G. Heiser, Etienne Le Sueur, A. Danis, Aleksander Budzynowski, T. Salomie, G. Alonso
Database management systems provide updates with guaranteed durability in the presence of OS crashes or power failures. Durability is achieved by performing synchronous writes to a transaction log on stable, non-volatile storage. The procedure is expensive and several techniques have been devised to ameliorate the impact on overall performance at the cost of increased system complexity. In this paper we explore the possibility of reducing the system complexity around logging by leveraging verification instead of using specialised/dedicated hardware or complicated optimisations. The prototype system, RapiLog, uses a dependable hypervisor based on seL4 to buffer log data outside the database system and its OS, and performs the physical disk writes asynchronously with respect to the operation of the database. RapiLog guarantees that the log data will eventually be written to the disk even if the database system or the underlying OS crash or electrical power is cut. We evaluate RapiLog with multiple open-source and commercial database engines and find that performance is never degraded (beyond the virtualisation overhead), and at times is significantly improved.
{"title":"RapiLog: reducing system complexity through verification","authors":"G. Heiser, Etienne Le Sueur, A. Danis, Aleksander Budzynowski, T. Salomie, G. Alonso","doi":"10.1145/2465351.2465383","DOIUrl":"https://doi.org/10.1145/2465351.2465383","url":null,"abstract":"Database management systems provide updates with guaranteed durability in the presence of OS crashes or power failures. Durability is achieved by performing synchronous writes to a transaction log on stable, non-volatile storage. The procedure is expensive and several techniques have been devised to ameliorate the impact on overall performance at the cost of increased system complexity. In this paper we explore the possibility of reducing the system complexity around logging by leveraging verification instead of using specialised/dedicated hardware or complicated optimisations. The prototype system, RapiLog, uses a dependable hypervisor based on seL4 to buffer log data outside the database system and its OS, and performs the physical disk writes asynchronously with respect to the operation of the database. RapiLog guarantees that the log data will eventually be written to the disk even if the database system or the underlying OS crash or electrical power is cut. We evaluate RapiLog with multiple open-source and commercial database engines and find that performance is never degraded (beyond the virtualisation overhead), and at times is significantly improved.","PeriodicalId":20737,"journal":{"name":"Proceedings of the Eleventh European Conference on Computer Systems","volume":"23 1","pages":"323-336"},"PeriodicalIF":0.0,"publicationDate":"2013-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75406986","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Malte Schwarzkopf, A. Konwinski, M. Abd-El-Malek, J. Wilkes
Increasing scale and the need for rapid response to changing requirements are hard to meet with current monolithic cluster scheduler architectures. This restricts the rate at which new features can be deployed, decreases efficiency and utilization, and will eventually limit cluster growth. We present a novel approach to address these needs using parallelism, shared state, and lock-free optimistic concurrency control. We compare this approach to existing cluster scheduler designs, evaluate how much interference between schedulers occurs and how much it matters in practice, present some techniques to alleviate it, and finally discuss a use case highlighting the advantages of our approach -- all driven by real-life Google production workloads.
{"title":"Omega: flexible, scalable schedulers for large compute clusters","authors":"Malte Schwarzkopf, A. Konwinski, M. Abd-El-Malek, J. Wilkes","doi":"10.1145/2465351.2465386","DOIUrl":"https://doi.org/10.1145/2465351.2465386","url":null,"abstract":"Increasing scale and the need for rapid response to changing requirements are hard to meet with current monolithic cluster scheduler architectures. This restricts the rate at which new features can be deployed, decreases efficiency and utilization, and will eventually limit cluster growth. We present a novel approach to address these needs using parallelism, shared state, and lock-free optimistic concurrency control.\u0000 We compare this approach to existing cluster scheduler designs, evaluate how much interference between schedulers occurs and how much it matters in practice, present some techniques to alleviate it, and finally discuss a use case highlighting the advantages of our approach -- all driven by real-life Google production workloads.","PeriodicalId":20737,"journal":{"name":"Proceedings of the Eleventh European Conference on Computer Systems","volume":"39 1","pages":"351-364"},"PeriodicalIF":0.0,"publicationDate":"2013-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73779819","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
TimeStream is a distributed system designed specifically for low-latency continuous processing of big streaming data on a large cluster of commodity machines. The unique characteristics of this emerging application domain have led to a significantly different design from the popular MapReduce-style batch data processing. In particular, we advocate a powerful new abstraction called resilient substitution that caters to the specific needs in this new computation model to handle failure recovery and dynamic reconfiguration in response to load changes. Several real-world applications running on our prototype have been shown to scale robustly with low latency while at the same time maintaining the simple and concise declarative programming model. TimeStream handles an on-line advertising aggregation pipeline at a rate of 700,000 URLs per second with a 2-second delay, while performing sentiment analysis of Twitter data at a peak rate close to 10,000 tweets per second, with approximately 2-second delay.
{"title":"TimeStream: reliable stream computation in the cloud","authors":"Zhengping Qian, Yong He, Chunzhi Su, Zhuojie Wu, Hongyu Zhu, Taizhi Zhang, Lidong Zhou, Yuan Yu, Zheng Zhang","doi":"10.1145/2465351.2465353","DOIUrl":"https://doi.org/10.1145/2465351.2465353","url":null,"abstract":"TimeStream is a distributed system designed specifically for low-latency continuous processing of big streaming data on a large cluster of commodity machines. The unique characteristics of this emerging application domain have led to a significantly different design from the popular MapReduce-style batch data processing. In particular, we advocate a powerful new abstraction called resilient substitution that caters to the specific needs in this new computation model to handle failure recovery and dynamic reconfiguration in response to load changes. Several real-world applications running on our prototype have been shown to scale robustly with low latency while at the same time maintaining the simple and concise declarative programming model. TimeStream handles an on-line advertising aggregation pipeline at a rate of 700,000 URLs per second with a 2-second delay, while performing sentiment analysis of Twitter data at a peak rate close to 10,000 tweets per second, with approximately 2-second delay.","PeriodicalId":20737,"journal":{"name":"Proceedings of the Eleventh European Conference on Computer Systems","volume":"220 1","pages":"1-14"},"PeriodicalIF":0.0,"publicationDate":"2013-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73814942","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abhilash Jindal, Abhinav Pathak, Y. C. Hu, S. Midkiff
To maximally conserve the critical resource of battery energy, smartphone OSes implement an aggressive system suspend policy that suspends the whole system after a brief period of user inactivity. This burdens developers with the responsibility of keeping the system on, or waking it up, to execute time-sensitive code. Developer mistakes in using the explicit power management unavoidably give rise to energy bugs, which cause significant, unexpected battery drain. In this paper, we study a new class of energy bugs, called sleep conflicts, which can happen in smartphone device drivers. Sleep conflict happens when a component in a high power state is unable to transition back to the base power state because the system is suspended when the device driver code responsible for driving the transition is supposed to execute. We illustrate the root cause of sleep conflicts, develop a classification of the four types of sleep conflicts, and finally present a runtime system that performs sleep conflict avoidance, along with a simple yet effective pre-deployment testing scheme. We have implemented and evaluated our system on two Android smartphones. Our testing scheme detects several sleep conflicts in WiFi and vibrator drivers, and our runtime avoidance scheme effectively prevents sleep conflicts from draining the battery.
{"title":"Hypnos: understanding and treating sleep conflicts in smartphones","authors":"Abhilash Jindal, Abhinav Pathak, Y. C. Hu, S. Midkiff","doi":"10.1145/2465351.2465377","DOIUrl":"https://doi.org/10.1145/2465351.2465377","url":null,"abstract":"To maximally conserve the critical resource of battery energy, smartphone OSes implement an aggressive system suspend policy that suspends the whole system after a brief period of user inactivity. This burdens developers with the responsibility of keeping the system on, or waking it up, to execute time-sensitive code. Developer mistakes in using the explicit power management unavoidably give rise to energy bugs, which cause significant, unexpected battery drain.\u0000 In this paper, we study a new class of energy bugs, called sleep conflicts, which can happen in smartphone device drivers. Sleep conflict happens when a component in a high power state is unable to transition back to the base power state because the system is suspended when the device driver code responsible for driving the transition is supposed to execute. We illustrate the root cause of sleep conflicts, develop a classification of the four types of sleep conflicts, and finally present a runtime system that performs sleep conflict avoidance, along with a simple yet effective pre-deployment testing scheme. We have implemented and evaluated our system on two Android smartphones. Our testing scheme detects several sleep conflicts in WiFi and vibrator drivers, and our runtime avoidance scheme effectively prevents sleep conflicts from draining the battery.","PeriodicalId":20737,"journal":{"name":"Proceedings of the Eleventh European Conference on Computer Systems","volume":"3 1","pages":"253-266"},"PeriodicalIF":0.0,"publicationDate":"2013-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87976704","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
F. Cruz, Francisco Maia, M. Matos, R. Oliveira, J. Paulo, J. Pereira, R. Vilaça
NoSQL databases manage the bulk of data produced by modern Web applications such as social networks. This stems from their ability to partition and spread data to all available nodes, allowing NoSQL systems to scale. Unfortunately, current solutions' scale out is oblivious to the underlying data access patterns, resulting in both highly skewed load across nodes and suboptimal node configurations. In this paper, we first show that judicious placement of HBase partitions taking into account data access patterns can improve overall throughput by 35%. Next, we go beyond current state of the art elastic systems limited to uninformed replica addition and removal by: i) reconfiguring existing replicas according to access patterns and ii) adding replicas specifically configured to the expected access pattern. MeT is a prototype for a Cloud-enabled framework that can be used alone or in conjunction with OpenStack for the automatic and heterogeneous reconfiguration of a HBase deployment. Our evaluation, conducted using the YCSB workload generator and a TPC-C workload, shows that MeT is able to i) autonomously achieve the performance of a manual configured cluster and ii) quickly reconfigure the cluster according to unpredicted workload changes.
{"title":"MeT: workload aware elasticity for NoSQL","authors":"F. Cruz, Francisco Maia, M. Matos, R. Oliveira, J. Paulo, J. Pereira, R. Vilaça","doi":"10.1145/2465351.2465370","DOIUrl":"https://doi.org/10.1145/2465351.2465370","url":null,"abstract":"NoSQL databases manage the bulk of data produced by modern Web applications such as social networks. This stems from their ability to partition and spread data to all available nodes, allowing NoSQL systems to scale. Unfortunately, current solutions' scale out is oblivious to the underlying data access patterns, resulting in both highly skewed load across nodes and suboptimal node configurations.\u0000 In this paper, we first show that judicious placement of HBase partitions taking into account data access patterns can improve overall throughput by 35%. Next, we go beyond current state of the art elastic systems limited to uninformed replica addition and removal by: i) reconfiguring existing replicas according to access patterns and ii) adding replicas specifically configured to the expected access pattern.\u0000 MeT is a prototype for a Cloud-enabled framework that can be used alone or in conjunction with OpenStack for the automatic and heterogeneous reconfiguration of a HBase deployment. Our evaluation, conducted using the YCSB workload generator and a TPC-C workload, shows that MeT is able to i) autonomously achieve the performance of a manual configured cluster and ii) quickly reconfigure the cluster according to unpredicted workload changes.","PeriodicalId":20737,"journal":{"name":"Proceedings of the Eleventh European Conference on Computer Systems","volume":"7 1","pages":"183-196"},"PeriodicalIF":0.0,"publicationDate":"2013-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90306430","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Andrew Baumann, Dongyoon Lee, Pedro Fonseca, Lisa Glendenning, Jacob R. Lorch, Barry Bond, Reuben Olinsky, G. Hunt
Library OS (LibOS) architectures implement the OS personality as a user-mode library, giving each application the flexibility to choose its LibOS. This approach is appealing for many reasons, not least the ability to extend or customise the LibOS. Recent work with Drawbridge [29] showed that an existing commodity OS (Windows 7) could be refactored to produce a LibOS while retaining application compatibility. This paper presents Bascule, an architecture for LibOS extensions based on Drawbridge. Rather than relying on the application developer to customise a LibOS, Bascule allows OS-independent extensions to be attached at runtime. Extensions interpose on a narrow binary interface of primitive OS abstractions, such as files and virtual memory. Thus, they are independent of both guest and host OS, and composable at runtime. Since an extension runs in the same process as an application and its LibOS, it is safe and efficient. Bascule demonstrates extension reuse across diverse guest LibOSes (Windows and Linux) and host OSes (Windows and Barrelfish). Current extensions include file system translation, checkpointing, and architecture adaptation.
{"title":"Composing OS extensions safely and efficiently with Bascule","authors":"Andrew Baumann, Dongyoon Lee, Pedro Fonseca, Lisa Glendenning, Jacob R. Lorch, Barry Bond, Reuben Olinsky, G. Hunt","doi":"10.1145/2465351.2465375","DOIUrl":"https://doi.org/10.1145/2465351.2465375","url":null,"abstract":"Library OS (LibOS) architectures implement the OS personality as a user-mode library, giving each application the flexibility to choose its LibOS. This approach is appealing for many reasons, not least the ability to extend or customise the LibOS. Recent work with Drawbridge [29] showed that an existing commodity OS (Windows 7) could be refactored to produce a LibOS while retaining application compatibility.\u0000 This paper presents Bascule, an architecture for LibOS extensions based on Drawbridge. Rather than relying on the application developer to customise a LibOS, Bascule allows OS-independent extensions to be attached at runtime. Extensions interpose on a narrow binary interface of primitive OS abstractions, such as files and virtual memory. Thus, they are independent of both guest and host OS, and composable at runtime. Since an extension runs in the same process as an application and its LibOS, it is safe and efficient.\u0000 Bascule demonstrates extension reuse across diverse guest LibOSes (Windows and Linux) and host OSes (Windows and Barrelfish). Current extensions include file system translation, checkpointing, and architecture adaptation.","PeriodicalId":20737,"journal":{"name":"Proceedings of the Eleventh European Conference on Computer Systems","volume":"374 1","pages":"239-252"},"PeriodicalIF":0.0,"publicationDate":"2013-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77985937","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Mihir Nanavati, Mark Spear, Nathan Taylor, Shriram Rajagopalan, Dutch T. Meyer, W. Aiello, A. Warfield
As hardware parallelism continues to increase, CPU caches can no longer be considered as a transparent, hardware-level performance optimization. Cache impact on performance, in particular in the face of false sharing, is completely dependent on the software that is executing. To effectively support parallel workloads on cache coherent hardware, the operating system must begin to treat the CPU cache like other shared hardware resources, and manage it appropriately. We demonstrate a prototype example of such support by describing Plastic, a software-based system that detects, diagnoses, and transparently repairs false sharing as it occurs in running applications. Plastic solves two challenging problems. First, it is capable of rapid, low-overhead detection and diagnosis of false sharing in unmodified, running applications. Second, it resolves identified instances of false sharing by providing a sub-page granularity memory remapping facility within the system. Our implementation is capable of identifying and repairing pathological false sharing in under one second of execution and achieves speedups of 3-6x on known examples of false sharing in parallel benchmarks.
{"title":"Whose cache line is it anyway?: operating system support for live detection and repair of false sharing","authors":"Mihir Nanavati, Mark Spear, Nathan Taylor, Shriram Rajagopalan, Dutch T. Meyer, W. Aiello, A. Warfield","doi":"10.1145/2465351.2465366","DOIUrl":"https://doi.org/10.1145/2465351.2465366","url":null,"abstract":"As hardware parallelism continues to increase, CPU caches can no longer be considered as a transparent, hardware-level performance optimization. Cache impact on performance, in particular in the face of false sharing, is completely dependent on the software that is executing. To effectively support parallel workloads on cache coherent hardware, the operating system must begin to treat the CPU cache like other shared hardware resources, and manage it appropriately.\u0000 We demonstrate a prototype example of such support by describing Plastic, a software-based system that detects, diagnoses, and transparently repairs false sharing as it occurs in running applications. Plastic solves two challenging problems. First, it is capable of rapid, low-overhead detection and diagnosis of false sharing in unmodified, running applications. Second, it resolves identified instances of false sharing by providing a sub-page granularity memory remapping facility within the system. Our implementation is capable of identifying and repairing pathological false sharing in under one second of execution and achieves speedups of 3-6x on known examples of false sharing in parallel benchmarks.","PeriodicalId":20737,"journal":{"name":"Proceedings of the Eleventh European Conference on Computer Systems","volume":"33 1","pages":"141-154"},"PeriodicalIF":0.0,"publicationDate":"2013-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76678992","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Cloud-scale storage applications have strict requirements. On the one hand, they require scalable throughput; on the other hand, many applications would largely benefit from strong consistency. Since these requirements are sometimes considered contradictory, the subject has split the community with one side defending scalability at any cost (the "NoSQL" side), and the other side holding on time-proven transactional storage systems (the "SQL" side). In this paper, we present Augustus, a system that aims to bridge the sides by offering low-cost transactions with strong consistency and scalable throughput. Furthermore, Augustus assumes Byzantine failures to ensure data consistency even in the most hostile environments. We evaluated Augustus with a suite of micro-benchmarks, Buzzer (a Twitter-like service), and BFT Derby (an SQL engine based on Apache Derby).
{"title":"Augustus: scalable and robust storage for cloud applications","authors":"Ricardo Padilha, F. Pedone","doi":"10.1145/2465351.2465362","DOIUrl":"https://doi.org/10.1145/2465351.2465362","url":null,"abstract":"Cloud-scale storage applications have strict requirements. On the one hand, they require scalable throughput; on the other hand, many applications would largely benefit from strong consistency. Since these requirements are sometimes considered contradictory, the subject has split the community with one side defending scalability at any cost (the \"NoSQL\" side), and the other side holding on time-proven transactional storage systems (the \"SQL\" side). In this paper, we present Augustus, a system that aims to bridge the sides by offering low-cost transactions with strong consistency and scalable throughput. Furthermore, Augustus assumes Byzantine failures to ensure data consistency even in the most hostile environments. We evaluated Augustus with a suite of micro-benchmarks, Buzzer (a Twitter-like service), and BFT Derby (an SQL engine based on Apache Derby).","PeriodicalId":20737,"journal":{"name":"Proceedings of the Eleventh European Conference on Computer Systems","volume":"62 1","pages":"99-112"},"PeriodicalIF":0.0,"publicationDate":"2013-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88080817","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
RadixVM is a new virtual memory system design that enables fully concurrent operations on shared address spaces for multithreaded processes on cache-coherent multicore computers. Today, most operating systems serialize operations such as mmap and munmap, which forces application developers to split their multithreaded applications into multiprocess applications, hoard memory to avoid the overhead of returning it, and so on. RadixVM removes this burden from application developers by ensuring that address space operations on non-overlapping memory regions scale perfectly. It does so by combining three techniques: 1) it organizes metadata in a radix tree instead of a balanced tree to avoid unnecessary cache line movement; 2) it uses a novel memory-efficient distributed reference counting scheme; and 3) it uses a new scheme to target remote TLB shootdowns and to often avoid them altogether. Experiments on an 80 core machine show that RadixVM achieves perfect scalability for non-overlapping regions: if several threads mmap or munmap pages in parallel, they can run completely independently and induce no cache coherence traffic.
{"title":"RadixVM: scalable address spaces for multithreaded applications","authors":"A. Clements, M. Kaashoek, N. Zeldovich","doi":"10.1145/2465351.2465373","DOIUrl":"https://doi.org/10.1145/2465351.2465373","url":null,"abstract":"RadixVM is a new virtual memory system design that enables fully concurrent operations on shared address spaces for multithreaded processes on cache-coherent multicore computers. Today, most operating systems serialize operations such as mmap and munmap, which forces application developers to split their multithreaded applications into multiprocess applications, hoard memory to avoid the overhead of returning it, and so on. RadixVM removes this burden from application developers by ensuring that address space operations on non-overlapping memory regions scale perfectly. It does so by combining three techniques: 1) it organizes metadata in a radix tree instead of a balanced tree to avoid unnecessary cache line movement; 2) it uses a novel memory-efficient distributed reference counting scheme; and 3) it uses a new scheme to target remote TLB shootdowns and to often avoid them altogether. Experiments on an 80 core machine show that RadixVM achieves perfect scalability for non-overlapping regions: if several threads mmap or munmap pages in parallel, they can run completely independently and induce no cache coherence traffic.","PeriodicalId":20737,"journal":{"name":"Proceedings of the Eleventh European Conference on Computer Systems","volume":"27 1","pages":"211-224"},"PeriodicalIF":0.0,"publicationDate":"2013-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89596517","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Preserving the integrity of application data across updates is difficult if power outages and system crashes may occur during updates. Existing approaches such as relational databases and transactional key-value stores restrict programming flexibility by mandating narrow data access interfaces. We have designed, implemented, and evaluated an approach that strengthens the semantics of a standard operating system primitive while maintaining conceptual simplicity and supporting highly flexible programming: Failureatomic msync() commits changes to a memory-mapped file atomically, even in the presence of failures. Our Linux implementation of failure-atomic msync() has preserved application data integrity across hundreds of whole-machine power interruptions and exhibits good microbenchmark performance on both spinning disks and solid-state storage. Failure-atomic msync() supports higher layers of fully general programming abstraction, e.g., a persistent heap that easily slips beneath the C++ Standard Template Library. An STL
{"title":"Failure-atomic msync(): a simple and efficient mechanism for preserving the integrity of durable data","authors":"Stan Park, T. Kelly, Kai Shen","doi":"10.1145/2465351.2465374","DOIUrl":"https://doi.org/10.1145/2465351.2465374","url":null,"abstract":"Preserving the integrity of application data across updates is difficult if power outages and system crashes may occur during updates. Existing approaches such as relational databases and transactional key-value stores restrict programming flexibility by mandating narrow data access interfaces. We have designed, implemented, and evaluated an approach that strengthens the semantics of a standard operating system primitive while maintaining conceptual simplicity and supporting highly flexible programming: Failureatomic msync() commits changes to a memory-mapped file atomically, even in the presence of failures. Our Linux implementation of failure-atomic msync() has preserved application data integrity across hundreds of whole-machine power interruptions and exhibits good microbenchmark performance on both spinning disks and solid-state storage. Failure-atomic msync() supports higher layers of fully general programming abstraction, e.g., a persistent heap that easily slips beneath the C++ Standard Template Library. An STL <map> built atop failure-atomic msync() outperforms several local key-value stores that support transactional updates. We integrated failure-atomic msync() into the Kyoto Tycoon key-value server by modifying exactly one line of code; our modified server reduces response times by 26--43% compared to Tycoon's existing transaction support while providing the same data integrity guarantees. Compared to a Tycoon server setup that makes almost no I/O (and therefore provides no support for data durability and integrity over failures), failure-atomic msync() incurs a three-fold response time increase on a fast Flash-based SSD---an acceptable cost of data reliability for many.","PeriodicalId":20737,"journal":{"name":"Proceedings of the Eleventh European Conference on Computer Systems","volume":"41 1","pages":"225-238"},"PeriodicalIF":0.0,"publicationDate":"2013-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82720360","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：