Friday, July 10, 2015
Massive corporate outages: lessons from the past
So, the NYSE trading floor outage on Wednesday was due to a “botched move” – colloquial for “elevation” or “promotion”.
News stories have compared modern Internet systems uptime reliability to Ma Bell’s in the monopolist past – with the enormous manpower it took to keep reliability.
United Airlines's problems were due to a bad router. WSJ's were unexplained.
I remember how we used to brace for “moves” – with a Wednesday noon deadline for freezing changes, with most elevations only on Fridays for weekend cycles. In the early 1990s, most shops adopted change-control management (like CA-Librarian) that froze various components and guaranteed move integrity,
It’s also interesting right now how recollections from my last two years in a customer service workbench support (2000-2001) helps me solve problems “on the road” today.