selfish cloud clients. Ginseng incentivizes selfish clients to bid their true value for the memory they need when they need it. Ginseng continuously collects client bids, finds an efficient memory allocation, and re-allocates physical memory to the clients that value it the most. Ginseng achieves a 6.2×--15.8x improvement (83%--100% of the optimum) in aggregate client satisfaction when compared with state-of-the-art approaches for cloud memory allocation.
75 citations
Cites methods from "Application level ballooning for ef..."
...’s database [44] or a Java runtime with balloons [19, 27, 44] or CRAMM [58]....
[...]
19 Mar 2018
TL;DR: Illuminator is presented, an efficient memory manager that provides various subsystems, such as the page allocator, the ability to track all unmovable pages, which allows subsystems to make informed decisions and eliminate unnecessary work which in turn leads to cost-effective huge page allocations.
Abstract: The virtual-to-physical address translation overhead, a major performance bottleneck for modern workloads, can be effectively alleviated with huge pages. However, since huge pages must be mapped contiguously, OSs have not been able to use them well because of the memory fragmentation problem despite hardware support for huge pages being available for nearly two decades. This paper presents a comprehensive study of the interaction of fragmentation with huge pages in the Linux kernel. We observe that when huge pages are used, problems such as high CPU utilization and latency spikes occur because of unnecessary work (e.g., useless page migration) performed by memory management related subsystems due to the poor handling of unmovable (i.e., kernel) pages. This behavior is even more harmful in virtualized systems where unnecessary work may be performed in both guest and host OSs. We present Illuminator, an efficient memory manager that provides various subsystems, such as the page allocator, the ability to track all unmovable pages. It allows subsystems to make informed decisions and eliminate unnecessary work which in turn leads to cost-effective huge page allocations. Illuminator reduces the cost of compaction (up to 99%), improves application performance (up to 2.3x) and reduces the maximum latency of MySQL database server (by 30x). Importantly, this work shows the effectiveness of a simple solution for long-standing huge page related problems.
74 citations
Cites background from "Application level ballooning for ef..."
...Contrary to a popular belief in the academic literature that fragmentation is not a critical problem and OSs can efficiently recover from fragmentation with memory compaction (i.e., by relocating pages) [29, 31, 46], we show that fragmentation can indeed cause the aforementioned issues with huge pages....
[...]
...Hence, we can safely infer that the huge page support available in hardware for nearly two decades is not effectively utilized by OSs....
[...]
...OSs can leverage the hardware support for huge pages in two ways....
[...]
..., object lifetime) is, in general, not available to OSs [54, 62]....
[...]
...However, harnessing the full potential of huge pages requires efficient fragmentation mitigation by the OSs....
[...]
24 Sep 2017
TL;DR: This paper enumerates the challenges of remote memory, discusses their feasibility, explain how some of them are addressed by recent work, and indicates other promising ways to tackle them.
Abstract: As the latency of the network approaches that of memory, it becomes increasingly attractive for applications to use remote memory---random-access memory at another computer that is accessed using the virtual memory subsystem. This is an old idea whose time has come, in the age of fast networks. To work effectively, remote memory must address many technical challenges. In this paper, we enumerate these challenges, discuss their feasibility, explain how some of them are addressed by recent work, and indicate other promising ways to tackle them. Some challenges remain as open problems, while others deserve more study. In this paper, we hope to provide a broad research agenda around this topic, by proposing more problems than solutions.
70 citations
Cites background from "Application level ballooning for ef..."
...faults, though doing so is notably hard [26, 41]....
[...]
24 Feb 2014
TL;DR: VSwapper is implemented, a guest-agnostic memory swapper for virtual environments that allows efficient, uncooperative overcommitment and can achieve up to double the performance under changing load conditions.
Abstract: The number of guest virtual machines that can be consolidated on one physical host is typically limited by the memory size, motivating memory overcommitment. Guests are given a choice to either install a "balloon" driver to coordinate the overcommitment activity, or to experience degraded performance due to uncooperative swapping. Ballooning, however, is not a complete solution, as hosts must still fall back on uncooperative swapping in various circumstances. Additionally, ballooning takes time to accommodate change, and so guests might experience degraded performance under changing conditions. Our goal is to improve the performance of hosts when they fall back on uncooperative swapping and/or operate under changing load conditions. We carefully isolate and characterize the causes for the associated poor performance, which include various types of superfluous swap operations, decayed swap file sequentiality, and ineffective prefetch decisions upon page faults. We address these problems by implementing VSwapper, a guest-agnostic memory swapper for virtual environments that allows efficient, uncooperative overcommitment. With inactive ballooning, VSwapper yields up to an order of magnitude performance improvement. Combined with ballooning, VSwapper can achieve up to double the performance under changing load conditions.
66 citations
Cites background from "Application level ballooning for ef..."
...Application-level ballooning [58] can mitigate negative side effects of ballooning on applications that manage their own memory....
[...]
23 Apr 2018
TL;DR: An effortless way for disaggregating the CPU-memory couple, two of the most important resources in cloud computing, is proposed and can improve the energy efficiency of state-of-the-art consolidation techniques by up to 86%, with minimal additional complexity.
Abstract: In this paper, we propose an effortless way for disaggregating the CPU-memory couple, two of the most important resources in cloud computing. Instead of redesigning each resource board, the disaggregation is done at the power supply domain level. In other words, CPU and memory still share the same board, but their power supply domains are separated. Besides this disaggregation, we make the two following contributions: (1) the prototyping of a new ACPI sleep state (called zombie and noted Sz) which allows to suspend a server (thus save energy) while making its memory remotely accessible; and (2) the prototyping of a rack-level system software which allows the transparent utilization of the entire rack resources (avoiding resource waste). We experimentally evaluate the effectiveness of our solution and show that it can improve the energy efficiency of state-of-the-art consolidation techniques by up to 86%, with minimal additional complexity.
60 citations
Cites background from "Application level ballooning for ef..."
...This comes from the fact that most applications and operating systems are configured according to the RAM size they see at start time [51]....
[...]
..., the balloon driver inside VMs), and they have limited returns with diverse workloads [51]....
[...]
References
More filters
19 Oct 2003
TL;DR: Xen, an x86 virtual machine monitor which allows multiple commodity operating systems to share conventional hardware in a safe and resource managed fashion, but without sacrificing either performance or functionality, considerably outperform competing commercial and freely available solutions.
Abstract: Numerous systems have been designed which use virtualization to subdivide the ample resources of a modern computer. Some require specialized hardware, or cannot support commodity operating systems. Some target 100% binary compatibility at the expense of performance. Others sacrifice security or functionality for speed. Few offer resource isolation or performance guarantees; most provide only best-effort provisioning, risking denial of service.This paper presents Xen, an x86 virtual machine monitor which allows multiple commodity operating systems to share conventional hardware in a safe and resource managed fashion, but without sacrificing either performance or functionality. This is achieved by providing an idealized virtual machine abstraction to which operating systems such as Linux, BSD and Windows XP, can be ported with minimal effort.Our design is targeted at hosting up to 100 virtual machine instances simultaneously on a modern server. The virtualization approach taken by Xen is extremely efficient: we allow operating systems such as Linux and Windows XP to be hosted simultaneously for a negligible performance overhead --- at most a few percent compared with the unvirtualized case. We considerably outperform competing commercial and freely available solutions in a range of microbenchmarks and system-wide tests.
6,326 citations
09 Dec 2002
TL;DR: Several novel ESX Server mechanisms and policies for managing memory are introduced, including a ballooning technique that reclaims the pages considered least valuable by the operating system running in a virtual machine, and an idle memory tax that achieves efficient memory utilization.
Abstract: VMware ESX Server is a thin software layer designed to multiplex hardware resources efficiently among virtual machines running unmodified commodity operating systems. This paper introduces several novel ESX Server mechanisms and policies for managing memory. A ballooning technique reclaims the pages considered least valuable by the operating system running in a virtual machine. An idle memory tax achieves efficient memory utilization while maintaining performance isolation guarantees. Content-based page sharing and hot I/O page remapping exploit transparent page remapping to eliminate redundancy and reduce copying overheads. These techniques are combined to efficiently support virtual machine workloads that overcommit memory.
1,528 citations
Proceedings Article•
11 Apr 2007TL;DR: This work presents Sandpiper, a system that automates the task of monitoring and detecting hotspots, determining a new mapping of physical to virtual resources and initiating the necessary migrations, and implements a black- box approach that is fully OS- and application-agnostic and a gray-box approach that exploits OS-and- application-level statistics.
Abstract: Virtualization can provide significant benefits in data centers by enabling virtual machine migration to eliminate hotspots. We present Sandpiper, a system that automates the task of monitoring and detecting hotspots, determining a new mapping of physical to virtual resources and initiating the necessary migrations. Sandpiper implements a black-box approach that is fully OS- and application-agnostic and a gray-box approach that exploits OS- and application-level statistics. We implement our techniques in Xen and conduct a detailed evaluation using a mix of CPU, network and memory-intensive applications. Our results show that Sandpiper is able to resolve single server hotspots within 20 seconds and scales well to larger, data center environments. We also show that the gray-box approach can help Sandpiper make more informed decisions, particularly in response to memory pressure.
931 citations
20 Aug 2002
TL;DR: This work provides a framework to assess the abilities of an XML database to cope with a broad range of different query types typically encountered in real-world scenarios and offers a set of queries where each query is intended to challenge a particular aspect of the query processor.
Abstract: While standardization efforts for XML query languages have been progressing, researchers and users increasingly focus on the database technology that has to deliver on the new challenges that the abundance of XML documents poses to data management: validation, performance evaluation and optimization of XML query processors are the upcoming issues. Following a long tradition in database research, we provide a framework to assess the abilities of an XML database to cope with a broad range of different query types typically encountered in real-world scenarios. The benchmark can help both implementors and users to compare XML databases in a standardized application scenario. To this end, we offer a set of queries where each query is intended to challenge a particular aspect of the query processor. The overall workload we propose consists of a scalable document database and a concise, yet comprehensive set of queries which covers the major aspects of XML query processing ranging from textual features to data analysis queries and ad hoc queries. We complement our research with results we obtained from running the benchmark on several XML database platforms. These results are intended to give a first baseline and illustrate the state of the art.
822 citations
01 Apr 2009
TL;DR: Experimental evaluation with RUBiS and TPC-W benchmarks along with production-trace-driven workloads indicates that AutoControl can detect and mitigate CPU and disk I/O bottlenecks that occur over time and across multiple nodes by allocating each resource accordingly.
Abstract: Virtualized data centers enable sharing of resources among hosted applications. However, it is difficult to satisfy service-level objectives(SLOs) of applications on shared infrastructure, as application workloads and resource consumption patterns change over time. In this paper, we present AutoControl, a resource control system that automatically adapts to dynamic workload changes to achieve application SLOs. AutoControl is a combination of an online model estimator and a novel multi-input, multi-output (MIMO) resource controller. The model estimator captures the complex relationship between application performance and resource allocations, while the MIMO controller allocates the right amount of multiple virtualized resources to achieve application SLOs. Our experimental evaluation with RUBiS and TPC-W benchmarks along with production-trace-driven workloads indicates that AutoControl can detect and mitigate CPU and disk I/O bottlenecks that occur over time and across multiple nodes by allocating each resource accordingly. We also show that AutoControl can be used to provide service differentiation according to the application priorities during resource contention.
553 citations