I'm going to be doing some advanced research (and surely testing and implementation) on reliability of certain applications. Even up to a near-real-time disaster fail-over configuration for a very complex application infrastructure.
What's interesting about this particular situation is that the current production deployment is so complex that is is almost certain that we won't find some of the possible problems until we've tried the simulated failure. Some things are just likely dependent on the current infrastructure configuration, from load-balancing, routing, application tier configuration, certificates, and so forth. Given the recent denial of service (DoS) attacks on Social Networking sites that have very large distributed Internet presences (presumably HA), and the success of those attacks, I'm curious to dig in and see what we can learn and apply in practice for our services.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment