Objectives:	Learn the principles of evaluating the entire system stack, identifying single points-of-failure. Learn to examine the options for protecting against the failures without “breaking the bank”. Learn to build contingencies for when you encounter the dreaded double/triple/… point-of-failure.
Abstract:	This session will examine the architecture of the ORACLE database supporting the Environment Canada Scientific Archive, a multi Terabyte database, processing data 24/7, but without 24/7 support. We will discuss the architecture starting from the hardware up, defining the single points-of-failure and what can be done to protect against them. Topics covered will include server hardware and networking, RAC on Linux, ASM, storage hardware, backup & recovery and standby databases. Real-life stories of write cache failures and storage meltdowns highlight the need for multiple layers of contingencies and careful consideration of Return to Service commitments. We will also examine the challenges of doing all this with opportunistic funding only.
Audiernce Focus:	Architects
Expertise:	Intermediate
Track:	Architecture
Date:	Wednesday, April 16, 2008
Time:	9:45 AM - 10:45 AM