Why Does the Cloud Stop Computing? Lessons from Hundreds of Service Outages

muratbuffalo.blogspot.com

25 pointsdeafcalculus3 comments

When a service makes significant SLA promises, usually it is not a promise of uptime but a promise of what the service provider will pay you for the lack of service.

What are some examples of recovery blocks and N-version software mentioned in the blog post?

Here's a few links along these lines that might interest you. They are an intro, recovery-oriented computing, a survey that includes N-version, and a recent survey including some byzantine fault-tolerance.

http://crystal.uta.edu/~kumar/cse6306/papers/FaultTolerantDi...

http://roc.cs.berkeley.edu/papers/ROC_TR02-1175.pdf

http://www.pld.ttu.ee/IAF0030/Paper_4.pdf

http://sirocco2016.hiit.fi/preproceedings/Recent_Results_on_...

Reply on news.ycombinator.com