web cache and remote access

[Bryan H. Davidson]

Greetings -

Here at the University of Arkansas, Fayetteville, we have recently
encountered a serious problem with remote access to our online databases
that authenticate by our campus IP address. I would be very interested in
hearing whether other institutions or libraries have encountered a similar
problem.

The Problem:

All of our Internet traffic, as well as most of the libraries' Internet
traffic in the state, travels through a server in Little Rock before going
out into the world (and that is another story).  This in itself has not been
a major problem until now.

This weekend, they implemented something to the effect of  a "web caching
server" without any notification. As a result we immediately lost access to
all of our subscription online resources that authenticate by IP, because
the server altered the IP number. Once we figured out that this was the
cause of the problem, they had to add the URL of each database to an
"exception list" on the server. As a result, not only did we spend hours
scanning through all of our major vendors URLs to get them added to the
exception list, but we are now in the process of going through all of the
URLs for the individual ejournal titles that we subscribe to through
different publishers (about 50 right now).

This implies that every time we acquire a new e-journal from a different
vendor, purchase access to a new database, or even want to set up a database
trial, that we will have to contact DIS in Little Rock to have this URL
added to the exception list. As you might expect, the needs of the library
are not at the top of their list and as a result, they are not very quick
about making the necessary changes to the exception list.

Their argument:
They argue that 80% of web page requests are for static pages. The web
caching system is therefor supposed to increase network speed by serving up
a cached version of that page the next time it is requested, thus reducing
network traffic.

It seems to me that, for most libraries, the opposite is true -  80% or more
of a libraries' Internet activity goes to online subscription databases
where users perform dynamic, unique searches every time.  Even as I type I
am getting calls from our users who are being prompted for passwords where
they should be authenticated by IP.

Has anyone else encountered this problem. Should it even be a problem?

Much Thanks

~~~~~~~~~~~~~~~~~~~~~
Bryan H. Davidson
Electronic Products Librarian / Webmaster
University of Arkansas Libraries, Fayetteville
Ph. 501-575-4665