Unit:
Κατεύθυνση / ειδίκευση Διαχείριση Πληροφορίας και Δεδομένων (ΔΕΔ)Library of the School of Science
Author:
Φούφουλας Ιωάννης
Supervisors info:
Ιωάννης Ιωαννίδης Καθηγητής (Επιβλέπων), Αλέξης Δελής Καθηγητής
Original Title:
Ελαστικές μέθοδοι συμπίεσης για ερωτήματα αναλυτικής επεξεργασίας
Translated title:
Elastic compression techniques for analytical query workloads
Summary:
The goal of this thesis is the implementation of compression algorithms for
distributed systems that are used for online analytical processing (OLAP) . We
process relational data and we pay attention to 2 critical issues :
Data Layout. The data layout is very important since it can affect
query execution times. OLAP systems use mainly simple scans so we need
appropriate layouts for such queries.
Data movement. In these systems large amounts of data are transferred
over the network so a high compression will reduce significantly the data
movement cost.
During this thesis we examine different data layouts and compression algorithms
that are used by current systems and we propose some new algorithms and
approaches. We produce parameterized algorithms which offer tradeoffs between
compression level and execution times, so that the system on top may select
options according to the network speed and the load at every time.
Keywords:
Compression, Data layout, Online analytical processing (OLAP)), Data transfer, Serialization
Number of index pages:
5,6,7,8
