Problem: It is
becoming increasingly difficult to justify our necessary
machine upgrades to management in an independent and
objective way.
Solution:
Predictive projections based on historical resource use
offer a visual representation of the requirement that
can be easily understood by non-technical management
teams. A predictive graph based on
transactions-per-hour, allows managers and Management to
easily see that escalating resource use is based on
greater throughput of transactions by a static number of
users, or that sustained transaction rates are
increasing due to a rapidly growing user base.
Problem:
Is there any way to determine the likelihood that future
instances of managing an important group of jobs, such
as QZDASOINIT, will impact our CPU and by how
much?
Solution: Any
history of this type of problem can be utilized in the
predictive projection summary and a valuable guideline
produced as to how these jobs might perform in the
future and their associated CPU usage quantified. This
can be particularly helpful in determining the likely
impact on users who may be carrying out urgent
processing tasks at a particular time and if disrupted,
could directly impact business processes.
