We minimize the monetary cost as well as providing
fast response times. We formulate the problem and
propose exhaustive and heuristic algorithms,
namely, multiobjective branch-and-bound (MOD-
B&B) and multiobjective robust genetic algorithm
(MOD-GA) for the optimization of the problem. To
the best of our knowledge, the multiobjective design
of Cloud data warehouses is being solved for the
first time with such an approach. There are studies
that concern with the best virtual resource
deployment or with the minimal monetary cost of
workloads in static hardware resources individually.
However we combine both of these optimization
techniques together and obtain significant results as
they are presented in our study. It is possible to
design and expand the study with additional elastic
virtual resources such as I/O bandwidth and dynamic
RAMs.
REFERENCES
Amazon Web Services (AWS). aws.amazon.com.
Barham, P., Dragovic, B., Fraser, K., Hand, S., Harris, T.,
Ho, A., and Warfield, A. (2003). Xen and the art of
virtualization. ACM SIGOPS Operating Systems
Review, 37(5), 164-177.
Balazinska, M., Howe, B., and Suciu, D. (2011). Data
markets in the cloud: An opportunity for the database
community. PVLDB, 4(12).
Bayir, M. A., Toroslu, I. H., and Cosar, A. (2007). Genetic
Algorithm for the Multiple- Query Optimization
Problem. IEEE Transactions on Systems, Man, and
Cybernetics- Part C: Applications and Reviews, Vol.
37 (1):147-153.
Curino, C., Jones, E., Popa, R., Malviya, N., Wu, E.,
Madden, S., Balakrishnan, H., and Zeldovich, N.
(2011). Relational Cloud: A Database Service for the
Cloud. CIDR, pp.235-240.
D’Orazio, L., Bimonte, S., and Darmont, J. (2012). Cost
Models for View Materialization in the Cloud. In
Proceedings of the Workshop on Data Analytics in the
Cloud (EDBT-ICDT/DanaC).
Dokeroglu, T. (supervised by Ahmet Cosar) (2012).
Parallel Genetic Algorithms for the Optimization of
Multi-Way Chain Join Queries of Distributed
Databases, 38th VLDB Ph.D. Workshop, August 27-
31, Istanbul/TURKEY.
Holland, J. H. (1975) Adaptation in Natural and Artificial
Systems. University of Michigan Press, Ann Arbor,
MI, USA.
Soror, A. A., Minhas, U. F., Aboulnaga, A., Salem, K.,
Kokosielis, P., and Kamath, S. (2010). Automatic
virtual machine configuration for database workloads.
ACM Transactions on Database Systems (TODS),
35(1), 7.
Tosun, U., Dokeroglu, T., and Cosar, A. (2013). A robust
Island Parallel Genetic Algorithm for the Quadratic
Assignment Problem. International Journal of
Production Research, 1-17.
Weikum, G. and Vossen, G. (2002). Transactional
Information Systems. Morgan Kaufmann.
Xiong, P., Chi, Y., Zhu, S., Moon, H. J., Pu, C., and
Hacigumus, H. (2011). Intelligent management of
virtualized resources for database systems in cloud
environment. In Data Engineering (ICDE), IEEE 27th
International Conference on (pp. 87-98).
APPENDIX
Alternative query execution plans for TPC-H query Q3
TPC-H Q3 statement in accordance with the query execution
plan 1 where all of the tables are shipped to query issuing
node.
SELECT TOP 10 L ORDERKEY ,..., O SHIPPRIORITY
FROM [VM2].CUSTOMER C, [VM3].ORDERS O,
[VM1].LINEITEM L,
WHERE C.C MKTSEGMENT = ’BUILDING’
AND C.C CUSTKEY = O.O CUSTKEY
AND L.L ORDERKEY = O.O ORDERKEY
AND O.O ORDERDATE < ’1995-03-15’
AND L.L SHIPDATE > ’1995-03-15’
GROUP BY L.L ORDERKEY, O.O ORDERDATE, O.O
SHIPPRIORITY
ORDER BY REVENUE DESC, O.O ORDERDATE;
TPC-H Q3 statement in accordance with query execution plan
2 where CUSTOMER and ORDERS tables are joined at
virtual machine 3 and the resulting tuples are shipped to
virtual machine 2 to join with LINEITEM table.
SELECT TOP 10 L ORDERKEY ,..., O SHIPPRIORITY
FROM OPENQUERY ( [VM3]. ’SELECT O ORDERDATE, O
SHIPPRIORITY, O ORDERKEY
FROM [VM2].CUSTOMER C, [VM3].ORDERS O
WHERE C.C MKTSEGMENT = ’BUILDING’
AND C.C CUSTKEY = O.O CUSTKEY
AND O.O ORDERDATE < ’1995-03-15’
GROUP BY O.O ORDERDATE, O.O SHIPPRIORITY, O
ORDERKEY ORDER BY O.OORDERDATE) Remote1,
[VM1].LINEITEM L WHERE AND L.L ORDERKEY =
Remote1.O ORDERKEY AND L.L SHIPDATE > ’1995-03-15’
GROUP BY L.L ORDERKEY, Remote1.O ORDERDATE,
Remote1.O SHIPPRIORITY
ORDER BY REVENEU DESC’);
ICAART2014-InternationalConferenceonAgentsandArtificialIntelligence
576