Performance-Critical Systems (PCS) Initiative Introduction
Software-intensive systems must perform successfully under
adverse circumstances (e.g., under heavy load or in the presence of subsystem
failures). And yet, the behavior of systems under such
circumstances is often less than acceptable.
|
Benefits
- Potential performance and dependability problems are identified early
enough to make acceptable cost, schedule, and functionality tradeoffs.
- Performance and dependability anomalies are found infrequently during
system integration and even less frequently in operational use. Any problems
that do arise are readily corrected.
- System reliability and availability is increased through the use of more
effective design. In particular, the impact of various dependability
strategies on system performance can be modeled at a high level of system
design.
- Delivered systems are more easily adjusted to respond to changes in system
load and evolving requirements and technology.
- System tests reliably predict the behavior of heavily loaded systems.
|
|
Contact Us
Contact Terry Dailey to discuss collaboration opportunities.
Contact SEI Customer Relations for more information about PCS or the SEI.
|
|
Background
The need to address
performance and dependability issues is obvious in real-time systems (such as
flight control software), but failure to address these issues early can lead
to unexpected problems that make systems virtually
unusable until costly repairs are undertaken.
The significant performance problems that software-intensive systems
can experience usually arise from
deficiencies in the
- state of the art: the needed practices are not
sufficiently mature to detect or resolve the problems in advance
- state of the practice: existing and effective
practices are not routinely used
Goals
- mature technologies that can be used to detect and resolve
performance and dependability problems before a system is tested
or fielded
- provide education and training materials on methods for specifying, modeling, and analyzing system architectures and
components to predict system performance and dependability
properties
|
|
return to PCS main page
|
