Chapter1 MECHANISM: Understandingperformanceissuesandtheirimpact THETIME-TO-LIVEBASEDCONSISTENCY

Abstract The Web is a large distributed database were copies of objects are replicated and used in multiple places. The dominant consistency mechanism deployed for HTTP (Hyper Text Transfer Protocol) and DNS (Domain Name Service) records is Time-to-Live (TTL) based weak consistency. Each object has a lifetimeduration assigned to it by its origin server. A copy of the object fetched from its origin server is received with maximum time-to-live (TTL) that equals its lifetime duration. Cached copies have shorter TTLs since the age (elapsed time since fetched from the origin) is deducted from the objects lifetime duration. A request served by a cache constitutes a hit if the cache has a fresh copy of the object. Otherwise, the request is considered a miss and is propagated to another server. With HTTP, expired cached copies need to be validated, and if they turned out to be not modified, we refer to the request as a freshness miss. Cache performance is affected by TTL-based consistency. Cache misses translate into user-perceived latency. When a popular cached object expires, a cache can reduce user perceived latency by refreshing its copy proactively, before it is requested. For hierarchical caches, the number of cache misses depends in subtle ways on the age of the copies the cache receives. Thus, fresh copies obtained through another cache are less effective than fresh copies relieved from an authoritative server