Re:Immutable Infrastructure and Disposable Components
by KevinEldon on Tuesday July 02, 2013 @12:01PM
(#5978711)
And also PhoenixServer: http://martinfowler.com/bliki/PhoenixServer.html
Re:Immutable Infrastructure and Disposable Components
by mwcampbell on Tuesday July 02, 2013 @12:28PM
(#5978933)
I like the concept. But where I work, our infrastructure is currently hosted on leased dedicated servers, where (last time I checked) OS reloads require manual intervention from the hosting provider and cost something. So I guess the burning down and rebuilding would have to be done on a layer above the base OS, e.g. Docker containers.
Re:Immutable Infrastructure and Disposable Components
by mwcampbell on Tuesday July 02, 2013 @11:38AM
(#5978515)
I'm conflicted on this.
On the one hand, the phrase "snowflake server" encapsulates my observations about long-lived servers that are upgraded and extended in place.
On the other hand: "Perfect replication is the enemy of any robust system." -- http://en.wikiquote.org/wiki/Daniel_Suarez (read the other quotes from the same chapter for context)
Re:Immutable Infrastructure and Disposable Components
by tome on Tuesday July 02, 2013 @12:43PM
(#5979053)
But this "immutable infrastructure" idea doesn't mean that the infrastructure is perfectly replicated across machines. It aims for perfect replication through time. That's something quite different.
Re:Immutable Infrastructure and Disposable Components
by mwcampbell on Tuesday July 02, 2013 @12:52PM
(#5979131)
You have a point. Indeed, the practice of building specialized machine or container images, which seems to be the best way of achieving immutable infrastructure, may make servers more heterogeneous, since each image will contain only the pieces necessary to perform its specific function. Done right, immutable infrastructure may be more resilient in the face of parasites.
http://martinfowler.com/bliki/SnowflakeServer.html