jbminn

2 minute read

Cloud computing is here to stay.

Thank you, Captain Obvious.

Software (or platform or infrastructure) as a service, hosting, service providers… whatever term you use, it all ultimately refers to the same thing:

You’ve placed computing processes onto an internet-connected server that you don’t own

The reasons are almost completely economic in nature:

  • Providers like AWS and DigitalOcean can rent you a server. There's no CAPEX
  • Your computing expense is truly pay-as-you-go. Providers only bill for the real clock time used with no minimums
  • If you only need a computing resource for a couple days, you aren't forced to purchase any equipment that is then immediately under-utilized
  • Conversely, if you need to run an experiment requiring thousands of servers, you don't need to buy even one. Rent one thousand instances of the most powerful server you need, run your tests, then destory those instances
  • Nearly every provider has a well documented API that allows your team to quickly integrate these rented computing resources with near-zero custom development
  • Those APIs mean that a vibrant ecosystem of 3rd party services exist allowing you to drop in extra functionality as needed
  • AWS + DigitalOcean's dashboards are so easy to use your team requires no IT support. Zero. Turn off the pagers + send 'em home.

…and on and on.


What does this mean for dev teams inside legacy enterprises? The answer is simple but important:

Leagcy enterprises must not only adopt cloud computing in their enterprises, they must migrate some or all of their legacy applications & infrastructure to the cloud

What if a majority of the organizations’s computing infrastructure is itself legacy - old, physical servers with proprietary OSes running homegrown applications that don’t have any obvious migration path short of a complete rewrite? These old servers are often directly responsible for legacy app failiures - the server exhibits issues long since fixed in newer hardware and OSes, causing outages. How does a team remove the risk of the failing hardware while still running the legacy application?

Migrate the entire server to the cloud

This is precisely the problem Tincup solves. Enterprises can use Tincup to migrate a running legacy server from behind the firewall to the cloud.

comments powered by Disqus