What do you do when you deliver on time, under budget and the system still fails?
Anonymous
I noticed recently in Australia 2 big projects which delivered on time, under budget but when the chips were down, the systems failed. Both appear to be due to software being unable to cope with a larger than expected load. How can project managers "recover" from a situation where the perception is that something was missed? Saving Changes...
Michael WoodProject Manager / Business Analyst / Business Process Improvement Guru| Independent ContractorGig Harbor, Wa, United States
Stress testing is part of any development effort. The project did not come in on-time and under budget because it failed to function in the environment it for which it was intended. However, this assumes that the volume loads were properly disclosed. Recovery comes in two flavors. 1) Fixing the throughput problem and 2) Fixing the perception.
The first step is to get everyone focused on the problem and off of the people.
Step 2 is to understand why the architecture breaksdown under volume. It could be a hardware issue or it could be software. Hardware is easy - spread the data out over multiple servers or drives (separate controllers); up the memory, etc.
Software is something else. For example, if the application is a data warehouse and it was implemented with a normalized data structure then the effort is reengineering related. Or, if the database engine can't handle the stress loads then its a migration to a new DBMS.
The point is that the cause of the throughput issue must be fully understood before remedy can be affected.
To be sure there will be more costs, time and anguish to repair the failure.
Step 3 is to repair the relationship issues that may have been caused by the failure and to get a 3rd party's opinion as to the viability of the remedy (the project team is most likely less than credible).
Step 4 FIX IT
Step 5 LEARN from the experience
Saving Changes...
