Special section on data-intensive cloud infrastructure

Special section on data-intensive cloud infrastructure The VLDB Journal (2014) 23:843 DOI 10.1007/s00778-014-0371-0 GUEST EDITORIAL Ashraf Aboulnaga · Beng Chin Ooi · Patrick Valduriez Published online: 30 September 2014 © Springer-Verlag Berlin Heidelberg 2014 More and more individuals, companies and organizations are is modeled as a hypergraph, which allows drawing connec- relying on the cloud to store and manage their data, which tions to graph theoretic concepts. Using query span, i.e., the translates into increasing pressure on the cloud infrastruc- average number of machines involved in the execution of a ture. Cloud data can be very diverse, including a wide variety query or a transaction, as the metric to optimize, the authors of personal data collections, very large multimedia content develop data placement and replication algorithms as well repositories and very large datasets. Users and application as scalable techniques to reduce the overhead of partitioning developers can be in very high numbers, with little DBMS and query routing. To deal with workload changes, they also expertise. Data-intensive applications can be very diverse propose an incremental repartitioning technique. The exper- too, with requirements ranging from basic database capa- iment shows significant reduction in total resource consump- bilities to complex analytics over big data. In particular, the tion for OLAP workloads, and improved transaction latency pay-as-you-go model makes the cloud attractive for support- and overall throughput for OLTP workloads. ing novel large-scale elastic applications. The second paper deals with applications such as bioinfor- NoSQL solutions for the cloud, for instance, have traded matics, time series and web log analysis, which require the consistency and transactional guarantees for scalability. extraction of frequent patterns, called motifs, from one very However, the grand challenge for a data-intensive cloud long (i.e., several gigabytes) sequence. It presents ACME, a infrastructure is to provide ease of use, consistency, privacy, parallel cloud-oriented system for extracting such motifs. scalability and elasticity, simultaneously, over cloud data. ACME uses a combinatorial approach that scales to giga- Addressing this challenge requires novel solutions across the byte long sequences, and is the first to support supermaximal spectrum of data management techniques, including massive motifs. ACME can be deployed on thousands of CPUs in data storage, elastic parallel query processing, transactions the cloud and includes an automatic tuning mechanism that over data replicated at geographically distributed sites, secu- suggests the appropriate number of CPUs to utilize, in order rity and privacy, and efficient data loading and access. This to meet the user runtime constraints while minimizing cloud special section focuses on recent advances in research and resources usage. The experiments show that, compared to development in data-intensive cloud infrastructures. the state of the art, ACME supports 3 orders of magnitude The first paper proposes a workload-aware data place- longer sequences, scales out very well and supports elastic ment and replication approach for minimizing resource con- deployment in the cloud. sumption in cloud data management systems. The workload A. Aboulnaga (B) · B. C. Ooi · P. Valduriez Waterloo, Canada e-mail: ashraf@uwaterloo.ca http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The VLDB Journal Springer Journals

Special section on data-intensive cloud infrastructure

Free
1 page
Loading next page...
1 Page
 
/lp/springer_journal/special-section-on-data-intensive-cloud-infrastructure-4BeARy07Gn
Publisher
Springer Berlin Heidelberg
Copyright
Copyright © 2014 by Springer-Verlag Berlin Heidelberg
Subject
Computer Science; Database Management
ISSN
1066-8888
eISSN
0949-877X
D.O.I.
10.1007/s00778-014-0371-0
Publisher site
See Article on Publisher Site

Abstract

The VLDB Journal (2014) 23:843 DOI 10.1007/s00778-014-0371-0 GUEST EDITORIAL Ashraf Aboulnaga · Beng Chin Ooi · Patrick Valduriez Published online: 30 September 2014 © Springer-Verlag Berlin Heidelberg 2014 More and more individuals, companies and organizations are is modeled as a hypergraph, which allows drawing connec- relying on the cloud to store and manage their data, which tions to graph theoretic concepts. Using query span, i.e., the translates into increasing pressure on the cloud infrastruc- average number of machines involved in the execution of a ture. Cloud data can be very diverse, including a wide variety query or a transaction, as the metric to optimize, the authors of personal data collections, very large multimedia content develop data placement and replication algorithms as well repositories and very large datasets. Users and application as scalable techniques to reduce the overhead of partitioning developers can be in very high numbers, with little DBMS and query routing. To deal with workload changes, they also expertise. Data-intensive applications can be very diverse propose an incremental repartitioning technique. The exper- too, with requirements ranging from basic database capa- iment shows significant reduction in total resource consump- bilities to complex analytics over big data. In particular, the tion for OLAP workloads, and improved transaction latency pay-as-you-go model makes the cloud attractive for support- and overall throughput for OLTP workloads. ing novel large-scale elastic applications. The second paper deals with applications such as bioinfor- NoSQL solutions for the cloud, for instance, have traded matics, time series and web log analysis, which require the consistency and transactional guarantees for scalability. extraction of frequent patterns, called motifs, from one very However, the grand challenge for a data-intensive cloud long (i.e., several gigabytes) sequence. It presents ACME, a infrastructure is to provide ease of use, consistency, privacy, parallel cloud-oriented system for extracting such motifs. scalability and elasticity, simultaneously, over cloud data. ACME uses a combinatorial approach that scales to giga- Addressing this challenge requires novel solutions across the byte long sequences, and is the first to support supermaximal spectrum of data management techniques, including massive motifs. ACME can be deployed on thousands of CPUs in data storage, elastic parallel query processing, transactions the cloud and includes an automatic tuning mechanism that over data replicated at geographically distributed sites, secu- suggests the appropriate number of CPUs to utilize, in order rity and privacy, and efficient data loading and access. This to meet the user runtime constraints while minimizing cloud special section focuses on recent advances in research and resources usage. The experiments show that, compared to development in data-intensive cloud infrastructures. the state of the art, ACME supports 3 orders of magnitude The first paper proposes a workload-aware data place- longer sequences, scales out very well and supports elastic ment and replication approach for minimizing resource con- deployment in the cloud. sumption in cloud data management systems. The workload A. Aboulnaga (B) · B. C. Ooi · P. Valduriez Waterloo, Canada e-mail: ashraf@uwaterloo.ca

Journal

The VLDB JournalSpringer Journals

Published: Dec 1, 2014

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve Freelancer

DeepDyve Pro

Price
FREE
$49/month

$360/year
Save searches from
Google Scholar,
PubMed
Create lists to
organize your research
Export lists, citations
Read DeepDyve articles
Abstract access only
Unlimited access to over
18 million full-text articles
Print
20 pages/month
PDF Discount
20% off