7 comments

[ 3.2 ms ] story [ 146 ms ] thread
all serverless products seem quite unhappy. looks like an internal dns issue
In the 1970's, Bob Kahn and Vint Cerf developed TCP/IP, making it possible to build a network of interconnected networks, or Internet, that is distributed, so it has no single point of failure: Any node could go down, and the network would route around it. DARPA funded all the research that culminated in TCP/IP, for obvious reasons.

Now, here we are, half a century later. Whenever a single entity on the Internet, AWS, goes down, it takes down with it many, if not most, of the services on which I rely every day. The Internet obviously continues to function, but... all these applications/services running on it are in effect centralized in a single "blob of failure." Sigh.

Keep your own eggs distributed across multiple boxes, and make your vendors who don't accountable.