software performance testing
When I think of performance testing I am not thinking of testing how well a person performs. No. I am thinking of how well an application or process performs. Performance testing includes load testing, stress testing, volume testing, resilience testing (well some types anyway) and failover testing. So do the builders and implementors of applications do enough performance testing? No. Simple as that. No.
“Through 2004, 20 per cent of enterprise mission-critical applications will experience severe performance problems that could have been avoided.” That’s from the Gartner Group, Note Number: R-21-550. It would be interesting to see if they can compare this prediction with the actual experience now that 2004 is completed. If they didn’t get their predictions about right would they still be in business? I suppose if everyone else was wrong too they would survive. But would they be a leader? So lets take it as pretty likely that 1 in 5 mission critical applications had a performance problem and lets assume that a number of these could have been detected by performance testing and monitoring before the problem occured. Lets also estimate the number of large companies in the UK that could be impacted by application performance problems and the number of mission critical applications they depend upon. Finally estimate the averge impact on a large corporation of a performance problem (use whatever metrics you like: lost income, downtime, etc etc). Now your mission (should you choose to accept it) is to estimate the cost to the UK of not doing performance testing and performance monitiring effectively. Please send your answers here and I’ll collate them all for a future blog.
Permalink Comments off