Webserver

Based on the File protocol workload model, the Webserver Example is designed to mimic an in webserver in terms of block sizes, command ratios, and typical file sizes. There is no need to take into account access within files because most files are read/written completely rather than in portions.

Webserver access is often very read heavy. The amount of metadata content in this example is quite low compared to other example where the amount of meta command access can often be as high as 95% of the traffic. This depends on the selected webservers caching algorithms. Webservers with high amounts of meta traffic are often GetAttr (as high as 62%), Lookups and Access heavy so a “commands distribution” setting should be used if the metadata is increased.