CompSci 296.2 Self-Managing Systems Shivnath Babu In this talk my aims are: - to give an overview of the work that we have been doing at Stanford on adaptive processing in the stream system and what we are doing now [past and current work] - facilitate some interesting discussion. Hope the title is provoking enough I will not be comparing our system with eddies explicitly, although I am doing some work in that respect with DeWitt and one of his students.
Today TA hours Scribing Paper discussion Summary of current work in self-managing systems Ideas for projects Will be continued on Thursday (but there is a different paper for Thursday)
Paper Summary Vision and overview paper Lists motivating factors Describes autonomic computing Evolving towards autonomic behavior Research issues
Motivating Factors for Autonomic Computing Increasing size and complexity Increasing administration cost Increasing time for administration (clarification) Operator errors outages Hard to deal with change
Time Distribution for Database Mgmt.
Ongoing Database Administration Ongoing database administration tasks, such as performance tuning, space management, system resource tuning and backup & recovery, accounts for the biggest chunk of a database administrator’s time. According to a survey conducted by Oracle, DBAs typically spend about 55% of their time performing these activities.
Irving Wladawsky-Berger Autonomic Computing “… So that instead of the technology behaving in its usual pedantic way and requiring a human being to do everything for it, it starts behaving more like the `intelligent' computer we all expect it to be, and starts taking care of its own needs. If it doesn't feel well, it does something. If someone is attacking it, the system recognizes it and deals with the attack. If it needs more computing power, it just goes and gets it, and it doesn't keep looking for human beings to step in.” Irving Wladawsky-Berger
Fundamentals of Autonomic Computing Self-configuring Dynamic addition, change Self-healing Recovering from “failure” Self-optimizing (Query) optimizers, index “advisor”, dealing with change Self-protecting Intruders, operator errors
Evolution towards Autonomic Comp. Basic Managed Predictive Adaptive Autonomic Standards
Strong Points Presentation
Weak Points Separating “mechanisms” from “policies” – separating goals from steps Comparisons with competing projects In-depth discussion of a system (e.g., databases, web servers) No discussion of whether autonomic computing is the right solution to the problem
Discussion Goals Vs. steps Self-* (consider different systems/settings) Relative importance, mechanisms, algorithms Evolution Vs. revolution Facts, e.g., causes of outages Placing current systems on the autonomic spectrum Complexity of autonomic computing itself