3. Introduction
A
software system that operates on its own or with a minimum of
human interference according to a set of rules
To
increase productivity while minimizing complexity for users,
capable of running themselves and adjusting to varying
circumstances
Control
theory, adaptive algorithms, software agents, robotics,
fault-tolerant computing, machine learning, artificial intelligence,
and many more
3
5. The Growing Complexity Problem
•
Average increase of computing devices 38% per annum
•
Ratio of Labour costs to Equipment costs around 18:1
•
Manual control is time-consuming, expensive, and errorprone
•
The distributed, diverse applications
heterogeneous tasks
deal with
5
6. The Evolution Problem
•
Maintenance and Evolutions of critical and legacy
systems
•
Keep the values of quality attributes within desired ranges
•
Monitor or verify requirements (functional or nonfunctional) over long periods of time
•
Adapt safety-critical systems without halting them
6
7. Autonomic Human Nervous System
•
•
A homeostatic system is an open system that maintains its
structure and functions by means dynamic equilibriums
Rigorously controlled by interdependent regulation mechanisms
•
A series of modifications that are equal in size and opposite in
direction to those that created the disturbance
•
The activities to maintain blood glucose level is an tremendous
example
7
8. Blood-Glucose Concentration Regulation
Glucose Concentration in Blood
(in percentage)
Activities
Less than 0.06
Tissue Starvations
Liver converts Glycogen to Glucose
More than normal
Pancreas secretes Insulin
Muscles and Skin disposes the excess
Greater than 0.18
Kidney excretes excess into urine
8
9. Ashby’s Ultrastable System
•
The goal of the adaptive behaviour is the survivability of the system
•
The system will always work towards returning to the original
equilibrium state in case of disturbances:o Frequent small impulses to the main variables
o Occasional step changes to its parameters
Fig. Essential variables
9
11. Human Nervous System as an Ashby’s
Ultrastable System
Fig. Nervous system as part of an Ultrastable system
11
12. Conceptual Model
•
A system that operates and serves its purpose by managing itself
without external intervention even in case of environmental
changes
Fig. Conceptual prototype of Autonomic System
12
13. Architecture
•
A closed control loop in a self-managing system monitors some
resource and autonomously tries to keep its parameters within a
desired range
Desired
Range
?
Control
Resource
Measure
Fig. Control Loop
13
18. Miscellaneous Characteristics
Other recommended attributes include
• Automaticity
• Adaptive
• Aware
• Reflexivity
• Transparency
• Open Source
• Autonomicity and Evolvability
• Easy to train and learn
18
19. An ode to Policies
•
Policies are a form of guidance used to determine decisions and actions
Fig. States and Actions
•
Action, Goal and Utility policies are its types
19
20. Contemporary versus Autonomic
Computing
Concept
Current Computing
Autonomic Computing
Self-configuration
Time consuming and error prone
System adjusts automatically or
by policies
Self-optimization
Manually set, ever increasing
parameters
Improve their own performance
and efficiency
Self-healing
Few weeks to solve problems
System automatically detects,
diagnoses, and repairs problems
Self-protection
Manual detection of and recovery System automatically defends
from attacks
against malicious attacks
20
23. Benefits of Autonomic Computing
Short-term I/T related benefits
•
•
•
•
•
•
•
•
Simplified user experience
Cost-savings - scale to use
Scaled power, storage and costs that optimize h/w & s/w usage
Full use of idle processing power
Natural language queries allow deeper and more accurate returns
Seamless access to multiple file types
Stability, High availability. High security system
Fewer system or network errors due to self-healing
23
24. Benefits of Autonomic Computing
(contd.)
Long-term, Higher Order Benefits
•
•
•
•
•
Realize the vision of enablement by shifting available resources to higherorder business
Embedding autonomic capabilities in client or access devices, servers,
storage systems, middleware, and the network itself. Constructing autonomic
federated systems
Achieving end-to-end service level management
Collaboration and global problem-solving
Massive simulation and complex calculations which require processors to run
24X7 for as long as a year at a time
24
27. Conclusions
•
Users demand and crave simplicity in computing solutions
•
A system used by millions of people each day and administered
by a half-time person seems attainable with the notion of
automatic updates
•
It will take another decade for the proliferation of Autonomicity
in existing systems
27
28. References
[1] IBM Corporation. An architectural blueprint for autonomic computing.
April 2003.
[2] Manish Parashar and Salim Hariri , Autonomic Computing: An Overview,
The Applied Software Systems Laboratory, Rutgers University, Piscataway
NJ, University of Arizona, Tucson, AZ, USA.
[3] Hausi A. Müller, Liam O’Brien, Mark Klein and Bill Wood , Autonomic
Computing, April 2006
[4] IBM, “Autonomic Computing: IBM’s Perspective on the State of
Information Technology”,
http://www1.ibm.com/industries/government/doc/content /resource
/thought/278606109.html.
[5] IEEE Computer Magazine, Jan 2003
28