基于DB2的DBaaS系统中计算资源隔离方法研究与实现
|
摘要
计算资源隔离是云计算领域的一个重要研究课题,通常应用虚拟机技术来解决该问题。数据云(Database as a Service,简称DBaaS)属于云计算平台的一种,其中的计算资源隔离一般也应用虚拟机技术,用户的数据库服务运行在虚拟机上,用户以虚拟机为单位租用服务。但虚拟机技术有较多缺点,如资源浪费、性能降低、计算资源分配不灵活等。
本文提出了一种新的计算资源隔离方案,用于基于DB2的DBaaS系统中。用户以数据库为单位租用服务,多个用户的数据库运行在同一实例下。它是一种更细粒度的计算资源隔离方案,较之于虚拟机的解决方案,它的资源共享率更高、性能更高,资源配置更灵活。本文提出了虚拟资源容器DBR,用于不同用户之间的计算资源隔离,这些计算资源包括CPU、内存、磁盘I/O速度、磁盘使用量。DBR运行于操作系统上,可以动态的创建,删除,修改计算资源参数。每个数据库用户对应一个DBR,用户的所有任务运行在自己的DBR中,在DBR中为每个用户分配资源,用户任务受到DBR资源阈值的控制。DBR可以被集成到基于DB2的DBaaS系统中,经实验表明DBR可以对不同用户的计算资源进行隔离,并提高用户性能的可预见性,能够保证用户服务质量。
One of the most important research subject in cloud computing is computing resource isolation. The traditional solution is using virtual machine technology. Database as s service is cloud computing paltform which also use virtual machine technology to islolate computing resource. Database service runs on virtual machine. Customers subscibe database service by virtual machine. Virtual machine has some drawbacks,waste of resource, low performance and inconvenience of resource distribution .etc.
A new resolution of computing resource isolation is proposed in this paper,which is used in DB2 based Database as a Service system. Customers subscribe database service by database rather than virtual machine and different customers may share the same DB2 instance. It is a fine graind solution compared to virtual machine under which computing resource can be shared more common and used more effective,thus the performance is high. We put forward a concept called DBR(Database Resource Unit) in this paper which includes CPU,memory,disk I/O bandwidth and disk space isolation. Eevry customer has its own DBR corresponding to their database, whose tasks will run in it. DBR runs on Linux Operating System which can be created, distroyed dynamically and its parameter can also be configured dynamically. Customer’s computing resource is allocated to DBR and their tasks are constrained by DBR. DBR can be integrated into DB2 based Database as a Service. Experiment results have proved that DBR can effectively isolate customer’s computing resource, improve performance prediction and guarantee service quality.
本文提出了一种新的计算资源隔离方案,用于基于DB2的DBaaS系统中。用户以数据库为单位租用服务,多个用户的数据库运行在同一实例下。它是一种更细粒度的计算资源隔离方案,较之于虚拟机的解决方案,它的资源共享率更高、性能更高,资源配置更灵活。本文提出了虚拟资源容器DBR,用于不同用户之间的计算资源隔离,这些计算资源包括CPU、内存、磁盘I/O速度、磁盘使用量。DBR运行于操作系统上,可以动态的创建,删除,修改计算资源参数。每个数据库用户对应一个DBR,用户的所有任务运行在自己的DBR中,在DBR中为每个用户分配资源,用户任务受到DBR资源阈值的控制。DBR可以被集成到基于DB2的DBaaS系统中,经实验表明DBR可以对不同用户的计算资源进行隔离,并提高用户性能的可预见性,能够保证用户服务质量。
One of the most important research subject in cloud computing is computing resource isolation. The traditional solution is using virtual machine technology. Database as s service is cloud computing paltform which also use virtual machine technology to islolate computing resource. Database service runs on virtual machine. Customers subscibe database service by virtual machine. Virtual machine has some drawbacks,waste of resource, low performance and inconvenience of resource distribution .etc.
A new resolution of computing resource isolation is proposed in this paper,which is used in DB2 based Database as a Service system. Customers subscribe database service by database rather than virtual machine and different customers may share the same DB2 instance. It is a fine graind solution compared to virtual machine under which computing resource can be shared more common and used more effective,thus the performance is high. We put forward a concept called DBR(Database Resource Unit) in this paper which includes CPU,memory,disk I/O bandwidth and disk space isolation. Eevry customer has its own DBR corresponding to their database, whose tasks will run in it. DBR runs on Linux Operating System which can be created, distroyed dynamically and its parameter can also be configured dynamically. Customer’s computing resource is allocated to DBR and their tasks are constrained by DBR. DBR can be integrated into DB2 based Database as a Service. Experiment results have proved that DBR can effectively isolate customer’s computing resource, improve performance prediction and guarantee service quality.
引文
[1]Michael Armbrust, Armando Fox, Rean Griffith, A View of Cloud Computing. communications of the acm,.vol 53 no 4, 2010-4
[2]B arroso, L.A., and Holzle, U. The case for energyproportional computing. IEEE Computer 40, 2007-12
[3]B rodkin, J. Loss of customer data spurs closure of online storage service’The Linkup.’Network World, 2008-8
[4]Freedom OS, Large data set transfer to the cloud, http://freedomoss.com/clouddataingestion,2009-4.
[5]G arfinkel, S. An Evaluation of Amazon’s Grid Computing Services: EC2, S3 and SQS. Tech. Rep. TR-08-07, Harvard University, 2007-8
[6]McCalpin, J. Memory bandwidth and machine balance in current high performance computers. IEEE Technical Committee on Computer Architecture Newsletter 1995, 19–25
[7] B. F. Cooper, A. Silberstein, E. Tam, R. Ramakrishnan, and R. Sears. Benchmarking cloud serving systems with ycsb.SoCC, 2010
[8]C. Curino, E. Jones, Y. Zhang, E. Wu, and S. Madden. Relationalcloud: The case for a database service, New England Database Summit, 2010
[9]D. DeWitt and J. Gray. Parallel database systems: the future of high performance database systems. Comm. ACM, 1992
[10]S. Ghandeharizadeh and D. J. DeWitt. Hybrid-range partitioning strategy: a new declustering strategy for multiprocessor databases machines. In VLDB, 1990
[11]M. Koyut¨urk and C. Aykanat. Iterative-improvement-based declustering heuristics for multi-disk databases. Journal of Information Systems, 2005. 30(1):47–70
[12]J. Rao, C. Zhang, N. Megiddo, and G. Lohman. Automating physical database design in a parallel database. In SIGMOD, 2002
[13]D. C. Zilio. Physical database design decision algorithms and concurrent reorganization for parallel database systems. In PhD thesis, 1998
[14]Yu Chen Zhou, Xin Peng Liu, Xi Ning Wang, Business Process Centric Platform-as-a-Service Model and Technologies for Cloud Enabled Industry Solutions, IEEE 3rd International Conference on Cloud Computing, 2010. 1-3
[15]Michael Armbrust, Armando Fox, Rean Griffith, A View of Cloud Computing. In:communications of the acm, vol 53 no 4, 2010-4
[16]Iperf, Network Throughput measurement tool, http://sourceforge.net/projects/iperf, Accessed April, 2009
[17]Schanzenbach, D. & Casanova, H. Accuracy and Responsiveness of CPU Sharing Using Xen’s Cap Values Computer and Information Sciences Dept., University of Hawai at manoa, 2008
[18]Weng, C.; Wang, Z.; Li, M. & Lu, X. The hybrid scheduling framework for virtual machine systems VEE '09: Proceedings of the 2009 ACM SIGPLAN/SIGOPS international conference on Virtual execution environments, ACM, 2009, 111-120
[19]Kim, H.; Lim, H.; Jeong, J.; Jo, H. & Lee, J. Task-aware virtual machine scheduling for I/O performance. VEE '09: Proceedings of the 2009 ACM SIGPLAN/SIGOPS international conference on Virtual execution environments, ACM, 2009, 101-110
[20]Menon, A.; Santos, J. R.; Turner, Y. Diagnosing performance overheads in the xen virtual machine environment VEE '05: Proceedings of the 1st ACM/USENIX international conference on Virtual execution environments, ACM,2005, 13-23
[21]Cherkasova, L. & Gardner, R.Measuring CPU overhead for I/O processing in the Xen virtual machine monitor ATEC '05: Proceedings of the annual conference on USENIX Annual Technical Conference, USENIX Association, 2005, 24-24
[22]Pradeep Padala, Kai-Yuan Hou, Automated Control of Multiple Virtualized Resources,HP Laboratories, 2008
[23]Chris Hyser, Bret McKee. Autonomic Virtual Machine Placement in the Data Center HP Laboratories, 2008-2
[24]Linux cgroups, http://www.kernel.org/doc/Documentation/cgroups/
[25]Leaky bucket, http://en.wikipedia.org/wiki/Leaky_bucket
[26]Whei-Jen Chen, Bill Comeau, H T Morgan. DB2 Workload Manager for Linux, UNIX, and Windows, 2008-7. 10-79
[27]Carlo Curino, Evan P.C.,Jones Raluca. Relational Cloud, A Database as a Service for the Cloud, 2005
[28]Carlo Curino,Evan Jones,Yang Zhang. Schism: a Workload Driven Approach to Database Replication and Partitioning, 2009
[29]Gaurav Somani,Sanjay Chaudhary. Application Performance Isolation in Virtualization. IEEE International Conference on Cloud Computing. 2009. 1-4
[30] Wenhong Tian, Sheng Su1, Guoming Lu. A Framework for Implementing and Managing Platform as a Service in a Virtual. Second International Workshop on Education Technology and Computer Science. 2010. 1-3
[2]B arroso, L.A., and Holzle, U. The case for energyproportional computing. IEEE Computer 40, 2007-12
[3]B rodkin, J. Loss of customer data spurs closure of online storage service’The Linkup.’Network World, 2008-8
[4]Freedom OS, Large data set transfer to the cloud, http://freedomoss.com/clouddataingestion,2009-4.
[5]G arfinkel, S. An Evaluation of Amazon’s Grid Computing Services: EC2, S3 and SQS. Tech. Rep. TR-08-07, Harvard University, 2007-8
[6]McCalpin, J. Memory bandwidth and machine balance in current high performance computers. IEEE Technical Committee on Computer Architecture Newsletter 1995, 19–25
[7] B. F. Cooper, A. Silberstein, E. Tam, R. Ramakrishnan, and R. Sears. Benchmarking cloud serving systems with ycsb.SoCC, 2010
[8]C. Curino, E. Jones, Y. Zhang, E. Wu, and S. Madden. Relationalcloud: The case for a database service, New England Database Summit, 2010
[9]D. DeWitt and J. Gray. Parallel database systems: the future of high performance database systems. Comm. ACM, 1992
[10]S. Ghandeharizadeh and D. J. DeWitt. Hybrid-range partitioning strategy: a new declustering strategy for multiprocessor databases machines. In VLDB, 1990
[11]M. Koyut¨urk and C. Aykanat. Iterative-improvement-based declustering heuristics for multi-disk databases. Journal of Information Systems, 2005. 30(1):47–70
[12]J. Rao, C. Zhang, N. Megiddo, and G. Lohman. Automating physical database design in a parallel database. In SIGMOD, 2002
[13]D. C. Zilio. Physical database design decision algorithms and concurrent reorganization for parallel database systems. In PhD thesis, 1998
[14]Yu Chen Zhou, Xin Peng Liu, Xi Ning Wang, Business Process Centric Platform-as-a-Service Model and Technologies for Cloud Enabled Industry Solutions, IEEE 3rd International Conference on Cloud Computing, 2010. 1-3
[15]Michael Armbrust, Armando Fox, Rean Griffith, A View of Cloud Computing. In:communications of the acm, vol 53 no 4, 2010-4
[16]Iperf, Network Throughput measurement tool, http://sourceforge.net/projects/iperf, Accessed April, 2009
[17]Schanzenbach, D. & Casanova, H. Accuracy and Responsiveness of CPU Sharing Using Xen’s Cap Values Computer and Information Sciences Dept., University of Hawai at manoa, 2008
[18]Weng, C.; Wang, Z.; Li, M. & Lu, X. The hybrid scheduling framework for virtual machine systems VEE '09: Proceedings of the 2009 ACM SIGPLAN/SIGOPS international conference on Virtual execution environments, ACM, 2009, 111-120
[19]Kim, H.; Lim, H.; Jeong, J.; Jo, H. & Lee, J. Task-aware virtual machine scheduling for I/O performance. VEE '09: Proceedings of the 2009 ACM SIGPLAN/SIGOPS international conference on Virtual execution environments, ACM, 2009, 101-110
[20]Menon, A.; Santos, J. R.; Turner, Y. Diagnosing performance overheads in the xen virtual machine environment VEE '05: Proceedings of the 1st ACM/USENIX international conference on Virtual execution environments, ACM,2005, 13-23
[21]Cherkasova, L. & Gardner, R.Measuring CPU overhead for I/O processing in the Xen virtual machine monitor ATEC '05: Proceedings of the annual conference on USENIX Annual Technical Conference, USENIX Association, 2005, 24-24
[22]Pradeep Padala, Kai-Yuan Hou, Automated Control of Multiple Virtualized Resources,HP Laboratories, 2008
[23]Chris Hyser, Bret McKee. Autonomic Virtual Machine Placement in the Data Center HP Laboratories, 2008-2
[24]Linux cgroups, http://www.kernel.org/doc/Documentation/cgroups/
[25]Leaky bucket, http://en.wikipedia.org/wiki/Leaky_bucket
[26]Whei-Jen Chen, Bill Comeau, H T Morgan. DB2 Workload Manager for Linux, UNIX, and Windows, 2008-7. 10-79
[27]Carlo Curino, Evan P.C.,Jones Raluca. Relational Cloud, A Database as a Service for the Cloud, 2005
[28]Carlo Curino,Evan Jones,Yang Zhang. Schism: a Workload Driven Approach to Database Replication and Partitioning, 2009
[29]Gaurav Somani,Sanjay Chaudhary. Application Performance Isolation in Virtualization. IEEE International Conference on Cloud Computing. 2009. 1-4
[30] Wenhong Tian, Sheng Su1, Guoming Lu. A Framework for Implementing and Managing Platform as a Service in a Virtual. Second International Workshop on Education Technology and Computer Science. 2010. 1-3