Information Technology Reference
In-Depth Information
requirements can be diverse, but include requirements on future capacity. These re-
quirements are passed through the value chain in reverse direction, from customer
(market demands) to sales and further on to production and procurement. To account
for this kind of interactive control, we need another management cycle (Fig. 5).
The manager interacts with other managers in the value chain at customer and sup-
plier side. The customer manager's requests do not concern a particular service
instance, but a certain state or quality of the service as such. For instance, that the
service has a certain capacity at a specified time. This leads to certain
commitments
that are part of a REA
contract
. The synchronization of contracts is what is called
scheduling. The purpose of a schedule is to make sure that for all services the needed
resources are identified, as well as when they will be needed [6:108]. A schedule is a
collection of increment and decrement commitments, as well as mitigation plans. The
increment commitments indicate the availability of the service at some future time, or
the availability of the resource produced by the service. Decrement commitments
concern the resources (subservices or resources produced by subservices) needed to
fulfill the increment commitments. These decrement commitments must be gained
from the managers of the supplying subservices. In our conceptualization, the sche-
dule is not a separate entity but the combination of these contractual commitments.
Note that the scheduling usually runs independently from the operational service. It
only prevents the operational service to break down when the actual service requests
come. However, the scheduling may influence the operational service. For instance, if
the subservice providers are not able to commit to the required resource capacity, this
is forwarded as such via the contract monitoring, so that the service policy enforce-
ment can pro-actively find and bind other suppliers.
A second important subclass of interactive control distinguished by Simons is the
ongoing conversations on probing the assumptions underlying the diagnostic control
settings. One of the manager's interactive control tasks is to adapt the service policy
when its assumptions do not hold anymore or to anticipate such a break-down. This
can be realized by a
discourse
between managers, akin to the above-mentioned know-
ledge plane [4].
As mentioned earlier, more control systems could be distinguished - boundary sys-
tems and belief systems. Whether these can be realized as special cases of the other
ones, or deserve to be identified independently, is a question for future research.
3.5 Design Method
Following the fractal approach, a comprehensive design method for management
services (we ignore other aspects here) looks roughly as follows:
Step 1
Identify core business services
Step 2
Identify coordination and management services per core service
Step 3
Identify management subservices per management service
Step 4
Identify software services that may support any core service,
coordination service, or management (sub) service
Step 5
Identify software services that manage the software services from
Step 4
as well as subservices of these management services
The first step is general. To identify core business services, it is recommendable to
use the value cycle of the enterprise as a reference. In step 2, management services are
