web analytics

Data, data everywhere 

Twitter LinkedIn RSS

CFP: Fifth Workshop on Big Data Benchmarking (WBDB 2014)

The Fifth International Workshop on Big Data Benchmarking will be held in Potsdam, Germany at the Hasso Platner Institut from August 5-6, 2014. The WBDB workshops are designed to make progress towards the development of industry-standard benchmarks for evaluating hardware and software solutions for big data applications.

Topics to be discussed at the Workshop include, but are not limited to:

  • Data features: New feature sets of data including, high-dimensional data, sparse data, event-based data, and enormous data sizes.
  • System characteristics: System-level issues including, large-scale and evolving system configurations, shifting loads, and heterogeneous technologies for big data and cloud platforms.
  • Implementation options: Different implementation options such as SQL, NoSQL, Hadoop software ecosystem, and different implementations of HDFS.
  • Workload: Representative big data business problems and corresponding benchmark implementations. Specification of benchmark applications that represent the different modalities of big data, including graphs, streams, scientific data, and document collections.
  • Hardware options: Evaluation of new options in hardware including different types of HDD, SSD, and main memory, and large-memory systems, and new platform options that include dedicated commodity clusters and cloud platforms.
  • Synthetic data generation: Models and procedures for generating large-scale synthetic data with requisite properties.
  • Benchmark execution rules: E.g. data scale factors, benchmark versioning to account for rapidly evolving workloads and system configurations, benchmark metrics.
  • Metrics for efficiency: Measuring the efficiency of the solution, e.g. based on costs of acquisition, ownership, energy and/or other factors, while encouraging innovation and avoiding benchmark escalations that favor large inefficient configuration over small efficient configurations.
  • Evaluation frameworks: Tool chains, suites and frameworks for evaluating big data systems.
  • Early implementations: Of the Deep Analytics Pipeline or BigBenchand lessons learned in benchmarking big data applications.
  • Enhancements: Proposals to augment these benchmarks, e.g. by  adding more data genres (e.g. graphs), or incorporating a range of machine learning and other algorithms, will be entertained and are encouraged.

Organizing committee

  • Chaitan Baru, San Diego Supercomputer Center (SDSC) UC San Diego
  • Tilmann Rabl, Middleware Services Research Group (MSRG), University of Toronto
  • Kai Sachs, SAP AG

Local Arrangements

  • Matthias Uflacker, Hasso Plattner Institute

Publicity Chair

  • Henning Schmitz, SAP Innovations Center

Publication Chair

  • Meikel Poess, Oracle

Short versions of papers (4-8 pages) should be submitted by May 4, 2014, using the Easychair conference management system. Notification of acceptance is scheduled for June 8. Camera-ready copies of each paper, in Springer LCNS format, are expected to be received on July 20, prior to the Workshop.

 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
No Comments  comments 

Women Who Code

A new organization for women in computing is starting up in Kitchener-Waterloo. Women Who Code is dedicated to educating and inspiring women to pursue and excel in technology careers.

The Waterloo chapter of Women Who Code will have their launch event next Tuesday, March 18th 2014, starting at 7:30pm, in Area 151 of the Communitech Hub. One of the goals of Women Who Code is to establish a network of women who are interested in careers in technology and who are interested in helping each other succeed in their studies and their careers.

Female students in Conestoga’s various IT and Engineering programs are encouraged to attend the K-W Women Who Code launch event, and meet and network with like-minded women in computing.

For more information and to register for the event, visit the Kitchener-Waterloo WWC Web site: http://www.meetup.com/Women-Who-Code-Waterloo/. You can follow Women Who Code on Twitter by following @WomenWhoCode.

Thanks to my friend Joanne Atlee, a Professor of Computer Science at the University of Waterloo, for sending this my way.

 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
No Comments  comments 

Set-level operations do matter

A recent email exchange with software consultant Breck Carter regarding the performance of user-defined functions in SQL has prompted me to re-post this article that I originally wrote in 2008. User-defined functions in SQL have their uses; unfortunately their performance characteristics are highly variable because it is very difficult impossible for a query optimizer to take their execution characteristics into account when deciding on the best access plan for a given query. For example, a change to the join order lower down in an execution plan may completely alter the ordering of rows in an intermediate result. If a user-defined function sits above the join, then the execution characteristics of the UDF may be negatively impacted, particularly if the ordering of the rows negates any benefit of memoization – the practice of caching prior execution results of the UDF for subsequent re-use, rather than naively re-executing the function each time. The re-execution of a user-defined function in SQL is prohibitively expensive due to the creation and tear-down of the UDF’s execution context (essentially the same as when executing a stored procedure), which destroys memory and CPU cache locality when executing the query plan.

Anticipating the performance of such complex queries in a production environment can be exceedingly difficult because the performance of the UDF may be affected not only by the size of the underlying tables but also on the amount of correlation between the values referenced within the function, which impacts the benefit of memoization. Testing such queries over small test databases will typically not alert the developer to the possibility of performance issues in a production setting.

The remainder of this blog entry is copyright by Sybase, Inc., an SAP Company, and first appeared on Glenn Paulley’s Sybase blog (http://iablog.sybase.com/paulley/) on 16 May 2008. It is reprinted here with permission.

Set-level operations do matter

One of the problems that the ad-hoc relational mappings offered by the Hibernate object persistence library causes is the generation of queries that contain subqueries – sometimes in a plentiful number. Subqueries can be very difficult to optimize because it remains a challenging research problem to estimate both their cost and their selectivity. SQL Anywhere contains quite sophisticated query rewrite optimizations for subqueries. However, for most commercial products, including SQL Anywhere, the optimization and execution of subqueries remains problematic.

This week I’ve been doing some problem determination for a customer application, and what I want to convey in this post is that the impact of subquery execution can be pronounced, to the extent that it may affect the scalability of the application.

Read more…

 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
1 Comment  comments 

IEEE Technology Management Council to become IEEE Society

In a recent announcement, Jennifer Trelewicz, President of the IEEE Technology Management Council, stated that the IEEE Technical Activities Board has approved the transition of the IEEE Technology Management Council to an IEEE Society. Once ratified by the IEEE Board of Directors, the transition of the TMC to an IEEE Society will take place on 1 January 2015.

The IEEE TMC is that part of the IEEE devoted to the management of engineering projects and functions, and are the publishers of two journals devoted to engineering management: the IEEE Transactions on Engineering Management journal, and the IEEE Engineering Management Review.

The TMC also sponsors the annual IEEE Technology Management Conference, which this year will be held in Chicago from June 12-15. The ITMC Conference is where one will hear about the latest work regarding Agile software development methodologies, Innovation and Technology Management in public organizations, Six Sigma, and global (geographically dispersed) software project development.

 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
No Comments  comments 

2014 Conestoga web site development competition

The 2014 edition of the annual Conestoga Web Site Development Competition will be held the evening of Tuesday, February 25 2014 from 6:30pm-8:30pm in room 2A206 at the Doon campus.

The website development competition provides students with the opportunity to demonstrate, through practical application, their skills in website design and development. Students are asked to design and develop an elementary web application using graphics software, (X)HTML, CSS, JavaScript, PHP, and MySQL. The top two competitors of this competition qualify to compete at the provincial Skills Ontario competition and will be provided with training materials from the professional textbook publisher Mike Murach & Associates.

For this College competition, students will not be completing every part of the required solution from scratch. Rather, students will be given parts of the application at varying levels of completion, and they will be expected to have a working solution at the end of the two hours. This generally entails a simple application that demonstrates the use of persistent data, presented through accessible, usable, and valid web pages. Also, students will not be able to bring anything with them to the competition (for example, source code), but will have access to documentation from the World Wide Web Consortium, Mozilla Developer Network, The PHP Group, and Oracle.

Contact Dalibor Dvorski (ddvorski@conestogac.on.ca) for more information and to register.

 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
No Comments  comments 

Call for papers: Second International Workshop on Testing the Cloud

Following a successful first workshop, the Second International Workshop on Testing the Cloud will be held in Cleveland, Ohio on 31 March 2014, co-located with the 2014 ICST Conference (International Conference on Software Testing, Verification and Validation).

Important Dates

  • Paper submission Deadline: Jan 26th 2014
  • Author Notification of Acceptance: Feb 9th 2014
  • Camera-Ready Copy Due: Feb 23rd 2014
  • Workshop Date: March 31st 2014

Call for papers

Cloud computing is everywhere, inevitable: originally a layered abstraction of a heterogeneous environment, it has become the paradigm of a large-scale data-oriented system. And while it has some interesting features (easy deployment of applications, resiliency, security, performance, scalability, elasticity, etc.), testing its robustness and its reliability is a major challenge. The Cloud is an intricate collection of interconnected and virtualised computers, connected services, complex service-level agreements. From a testing perspective, the Cloud is then a complex composition of complex systems, and one can wonder whether anything like a global testing is possible? But if the answer is no, what can we conclude from partial tests? The question of testing this large, network-based, dynamic, composition of computers, virtual machines, servers, services, SLAs, seems particularly difficult. And critical for Cloud vendors: customers’ trust is indeed critical for companies implementing Clouds, and they have to ensure that the system has all the security and performance characteristics the marketing department highlights. This problem is a perfect example of cross concerns between academia and product companies, and it covers a broad range of topics, from software development to code analysis, performance monitoring to formal model for system testing, and so on.

In TTC, we aim at bringing together researchers and practitioners interested in this difficult question of testing the cloud, i.e., a complex distributed, dynamic and interconnected system. Hence, we call for regular scientific submissions, but also for industrial experience feedback. We are interested in contributions related to ‘testing the Cloud’ (i.e., testing the Cloud itself, for instance, its infrastructure), ‘testing in the Cloud’ (i.e., testing applications that are deployed in the Cloud), and ‘testing with the Cloud’ (e.g., using the Cloud capabilities to perform stress testing on an application). All the submissions describing approaches used in the industry, defining new methods to facilitate testing or identifying new challenges are relevant.

Topics of Interest

“Testing the Cloud” covers many different topics. We welcome academic and industrial contributions that are relevant. We will run regular academic sessions, but we are also likely to have a more industry-focused session where it will be possible to describe solutions deployed in product companies or best practices followed by practitioners. Topics include, but are not limited to, the development and application of testing cloud-based systems, e.g.

  • Domain-specific languages for testing
  • Fault injection
  • Formal specification and verification of programming libraries and programs
  • Functional and structural testing
  • New tools for testing
  • Performance testing
  • Programming techniques and methodologies (that decrease the need for testing)
  • Replay techniques for multi-threaded applications
  • Static and dynamic program analysis (including code review)
  • Test generation algorithms and tools
  • Security testing of concurrent systems and applications in the cloud
  • Load Testing
  • Live-testing
  • Test environment/Production environment
  • Test Monitoring
  • Test and customer relationship management
Submission and Proceedings

Regular technical papers must be prepared in ACM conference format and must not exceed 6 pages. All submissions must be in English. Submissions that do not adhere to these guidelines or that violate formatting will be declined without review.

Submit your paper via the paper submission website. Papers are due on January 26th 2014 at 11:59:59 PM, Western European Time (WET). We also welcome abstracts of presentations focusing on real solutions/products and best practices. Those can be short (max. 4 pages) and must show what the relevant things the audience could take away from the talk: how a software can address testing the cloud, what process is used in a company to make sure testing the cloud is optimized, etc.

Program Committee

Organizers and General Chairs: Jeff Offutt, George Mason University; Christina Thorpe, University College, Dublin, Ireland; Anthony Ventresque, University College, Dublin, Ireland.

Program Committee:

  • Bram Adams, École Polytechnique de Montréal, Canada
  • Carsten Binnig, DHBW Mannheim, Germany
  • Ivona Brandic, Vienna UT, Austria
  • Francisco Brasileiro, UFCG, Brasil
  • Achim Brucker, SAP AG, Karlsruhe, Germany
  • Rodrigo Calheiros, Uni. of Melbourne, Australia
  • Thomas Cerqueus, UCD, Ireland
  • Christoph Csallner, Uni. of Texas at Arlington, USA
  • Erik Elmroth, Umeå University, Sweden
  • Vahid Garousi, University of Calgary, Canada
  • Ahmed E. Hassan, Queen’s University, Canada
  • Jack ZhenMing Jiang, York UniversitY, Canada
  • Gregory Kapfhammer, Allegheny College, USA
  • Diwakar Krishnamurthy, Uni. of Calgary, Canada
  • Marin Litoiu, York University, Canada
  • Tejeddine Mouelhi, Uni. of Luxembourg, Luxembourg
  • Henry Muccini, University of L’Aquila, Italy
  • John Murphy, University College Dublin, Ireland
  • Aisling O’ Driscoll, CIT Cork, Ireland
  • Glenn Paulley, Conestoga College, Canada
  • Paul Strooper, University of Queensland, Australia
  • Stefanie Scherzinger, Regensburg Uni. of Applied Sciences, Germany
  • Gerson Sunyé, Uni. of Nantes, France
  • Toyotaro Suzumura, IBM Dublin and UCD, Ireland
  • Luis M Vaquero, HP Labs, Bristol, UK
  • Miao Wang, UCD Dublin, Ireland
 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
No Comments  comments 

Women in computing: forthcoming events

Two forthcoming events to support women in computing will be held in Waterloo Region this fall.

The first event is the inaugural meeting of the Waterloo Region Women in Trades Association, which has been spearheaded Judi Westcott and Laura Potje of Conestoga’s School of Engineering and IT. The purpose of the Women in Trades Association is to create a network of women, and supporters of women, in the construction and industrial trades with the intent of networking, educating and promoting women in the construction and industrial trades.

The inaugural meeting of the Women in Trades Association will take place the evening of Thursday, 26 September 2013:

What: Waterloo Region Women in Trades Association meeting
Who: Women, and supporters of women, in industrial, motive power or construction trades
Date: Thursday, September 26 2013
Time: 7:00 – 9:00 pm
Place: Conestoga College, Cambridge Campus (new building south of the 401) room A1210.
Cost: $5 non-students/$2 students. Light refreshments will be served; any profit will be donated to the Waterloo Region Food Bank
Facebook page: Waterloo Region Women in Trades Association
Email: wrwita@gmail.com if you wish to be on our list for future e-mails

RSVPs are appreciated but not necessary; everyone is welcome.

The second event is the fourth annual Ontario Conference for Women in Computing (ONCWIC) which is being held at the University of Waterloo on November 8/9, 2013. Registration for this event is now open, and for a very reasonable student fee of $40 the conference will include meals, social events, and three keynote talks.

Female students at Conestoga should take advantage of these opportunities to network with other women in computing and other engineering disciplines, which can lead to a support network and hiring once your studies have been completed.

Thanks to IT program co-ordinator Liz Stacey for sending this my way.

 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
No Comments  comments 

Pitfalls of using Windows Task Manager

In the Software QA Techniques course I am teaching this semester I am including various aspects of performance measurement in the syllabus, since performance is one of the top three factors of database applications (the other two factors are performance, and its related factor, performance). No, those are not typos.

Over my career at Sybase I often responded to customers who had concerns about memory usage of SQL Anywhere on Windows machines. Their concerns were sometimes based on misinterpretations of TaskMan output (the Windows Task Manager), and those misinterpretations prompted a section in the Capacity Planning with SQL Anywhere whitepaper, which I co-authored with Ivan Bowman. I have reproduced that article here; once again my thanks to John Smirnios of SAP Waterloo who provided the details for this article.

The remainder of this blog entry is copyright by Sybase, Inc., an SAP Company, and first appeared on Glenn Paulley’s Sybase blog (http://iablog.sybase.com/paulley/) on 23 October 2008. It is reprinted here with permission.

Pitfalls of using Windows Task Manager

A common problem on Windows is identifying how much physical memory is in use by a SQL Anywhere database server. The Windows Task Manager, which lists all of the active processes on the system, does not actually report true memory usage in the “Memory Size’ column, but instead reports each task’s working set size. Roughly speaking, on Windows XP and other similar Windows operating systems, working set size is the amount of physical RAM resident in memory and recently in use by the process.

The issue here is how Windows defines “recently in use’. Windows periodically “trims’ the working set of a process by marking most or all of the memory allocated to it as “not present” – as if it has been swapped out. However, Windows doesn’t actually swap that memory out, although those memory pages will be among the first to be re-used for other purposes. If the process references that memory again, the process incurs a virtual memory fault. Windows catches this “soft fault” and, knowing that the process is actually using that piece of memory, increments the process’s working set accordingly. Typically, Task Manager will report a drastically reduced working set size for a process when that process is minimized: Windows is making an assumption – invalid for a database server – that a minimized application will utilize very little memory, and severely trims the working set size for that process. As a result, a process’s working set size as displayed by Task Manager can fluctuate significantly, and neither it nor the “VM Size” value reported by Task Manager are useful metrics to use in determining memory consumption of the server.

Read more…

 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
No Comments  comments 

ONCWIC Registration is open!

The fourth annual Ontario Conference for Women in Computing (ONCWIC) is being held at the University of Waterloo from November 8-9. The list of featured guest speakers is now available and attendees can look forward to three outstanding seminars. The guest speakers are:

  • Maria Klawe, President of Harvey Mudd College, formerly Dean of Engineering and Professor of Computer Science at Princeton University;
  • Carol Leaman, CEO of Axonify (www.axonify.com); and
  • Kelley Irwin, Vice President, Technology Solutions at TD Bank Group.

Registration for university or college students is only $40.00. For faculty and other attendees the cost is $75.00. Register now to avoid disappointment.

Thanks to my friend Kelly Lyons of the University of Toronto for sending this my way.

 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
1 Comment  comments 

What has driven women out of Computer Science?

Over the years I have tried to bring some attention to the disproportionate number of women that are enrolled in Computer Science and Computer Engineering programs, and now that I am an instructor at Conestoga College I get to see this first-hand. One anecdotal example from yesterday’s Computer Application Development (CAD) diploma program orientation: of the approximately 30 students in the room, only three were women, a significant drop from last year’s CAD cohort.

I want to re-post an article that I initially wrote three years ago (with links corrected if possible), and afterwards add some additional material taken from last year’s Taulbee survey, which is done by the Computing Research Association.

The remainder of this blog entry is copyright by Sybase, Inc., an SAP Company, and first appeared on Glenn Paulley’s Sybase blog (http://iablog.sybase.com/paulley/) on 22 May 2010. It is reprinted here with permission.

What has driven women out of Computer Science?

Around the office, my recent articles regarding the forthcoming ACM-W event at Queen’s University, and the sponsorship of the University of Waterloo’s Women in Computer Science organization by Sybase Waterloo, has sparked several comments – mostly from those who, having graduated from university some years ago, had not realized that the situation had become so dire.

To offer some quantitative data: this study by Rhian Davies, Mark Hancock, and Anne Condon of the University of British Columbia outlines Canadian CS enrollment data as of 1998 thru 2000, with Bachelor’s programs having approximately 20% female participation – which is down considerably from the highs of 40% in the mid-1980’s. Unfortunately, while overall enrollment in Computer Science programs is up – albeit barely – over the past five years, the downward trend in Computer Science enrollment for women has continued and the numbers for 2008, taken from the Taulbee survey commissioned by the Computing Research Association, indicates that women constitute only 11.8% of the students in Computer Science Bachelors’ programs in North American universities, and merely 10.7% in Computer Engineering. Yet, overall, women outnumber men at Canadian universities, and in general other technical programs do not suffer from this degree of gender skew.

Read more…

 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
No Comments  comments