CloudOptimizer: Multi-tenancy for I/O-bound OLAP workloads

Hatem A. Mahmoud, Hyun Jin Moon, Yun Chi, Hakan Hacigümüş, Divyakant Agrawal, Amr El-Abbadi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

10 Citations (Scopus)

Abstract

Consolidation of multiple databases on the same server allows service providers to save significant resources because many production database servers are often under-utilized. Recent research investigates the problem of minimizing the number of servers required to host a set of tenants when the working sets of tenants are kept in main memory (e.g., in-memory OLAP workloads, or OLTP workloads), thus the memory assigned to each tenant, as well as the I/O bandwidth and CPU time, are all dictated by the working set size of the tenant. Other research investigates the reverse problem when the number of servers is fixed, but the amount of resources allocated to different tenants on the same server needs to be configured to optimize a cost function. In this paper we investigate the problem when neither the number of servers nor the amount of resources allocated to each tenant are fixed. This problem arises when consolidating OLAP workloads of tenants whose service-level agreements (SLAs) allow for queries to be answered from disk. We study the trade-off between the amount of memory and the I/O bandwidth assigned to OLAP workloads, and develop a principled approach for allocating resources to tenants in a manner that minimizes the total number of servers required to host all tenants while satisfying the SLA of each tenant. We then explain how we modified InnoDB, the storage engine of MySQL, to be able to change the amount of resources allocated to each tenant at runtime, so as to account for fluctuations in workloads. Finally, we evaluate our approach experimentally using the TPC-H benchmark to demonstrate its effectiveness and accuracy.

Original languageEnglish
Title of host publicationACM International Conference Proceeding Series
Pages77-88
Number of pages12
DOIs
Publication statusPublished - 2 May 2013
Externally publishedYes
Event16th International Conference on Extending Database Technology, EDBT 2013 - Genoa, Italy
Duration: 18 Mar 201322 Mar 2013

Other

Other16th International Conference on Extending Database Technology, EDBT 2013
CountryItaly
CityGenoa
Period18/3/1322/3/13

Fingerprint

Servers
Data storage equipment
Bandwidth
Cost functions
Consolidation
Program processors
Engines

Keywords

  • Algorithms
  • C.4 [Performance of Systems]: Modeling techniques
  • Experimentation
  • H.2.4 [Database Management]: Systems
  • Measurement

ASJC Scopus subject areas

  • Human-Computer Interaction
  • Computer Networks and Communications
  • Computer Vision and Pattern Recognition
  • Software

Cite this

Mahmoud, H. A., Moon, H. J., Chi, Y., Hacigümüş, H., Agrawal, D., & El-Abbadi, A. (2013). CloudOptimizer: Multi-tenancy for I/O-bound OLAP workloads. In ACM International Conference Proceeding Series (pp. 77-88) https://doi.org/10.1145/2452376.2452386

CloudOptimizer : Multi-tenancy for I/O-bound OLAP workloads. / Mahmoud, Hatem A.; Moon, Hyun Jin; Chi, Yun; Hacigümüş, Hakan; Agrawal, Divyakant; El-Abbadi, Amr.

ACM International Conference Proceeding Series. 2013. p. 77-88.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Mahmoud, HA, Moon, HJ, Chi, Y, Hacigümüş, H, Agrawal, D & El-Abbadi, A 2013, CloudOptimizer: Multi-tenancy for I/O-bound OLAP workloads. in ACM International Conference Proceeding Series. pp. 77-88, 16th International Conference on Extending Database Technology, EDBT 2013, Genoa, Italy, 18/3/13. https://doi.org/10.1145/2452376.2452386
Mahmoud HA, Moon HJ, Chi Y, Hacigümüş H, Agrawal D, El-Abbadi A. CloudOptimizer: Multi-tenancy for I/O-bound OLAP workloads. In ACM International Conference Proceeding Series. 2013. p. 77-88 https://doi.org/10.1145/2452376.2452386
Mahmoud, Hatem A. ; Moon, Hyun Jin ; Chi, Yun ; Hacigümüş, Hakan ; Agrawal, Divyakant ; El-Abbadi, Amr. / CloudOptimizer : Multi-tenancy for I/O-bound OLAP workloads. ACM International Conference Proceeding Series. 2013. pp. 77-88
@inproceedings{a3f5d85fcf2941cab32255371599dd40,
title = "CloudOptimizer: Multi-tenancy for I/O-bound OLAP workloads",
abstract = "Consolidation of multiple databases on the same server allows service providers to save significant resources because many production database servers are often under-utilized. Recent research investigates the problem of minimizing the number of servers required to host a set of tenants when the working sets of tenants are kept in main memory (e.g., in-memory OLAP workloads, or OLTP workloads), thus the memory assigned to each tenant, as well as the I/O bandwidth and CPU time, are all dictated by the working set size of the tenant. Other research investigates the reverse problem when the number of servers is fixed, but the amount of resources allocated to different tenants on the same server needs to be configured to optimize a cost function. In this paper we investigate the problem when neither the number of servers nor the amount of resources allocated to each tenant are fixed. This problem arises when consolidating OLAP workloads of tenants whose service-level agreements (SLAs) allow for queries to be answered from disk. We study the trade-off between the amount of memory and the I/O bandwidth assigned to OLAP workloads, and develop a principled approach for allocating resources to tenants in a manner that minimizes the total number of servers required to host all tenants while satisfying the SLA of each tenant. We then explain how we modified InnoDB, the storage engine of MySQL, to be able to change the amount of resources allocated to each tenant at runtime, so as to account for fluctuations in workloads. Finally, we evaluate our approach experimentally using the TPC-H benchmark to demonstrate its effectiveness and accuracy.",
keywords = "Algorithms, C.4 [Performance of Systems]: Modeling techniques, Experimentation, H.2.4 [Database Management]: Systems, Measurement",
author = "Mahmoud, {Hatem A.} and Moon, {Hyun Jin} and Yun Chi and Hakan Hacig{\"u}m{\"u}ş and Divyakant Agrawal and Amr El-Abbadi",
year = "2013",
month = "5",
day = "2",
doi = "10.1145/2452376.2452386",
language = "English",
isbn = "9781450315975",
pages = "77--88",
booktitle = "ACM International Conference Proceeding Series",

}

TY - GEN

T1 - CloudOptimizer

T2 - Multi-tenancy for I/O-bound OLAP workloads

AU - Mahmoud, Hatem A.

AU - Moon, Hyun Jin

AU - Chi, Yun

AU - Hacigümüş, Hakan

AU - Agrawal, Divyakant

AU - El-Abbadi, Amr

PY - 2013/5/2

Y1 - 2013/5/2

N2 - Consolidation of multiple databases on the same server allows service providers to save significant resources because many production database servers are often under-utilized. Recent research investigates the problem of minimizing the number of servers required to host a set of tenants when the working sets of tenants are kept in main memory (e.g., in-memory OLAP workloads, or OLTP workloads), thus the memory assigned to each tenant, as well as the I/O bandwidth and CPU time, are all dictated by the working set size of the tenant. Other research investigates the reverse problem when the number of servers is fixed, but the amount of resources allocated to different tenants on the same server needs to be configured to optimize a cost function. In this paper we investigate the problem when neither the number of servers nor the amount of resources allocated to each tenant are fixed. This problem arises when consolidating OLAP workloads of tenants whose service-level agreements (SLAs) allow for queries to be answered from disk. We study the trade-off between the amount of memory and the I/O bandwidth assigned to OLAP workloads, and develop a principled approach for allocating resources to tenants in a manner that minimizes the total number of servers required to host all tenants while satisfying the SLA of each tenant. We then explain how we modified InnoDB, the storage engine of MySQL, to be able to change the amount of resources allocated to each tenant at runtime, so as to account for fluctuations in workloads. Finally, we evaluate our approach experimentally using the TPC-H benchmark to demonstrate its effectiveness and accuracy.

AB - Consolidation of multiple databases on the same server allows service providers to save significant resources because many production database servers are often under-utilized. Recent research investigates the problem of minimizing the number of servers required to host a set of tenants when the working sets of tenants are kept in main memory (e.g., in-memory OLAP workloads, or OLTP workloads), thus the memory assigned to each tenant, as well as the I/O bandwidth and CPU time, are all dictated by the working set size of the tenant. Other research investigates the reverse problem when the number of servers is fixed, but the amount of resources allocated to different tenants on the same server needs to be configured to optimize a cost function. In this paper we investigate the problem when neither the number of servers nor the amount of resources allocated to each tenant are fixed. This problem arises when consolidating OLAP workloads of tenants whose service-level agreements (SLAs) allow for queries to be answered from disk. We study the trade-off between the amount of memory and the I/O bandwidth assigned to OLAP workloads, and develop a principled approach for allocating resources to tenants in a manner that minimizes the total number of servers required to host all tenants while satisfying the SLA of each tenant. We then explain how we modified InnoDB, the storage engine of MySQL, to be able to change the amount of resources allocated to each tenant at runtime, so as to account for fluctuations in workloads. Finally, we evaluate our approach experimentally using the TPC-H benchmark to demonstrate its effectiveness and accuracy.

KW - Algorithms

KW - C.4 [Performance of Systems]: Modeling techniques

KW - Experimentation

KW - H.2.4 [Database Management]: Systems

KW - Measurement

UR - http://www.scopus.com/inward/record.url?scp=84876800162&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84876800162&partnerID=8YFLogxK

U2 - 10.1145/2452376.2452386

DO - 10.1145/2452376.2452386

M3 - Conference contribution

AN - SCOPUS:84876800162

SN - 9781450315975

SP - 77

EP - 88

BT - ACM International Conference Proceeding Series

ER -