Check that the data obeys the business rules of the system and that referential
integrity isn??™t compromised when the data is loaded. Two approaches can be taken:
??? Use the web server log to extract user data.
??“ Modify scripts to capture data and archive it.
??“ The data can then be rendered anonymous as required. This kind of data suffers however from
the problems that it is true for a particular website structure and popularity level. As the website
is modified, the URLs used by a user may no longer exist; should the site become heavily used
(and thus slow), users may use it differently.
???
Data auditing
. Apart from the commercial considerations of accountability consider how you will
test the auditability of data. Which logs hold it and how securely are they kept?
???
Tool acquisition and use
. Ensure your staff has had time to accustom themselves to the tools and
that the tools work. Highly conscientious staff may refuse to accept that a tool is failing and blame
themselves. Insist on demonstrations. Use outside agencies for one-off events but develop staff
for repeat events: consultants and agencies take their ability with them. Compromise: hire a
consultant to mentor your staff. Buy load-testing tools only if their repeated use is foreseen,
otherwise use open source tools (and give staff time to learn them) or (for one-off events) hire
specialists with their own tools.
???
Early execution
.
Pages:
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208