May 1999

End of Year Report
UPSCALE: Undergraduate Physics Students'
Computing and Learning Environment
Department of Physics
University of Toronto

Prepared by David Harrison

INTRODUCTION

UPSCALE provides computational services to undergraduate physics students in three main areas:

Delivery of information via the World Wide Web. This includes course home pages, problem set solutions, copies of overhead transparencies, laboratory instrument specifications, etc. As will be seen, almost all of this usage is via students accessing our web server remotely. Before we converted the information sub-system to web form, we were experiencing unacceptable pressure on the number of "seats" that we could provide our students in-house.
Data analysis, data testing, and evaluation. This is almost all via our in-house X-terminals.
Data acquisition, process control, computational physics. Used primarily by third and fourth year students, this is mostly via the Nortel Laboratory in Applied Physics, with UPSCALE's central computers acting as file and print servers for the 16 stations in that laboratory.

As is customary, this report summarises the undergraduate computing facility in the past academic year. We do not report on the active use of the Nortel laboratory by our upper-year students. In the final section, we discuss future directions for UPSCALE.

USAGE SUMMARY

Web Access

The following table summarises the delivery of information via the World Wide Web for the past three years. Virtually all of this usage is by our undergraduate students. The phrase "top level" accesses means, for example, when a student accesses the Solutions sub-system from its top-level page; a PHY138Y student might access the relevant part of the sub-system for their course from the PHY138Y home page or from a bookmark, and these accesses are not included.

I also present summaries for three courses which make heavy use of the web to deliver course materials; these courses are PHY100F, PHY138Y and JPU200Y. Almost all of the access to the JPU200Y home page occurred in the second term.

Also, in mid-February we established a Physics Virtual Bookshelf, which is a collection of course materials that we feel may be of interest to a wider audience than just the students in the course for which they were originally prepared; access to the Bookshelf is also summarised below.

What	1998-99	1997-98	1996-97
Files delivered	1,599,770	834,568	482,046
Percentage delivered in-house	3.0%	7.4%	36.9%
Solutions top-level accesses	22,467	18,919	19,395
Course/Lab home pages top-level accesses	42,228	27,070	12,283
PHY138Y home page top-level accesses	27,579 ~ 29/student	Not available	Not available
PHY100F home page top-level accesses	1949 ~ 19/student	Not available	Not available
JPU200Y home page top-level accesses	1806 ~ 26/student	Not available	Not available
Physics Virtual Bookshelf top-level accesses	659	Not applicable	Not applicable

Logins

The following table shows the usage figures for this year and last year via traditional logins. We have two categories of students accounts: the "x" accounts for first and second year students, and "special" accounts for upper year students.

What	1998-99	1997-98
Active "x" accounts	1940	1792
Active "special" accounts	116	117
Total student accounts	2056	1909
Number of "x" logins	21,499	22,576
Number of "special" logins	5687	12,341
Total logins	27,186	34,917
"x" connect hours	10,246	9436
"special" connect hours	1492	2725
Total connect hours	11,738	11,861

Virtually all of the above usage is via our in-house X-terminals. For example, only 579 logins last year were from students dialing in from their homes or elsewhere.

The figure shows the logins per week through the academic year. The Christmas break (weeks 14 - 17) and Reading Week (week 23) are clearly visible.
graph of logins per week

CONCLUSIONS AND FUTURE PLANS

The increase in usage of our web facility doubled again last year. The percentage of students accessing the facility from our own "seats" continues to decline.

As the I Year Laboratory continues to place increased emphasis on using UPSCALE's facilities for data analysis, the logins by these students remains fairly stable. Our upper year students are not logging in as much as previously; presumably they are instead using the web facility and the Nortel laboratory instead.

Our 7 year old web server was struggling to keep up with demand this year. In addition, our coax-based networks occasionally saturated. We have in place a new much faster web server which will be in production by September. In addition, our networks are currently being upgraded to 100 Mbps.