Pub Date : 2019-10-01DOI: 10.1109/SRDSW49218.2019.00012
Tânia Esteves, Ricardo Macedo, Alberto Faria, Bernardo Portela, J. Paulo, J. Pereira, Danny Harnik
Data confidentiality in cloud services is commonly ensured by encrypting information before uploading it. However, this approach limits the use of content-aware functionalities, such as deduplication and compression. Although this issue has been addressed individually for some of these functionalities, no unified framework for building secure storage systems exists that can leverage such operations over encrypted data. We present TrustFS, a programmable and modular stackable file system framework for implementing secure content-aware storage functionalities over hardware-assisted trusted execution environments. This framework extends the original SafeFS architecture to provide the isolated execution guarantees of Intel SGX. We demonstrate its usability by implementing an SGX-enabled stackable file system prototype while a preliminary evaluation shows that it incurs reasonable performance overhead when compared to conventional storage systems. Finally, we highlight open research challenges that must be further pursued in order for TrustFS to be fully adequate for building production-ready secure storage solutions.
{"title":"TrustFS: An SGX-Enabled Stackable File System Framework","authors":"Tânia Esteves, Ricardo Macedo, Alberto Faria, Bernardo Portela, J. Paulo, J. Pereira, Danny Harnik","doi":"10.1109/SRDSW49218.2019.00012","DOIUrl":"https://doi.org/10.1109/SRDSW49218.2019.00012","url":null,"abstract":"Data confidentiality in cloud services is commonly ensured by encrypting information before uploading it. However, this approach limits the use of content-aware functionalities, such as deduplication and compression. Although this issue has been addressed individually for some of these functionalities, no unified framework for building secure storage systems exists that can leverage such operations over encrypted data. We present TrustFS, a programmable and modular stackable file system framework for implementing secure content-aware storage functionalities over hardware-assisted trusted execution environments. This framework extends the original SafeFS architecture to provide the isolated execution guarantees of Intel SGX. We demonstrate its usability by implementing an SGX-enabled stackable file system prototype while a preliminary evaluation shows that it incurs reasonable performance overhead when compared to conventional storage systems. Finally, we highlight open research challenges that must be further pursued in order for TrustFS to be fully adequate for building production-ready secure storage solutions.","PeriodicalId":297328,"journal":{"name":"2019 38th International Symposium on Reliable Distributed Systems Workshops (SRDSW)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129403811","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}
Pub Date : 2019-10-01DOI: 10.1109/SRDSW49218.2019.00017
J. Ellul, Gordon J. Pace
Developing smart contract decentralised application based systems typically involves writing code for various platforms, from the smart contract code residing on the underlying distributed ledger technology implementation to back end oracles and front end websites or mobile apps. In addition to the different technologies used for the different parts, the programmer is also burdened with implementing communication channels between the various parts. In this paper we propose a unified programming model allowing for developers to build such systems through a single code artifact, using a macroprogramming approach.
{"title":"Towards A Unified Programming Model for Blockchain Smart Contract dApp Systems","authors":"J. Ellul, Gordon J. Pace","doi":"10.1109/SRDSW49218.2019.00017","DOIUrl":"https://doi.org/10.1109/SRDSW49218.2019.00017","url":null,"abstract":"Developing smart contract decentralised application based systems typically involves writing code for various platforms, from the smart contract code residing on the underlying distributed ledger technology implementation to back end oracles and front end websites or mobile apps. In addition to the different technologies used for the different parts, the programmer is also burdened with implementing communication channels between the various parts. In this paper we propose a unified programming model allowing for developers to build such systems through a single code artifact, using a macroprogramming approach.","PeriodicalId":297328,"journal":{"name":"2019 38th International Symposium on Reliable Distributed Systems Workshops (SRDSW)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131646115","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}
Pub Date : 2019-10-01DOI: 10.1109/SRDSW49218.2019.00016
Thiago Luiz Gontijo de Almeida, Pierre Francois, S. Frénot
In this article, we shed light on the peer-to-peer networks which support public blockchains stemming from Bitcoin forks. While the Bitcoin network has undergone a lot of attention, little has been discovered on the size, geographical spread, and general dependability of the networks supporting such forks. In this paper, we first discuss the various types of Bitcoin forks. We identify the case of hard forks which essentially consist in independent crypto-currencies that become completely separated from the reference Bitcoin network. We present a set of tools that are used to gather information on the Bitcoin forks networks. Finally, we provide preliminary analysis results regarding the size, IP layer localization, concentration, and overlap of ten Bitcoin forks.
{"title":"Forkmon: Monitoring the Networks Supporting Bitcoin Hard Forks","authors":"Thiago Luiz Gontijo de Almeida, Pierre Francois, S. Frénot","doi":"10.1109/SRDSW49218.2019.00016","DOIUrl":"https://doi.org/10.1109/SRDSW49218.2019.00016","url":null,"abstract":"In this article, we shed light on the peer-to-peer networks which support public blockchains stemming from Bitcoin forks. While the Bitcoin network has undergone a lot of attention, little has been discovered on the size, geographical spread, and general dependability of the networks supporting such forks. In this paper, we first discuss the various types of Bitcoin forks. We identify the case of hard forks which essentially consist in independent crypto-currencies that become completely separated from the reference Bitcoin network. We present a set of tools that are used to gather information on the Bitcoin forks networks. Finally, we provide preliminary analysis results regarding the size, IP layer localization, concentration, and overlap of ten Bitcoin forks.","PeriodicalId":297328,"journal":{"name":"2019 38th International Symposium on Reliable Distributed Systems Workshops (SRDSW)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132668157","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}
Pub Date : 2019-10-01DOI: 10.1109/SRDSW49218.2019.00015
Liuyang Ren, Paul A. S. Ward
The decentralization property of blockchains stems from the fact that each miner accepts or refuses transactions and blocks based on its own verification results. However, pooled mining causes blockchains to evolve into centralized systems because pool participants delegate their decision-making rights to pool managers. In this paper, we established and validated a model for Proof-of-Work mining, introduced the concept of equivalent blocks, and quantitatively derived that pooling effectively lowers the income variance of miners. We also analyzed Bitcoin and Ethereum data to prove that pooled mining has become prevalent in the real world. The percentage of pool-mined blocks increased from 49.91% to 91.12% within four months in Bitcoin and from 76.9% to 92.2% within five months in Ethereum. In July 2018, Bitcoin and Ethereum mining were dominated by only six and five pools respectively.
{"title":"Pooled Mining is Driving Blockchains Toward Centralized Systems","authors":"Liuyang Ren, Paul A. S. Ward","doi":"10.1109/SRDSW49218.2019.00015","DOIUrl":"https://doi.org/10.1109/SRDSW49218.2019.00015","url":null,"abstract":"The decentralization property of blockchains stems from the fact that each miner accepts or refuses transactions and blocks based on its own verification results. However, pooled mining causes blockchains to evolve into centralized systems because pool participants delegate their decision-making rights to pool managers. In this paper, we established and validated a model for Proof-of-Work mining, introduced the concept of equivalent blocks, and quantitatively derived that pooling effectively lowers the income variance of miners. We also analyzed Bitcoin and Ethereum data to prove that pooled mining has become prevalent in the real world. The percentage of pool-mined blocks increased from 49.91% to 91.12% within four months in Bitcoin and from 76.9% to 92.2% within five months in Ethereum. In July 2018, Bitcoin and Ethereum mining were dominated by only six and five pools respectively.","PeriodicalId":297328,"journal":{"name":"2019 38th International Symposium on Reliable Distributed Systems Workshops (SRDSW)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123483398","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}
Pub Date : 2019-10-01DOI: 10.1109/SRDSW49218.2019.00010
Xin Xie, Chentao Wu, Chao Li, Jie Li, M. Guo, Fang Xu
With the development of cloud computing, the reliability of disk arrays are increasingly concerned. Data centers usually use erasure codes to provide high reliability. However, most of reconstruction methods on disk arrays focus on single/ multiple disk(s) recovery, which ignores how to efficiently reconstruct the lost data such as Latent Sector Errors (LSEs), etc. In real situations, local stripe errors are much more common than disk failures. It has become an urgent problem that how to improve reconstruction efficiently for stripes. This paper proposes a comprehensive rearranging priority reconstruction(CRPR), which combines temporal locality, spatial locality and coding characteristics together. CRPR divides different blocks into various priorities and recovers them sequentially. To demonstrate the effectiveness of CRPR, we conduct several simulations via disksim. The simulations results show that, the comprehensive rearranging priority reconstruction method keeps up with previous methods and can save up to 63.9% in terms of waiting time.
{"title":"A Comprehensive Rearranging Priority Based Method To Accelerate the Reconstruction of RAID Arrays","authors":"Xin Xie, Chentao Wu, Chao Li, Jie Li, M. Guo, Fang Xu","doi":"10.1109/SRDSW49218.2019.00010","DOIUrl":"https://doi.org/10.1109/SRDSW49218.2019.00010","url":null,"abstract":"With the development of cloud computing, the reliability of disk arrays are increasingly concerned. Data centers usually use erasure codes to provide high reliability. However, most of reconstruction methods on disk arrays focus on single/ multiple disk(s) recovery, which ignores how to efficiently reconstruct the lost data such as Latent Sector Errors (LSEs), etc. In real situations, local stripe errors are much more common than disk failures. It has become an urgent problem that how to improve reconstruction efficiently for stripes. This paper proposes a comprehensive rearranging priority reconstruction(CRPR), which combines temporal locality, spatial locality and coding characteristics together. CRPR divides different blocks into various priorities and recovers them sequentially. To demonstrate the effectiveness of CRPR, we conduct several simulations via disksim. The simulations results show that, the comprehensive rearranging priority reconstruction method keeps up with previous methods and can save up to 63.9% in terms of waiting time.","PeriodicalId":297328,"journal":{"name":"2019 38th International Symposium on Reliable Distributed Systems Workshops (SRDSW)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127681304","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}
Pub Date : 2019-10-01DOI: 10.1109/SRDSW49218.2019.00011
V. Cogo, A. Bessani
Efficiently storing large data sets of human genomes is a long-term ambition from both the research and clinical life sciences communities. For instance, biobanks stock thousands to millions of biological physical samples and have been under pressure to store also their resulting digitized genomes. However, these and other life sciences institutions lack the infrastructure and expertise to efficiently store this data. Cloud computing is a natural economic alternative to private infrastructures, but it is not as good an alternative in terms of security and privacy. In this work, we present an end-to-end composite pipeline intended to enable the efficient, dependable cloud-based storage of human genomes by integrating three mechanisms we have recently proposed. These mechanisms encompass (1) a privacy-sensitivity detector for human genomes, (2) a similarity-based deduplication and delta-encoding algorithm for sequencing data, and (3) an auditability scheme to verify who has effectively read data in storage systems that use secure information dispersal. By integrating them with appropriate storage configurations, one can obtain reasonable privacy protection, security, and dependability guarantees at modest costs (e.g., less than $1/Genome/Year). Our preliminary analysis indicates that this pipeline costs only 3% more than non-replicated systems, 48% less than fully-replicating all data, and 31% less than secure information dispersal schemes.
{"title":"Enabling the Efficient, Dependable Cloud-Based Storage of Human Genomes","authors":"V. Cogo, A. Bessani","doi":"10.1109/SRDSW49218.2019.00011","DOIUrl":"https://doi.org/10.1109/SRDSW49218.2019.00011","url":null,"abstract":"Efficiently storing large data sets of human genomes is a long-term ambition from both the research and clinical life sciences communities. For instance, biobanks stock thousands to millions of biological physical samples and have been under pressure to store also their resulting digitized genomes. However, these and other life sciences institutions lack the infrastructure and expertise to efficiently store this data. Cloud computing is a natural economic alternative to private infrastructures, but it is not as good an alternative in terms of security and privacy. In this work, we present an end-to-end composite pipeline intended to enable the efficient, dependable cloud-based storage of human genomes by integrating three mechanisms we have recently proposed. These mechanisms encompass (1) a privacy-sensitivity detector for human genomes, (2) a similarity-based deduplication and delta-encoding algorithm for sequencing data, and (3) an auditability scheme to verify who has effectively read data in storage systems that use secure information dispersal. By integrating them with appropriate storage configurations, one can obtain reasonable privacy protection, security, and dependability guarantees at modest costs (e.g., less than $1/Genome/Year). Our preliminary analysis indicates that this pipeline costs only 3% more than non-replicated systems, 48% less than fully-replicating all data, and 31% less than secure information dispersal schemes.","PeriodicalId":297328,"journal":{"name":"2019 38th International Symposium on Reliable Distributed Systems Workshops (SRDSW)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127183983","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}
Pub Date : 2019-10-01DOI: 10.1109/SRDSW49218.2019.00014
Shota Numakura, Junya Nakamura, Ren Ohmura
Geographic state machine replication (SMR) is a replication method in which replicas of a service are located on multiple continents to improve the fault tolerance of a general service. Nowadays, geographic SMR is easily realized using public cloud services; SMR provides extraordinary resilience against catastrophic disasters. Previous studies have revealed that the geographic distribution of the replicas has a significant influence on the performance of the geographic SMR; however, the optimal way for a system integrator to deploy replicas remains unknown. In this paper, we propose a method to evaluate and rank replica deployments to assist a system integrator in deciding a final replica deployment. In the method, we also propose a novel evaluation function that estimates a latency of SMR protocols with round-trip time (RTT). To demonstrate the effectiveness of the proposed method, we build thousands of geographic SMRs on Amazon Web Services and present experimental results. The results show that the proposed method that estimates a latency based on RTTs can generate consistent rankings with reasonable calculation time.
地理状态机复制(SMR)是一种复制方法，其中服务的副本位于多个大洲，以提高一般服务的容错性。如今，地理SMR很容易通过公共云服务实现;SMR提供了非凡的抗灾能力。已有研究表明，副本的地理分布对地理SMR的绩效有显著影响;然而，系统集成商部署副本的最佳方式仍然未知。在本文中，我们提出了一种评估和排序副本部署的方法，以帮助系统集成商决定最终的副本部署。在该方法中，我们还提出了一个新的评估函数，该函数可以用往返时间(RTT)来估计SMR协议的延迟。为了证明该方法的有效性，我们在Amazon Web Services上构建了数千个地理smr，并给出了实验结果。结果表明，基于rtt估计时延的方法可以在合理的计算时间内生成一致的排名。
{"title":"Evaluation and Ranking of Replica Deployments in Geographic State Machine Replication","authors":"Shota Numakura, Junya Nakamura, Ren Ohmura","doi":"10.1109/SRDSW49218.2019.00014","DOIUrl":"https://doi.org/10.1109/SRDSW49218.2019.00014","url":null,"abstract":"Geographic state machine replication (SMR) is a replication method in which replicas of a service are located on multiple continents to improve the fault tolerance of a general service. Nowadays, geographic SMR is easily realized using public cloud services; SMR provides extraordinary resilience against catastrophic disasters. Previous studies have revealed that the geographic distribution of the replicas has a significant influence on the performance of the geographic SMR; however, the optimal way for a system integrator to deploy replicas remains unknown. In this paper, we propose a method to evaluate and rank replica deployments to assist a system integrator in deciding a final replica deployment. In the method, we also propose a novel evaluation function that estimates a latency of SMR protocols with round-trip time (RTT). To demonstrate the effectiveness of the proposed method, we build thousands of geographic SMRs on Amazon Web Services and present experimental results. The results show that the proposed method that estimates a latency based on RTTs can generate consistent rankings with reasonable calculation time.","PeriodicalId":297328,"journal":{"name":"2019 38th International Symposium on Reliable Distributed Systems Workshops (SRDSW)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125165048","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}
Pub Date : 2019-10-01DOI: 10.1109/SRDSW49218.2019.00008
Aris Chronarakis, Antonis Papaioannou, K. Magoutis
Incremental checkpointing (IC) is a fault-tolerance technique used in several stateful distributed stream processing systems. It relies on continuously logging state updates to a remote storage service and periodically compacting the update-log via a background process. We highlight a tradeoff between the intensity of compaction of the IC update-log (and the associated resource overhead) and its impact on recovery time in such systems. We also highlight the control parameters that can be used to adjust this tradeoff in the Apache Samza stream processing system, and demonstrate this tradeoff experimentally.
{"title":"On the Impact of log Compaction on Incrementally Checkpointing Stateful Stream-Processing Operators","authors":"Aris Chronarakis, Antonis Papaioannou, K. Magoutis","doi":"10.1109/SRDSW49218.2019.00008","DOIUrl":"https://doi.org/10.1109/SRDSW49218.2019.00008","url":null,"abstract":"Incremental checkpointing (IC) is a fault-tolerance technique used in several stateful distributed stream processing systems. It relies on continuously logging state updates to a remote storage service and periodically compacting the update-log via a background process. We highlight a tradeoff between the intensity of compaction of the IC update-log (and the associated resource overhead) and its impact on recovery time in such systems. We also highlight the control parameters that can be used to adjust this tradeoff in the Apache Samza stream processing system, and demonstrate this tradeoff experimentally.","PeriodicalId":297328,"journal":{"name":"2019 38th International Symposium on Reliable Distributed Systems Workshops (SRDSW)","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127240107","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}
Security Information and Event Management (SIEM) systems have been adopted by organizations to enable holistic monitoring of malicious activities in their IT infrastructures. SIEMs receive events from diverse devices of the organization's IT infrastructure (e.g., servers, firewalls, IDS), correlate these events, and present reports for security analysts. Given the large number of events collected by SIEMs, it is costly to store such data for long periods. Since organizations store a relatively limited time-frame of events, the forensic analysis capabilities severely become reduced. This concern limits the organizations' ability to store important information about the past cybersecurity-related activity, limiting forensic analysis. A possible solution for this issue is to leverage public cloud storage services, exploiting their low cost and "infinite" scalability. We present SLiCER an archival system for long-term storage that makes use of a multi-cloud-based storage system to guarantee data security and ensures cost-effectiveness by grouping events in blocks and using indexing techniques to recover them. The system was evaluated using a real dataset and the results show that it is significantly more cost-efficient than competing alternatives.
{"title":"A Cost-Effective Cloud Event Archival for SIEMs","authors":"Adriano Serckumecka, Ibéria Medeiros, Bernardo Ferreira","doi":"10.1109/SRDSW49218.2019.00013","DOIUrl":"https://doi.org/10.1109/SRDSW49218.2019.00013","url":null,"abstract":"Security Information and Event Management (SIEM) systems have been adopted by organizations to enable holistic monitoring of malicious activities in their IT infrastructures. SIEMs receive events from diverse devices of the organization's IT infrastructure (e.g., servers, firewalls, IDS), correlate these events, and present reports for security analysts. Given the large number of events collected by SIEMs, it is costly to store such data for long periods. Since organizations store a relatively limited time-frame of events, the forensic analysis capabilities severely become reduced. This concern limits the organizations' ability to store important information about the past cybersecurity-related activity, limiting forensic analysis. A possible solution for this issue is to leverage public cloud storage services, exploiting their low cost and \"infinite\" scalability. We present SLiCER an archival system for long-term storage that makes use of a multi-cloud-based storage system to guarantee data security and ensures cost-effectiveness by grouping events in blocks and using indexing techniques to recover them. The system was evaluated using a real dataset and the results show that it is significantly more cost-efficient than competing alternatives.","PeriodicalId":297328,"journal":{"name":"2019 38th International Symposium on Reliable Distributed Systems Workshops (SRDSW)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129242754","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}
Pub Date : 2019-10-01DOI: 10.1109/SRDSW49218.2019.00009
Ricardo Macedo, Alberto Faria, J. Paulo, J. Pereira
Modern storage infrastructures feature long and complicated I/O paths composed of several layers, each employing their own optimizations to serve varied applications with fluctuating requirements. However, as these layers do not have global infrastructure visibility, they are unable to optimally tune their behavior to achieve maximum performance. Background storage tasks, in particular, can rapidly overload shared resources, but are executed either periodically or whenever a certain threshold is hit regardless of the overall load on the system. In this paper, we argue that to achieve optimal holistic performance, these tasks should be dynamically programmable and handled by a controller with global visibility. To support this argument, we evaluate the impact on performance of compaction and checkpointing in the context of HBase and PostgreSQL. We find that these tasks can respectively increase 99th percentile latencies by 955.2% and 61.9%. We also identify future research directions to achieve programmable background tasks.
{"title":"A Case for Dynamically Programmable Storage Background Tasks","authors":"Ricardo Macedo, Alberto Faria, J. Paulo, J. Pereira","doi":"10.1109/SRDSW49218.2019.00009","DOIUrl":"https://doi.org/10.1109/SRDSW49218.2019.00009","url":null,"abstract":"Modern storage infrastructures feature long and complicated I/O paths composed of several layers, each employing their own optimizations to serve varied applications with fluctuating requirements. However, as these layers do not have global infrastructure visibility, they are unable to optimally tune their behavior to achieve maximum performance. Background storage tasks, in particular, can rapidly overload shared resources, but are executed either periodically or whenever a certain threshold is hit regardless of the overall load on the system. In this paper, we argue that to achieve optimal holistic performance, these tasks should be dynamically programmable and handled by a controller with global visibility. To support this argument, we evaluate the impact on performance of compaction and checkpointing in the context of HBase and PostgreSQL. We find that these tasks can respectively increase 99th percentile latencies by 955.2% and 61.9%. We also identify future research directions to achieve programmable background tasks.","PeriodicalId":297328,"journal":{"name":"2019 38th International Symposium on Reliable Distributed Systems Workshops (SRDSW)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123135527","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：