??“ How you will test mirror sites, particularly if one crashes and traffic is re-routed
??“ How pages and scripts need to be restructured to minimize latency (see section A.7)
??“ How you will manage to upgrade servers and server farms while keeping a 24/7 site. Will the
system reconfigure itself and load-balance correctly?
??“ How you will test to show that these issues have been identified and resolved.
???
Disaster recovery
: how will you test the transfer to an emergency site?
???
Division of responsibilities
. You must ensure that there is a clear division of responsibilities
between testers and developers. It is in principle uneconomic to have testers do anything other
than look for bugs. It should not therefore be testers??™ primary responsibility either to locate the
Testing and the Web
71
source of bugs or to review code in an effort to imagine where bugs might be found. This is work
which is far better left to developers.
???
Quality of Service constraints
. The user of a service stipulates a SLA with the provider and the
latter agrees to ensure proper levels of QoS. QoS testing of a website attempts to break SLA
constraints, by provoking the system to take different paths through the service workflow, which
(may) produce different QoS values.
???
Size
. See [Reifer] for an adaptation of the CoCoMo model to the web. This yields an effort/
duration estimate 1/3 of which should be allocated to testing.
Pages:
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210