Re:Amazon ElastiCache Details - Managed Memcache
by wpietri on Tuesday August 23, 2011 @12:34AM
(#2915353)
As a startup guy and recovering sysadmin, I really admire the Amazon Web Services effort. It started out small and ugly, but they have marched ever onward, slowly improving things and adding new features. Five years of steady plodding can get you pretty far.
Re:Amazon ElastiCache Details - Managed Memcache
by wpietri on Tuesday August 23, 2011 @01:37AM
(#2915487)
It's entirely meant as a compliment. There's no way I want an infrastructure supplier to seem hurried. Internally it may have seemed as fast as possible, but externally it was never showed signs of being faster than possible, which is an amazing achievement. For our industry, almost miraculous.
Those are all legitimate gripes, and one by one Amazon has knocked them all down. Had you tried solving all of the problems before launching, I imagine you never would have gotten anywhere. The steady pace and it's-ready-when-it's-ready releases are definitely inspiring to me.
Re:Amazon ElastiCache Details - Managed Memcache
by petervandijck on Tuesday August 23, 2011 @11:20AM
(#2916909)
I'm always very impressed with the continues release of awesome features. I'd love to see your roadmap! Just knowing that, whatever you need or desire, AWS will probably release it at some point soon-ish, is awesome.
Re:Amazon ElastiCache Details - Managed Memcache
by rcaught on Tuesday August 23, 2011 @01:48AM
(#2915513)
Except running memcached on an EC2 instance is not equivalent to the ElastiCache offering. With ElastiCache you are paying for a redundant, managed service that can be scaled and monitored through a few clicks of a webpage or API.
Outsourcing these type of things is potentially justifiable to a developer whose wallet is full and many hats touch the ceiling.
Re:Amazon ElastiCache Details - Managed Memcache
by jacques_chester on Tuesday August 23, 2011 @05:14AM
(#2915933)
> In addition, Amazon ElastiCache automatically detects and replaces failed Cache Nodes, providing a resilient system that mitigates the risk of overloaded databases, which slow website and application load times.
And I think you folks were very smart to start small and improve incrementally. As an example of what I mean, take this Slashdot comment from the EC2 launch a few years back: http://developers.slashdot.org/comments.pl?sid=329797&ci...
Those are all legitimate gripes, and one by one Amazon has knocked them all down. Had you tried solving all of the problems before launching, I imagine you never would have gotten anywhere. The steady pace and it's-ready-when-it's-ready releases are definitely inspiring to me.
Outsourcing these type of things is potentially justifiable to a developer whose wallet is full and many hats touch the ceiling.
Obviously there is overhead, but if you have machines already and they are not maxed out on RAM then the $10/GB price is not far of the real cost.