Performance tracking and monitoring is often a problem with many applications. Especially in systems with distributed SOA architecture, its difficult to identify the services causing performance hit. So its nice, wise and safe always to profile heavy operations in exterprise applications. Theres always a bad day when suddenly system starts showing its true colors by giving sub-optimal performance in production environment. Life is easy at this point, if all the components have runtime performance metrics available for analysis.
Read more…
I am blogging after a while and this time nothing related to human psychology. Lets talk something about technology.
At my work, ‘m many times involved in performance tuning of the applications. May it be a batch application or UI app, the culprit is always considered as the Data Layer. Its not true always. I worked on one UI application, where we tuned the data layer to the best possible state, but no luck. Recently I came across YSlow. YSlow checks how the HTTP protocol is implemented. ‘m almost a year late on this, but as rightly said Better Late Than Never.
Check this out!!
http://developer.yahoo.net/blog/archives/2007/08/yslow-podcast-screencast.html