Architecture 201

Service Oriented Architectures

Fault tolerance, fault protection, masking, dependability fundamentals

Fail open, fail closed

Perspective: node, network, cluster, application

Caching Concerns

Static assets

Data

Eviction and replacement policies and evaluation

Approaches

(TTL, purge-on-write, no-purge versioning, constantly churning cache versus contained, working set sizing)

Crash only

Synchronous vs. Asynchronous

Business continuity vs. Disaster Recovery

Designing for Scalability: Horizontal, Vertical

Simplicity

Performance

Tiered architectures

MTTR > MTBF

http://www.kitchensoap.com/2010/11/07/mttr-mtbf-for-most-types-of-f/