Information Technology Reference
In-Depth Information
3.4.8 Operational Readiness and Pilot Operation
Our discussions up to now have assumed that the acceptance procedures for a given
release should suffice to clear software for regular operation and production. For
highly complex systems with many interfaces and individual modules to be tested
separately demands are higher. In this case the customer requires a higher degree of
security. To meet this, an additional operational readiness check, including or not a
pilot operation phase, may be provided.
3.4.8.1 Operational Readiness Check
The aim is to make sure that all services necessary for the deployment of a system
as ordered by the customer have been provided by the supplier and are ready for
operation. This may include more than just software: In addition there may be:
Utilities necessary to support production
Access security concepts
Multi-client capability
Separate standard reporting facilities
Documentation
In certain cases production schedules and operating system support
Hotline and other support functions.
Since all functional tests have been passed already at this point in time the
operational readiness check need no longer include those. A stock taking of the
overall organisational, technical and operational situation for a given date should
suffice. This will be documented separately (see template “Operational Readiness
Declaration” under Sect.
3.5
).
3.4.8.2 Pilot Operation
Equally, for pilot operations dedicated tests are no longer required, since accep-
tance procedures will have come to an end. But prior to it operational readiness
checks as outlined above should have been done. The type of pilot operation
depends on the deployment philosophy initially agreed upon. It therefore could
for example take place on a parallel system. This is recommended in case that the
existing production configuration has to be migrated only. The disadvantage is that
for a realistic pilot operation the complete production environment with all its
operational facets has to be mapped, all professional resources being usually
employed in daily business have to be recruited.
Another possibility is the monitoring of the productive system (after acceptance!)
under load. The disadvantage is obvious: it resides in the risk starting operations with
all business processes without prior rehearsal. In practice one has to define two
