View Full Document

11 views

Unformatted text preview:

End User Web Availability Matthew Merzbacher visiting researcher UCB CS Dan Patterson undergraduate UCB Measuring Availability Measuring nines of uptime is insufficient Reflects unrealistic operating conditions Need to capture end user s experience Server Network Client Client Machine and Client Software Experiment Hourly small web transactions from two sites Mills CS Berkeley CS To a variety of sites including Internet Retailer including international sites Search Engine Directory Service also including int l sites Ran for 6 months Results Success completeness and speed Raw Ignoring client problems All Retailer Search Directory 93049 93108 93456 92668 98877 98870 99349 98569 Ignoring client and network problems 99908 99756 1 0000 99972 Ignoring client network and one time errors 99945 99845 1 0000 99991 Put Another Way Client Errors Severe Network Traffic Corporate Errors Medium Network Traffic Server Errors Client Hardware Problems Dominated User Experience System wide crashes Administration errors Power outages And many many more What About Speed 60 50 Time sec 40 retailer 30 search directory 20 10 0 0 95 0 96 0 97 0 98 Availability 0 99 1 Does Retry Help Error Type Total Client 0 26667 Medium Network 0 86207 Severe Network 0 78947 Server 0 91111 Corporate 0 42105 Retailer 0 27059 0 87037 0 92308 0 78571 0 31250 Search 0 26471 0 92857 1 00 1 00 1 00 Directory 0 26471 0 83178 0 68889 0 96 n a What Guides Retry Uniqueness of data Importance to user Loyalty of user Transience of information And more


Access the best Study Guides, Lecture Notes and Practice Exams

Loading Unlocking...
Login

Join to view End-User Web Availability and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view End-User Web Availability and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?