Scalable compute services.
Simple object storage.
Run managed Kubernetes clusters.
Tools & Integrations
Automate your infrastructure.
Deploy pre-built applications.
Connect, share and learn
DevOps and development guides
Questions & Answers
Development and systems Q&A
by Justin Ellingwood
For example, NYC1 -> SFO1. In the event where NYC1 goes down during a Sharknado, the IP would automatically transition to SFO1 or TOR1 or whatever I configure, where a hot-standby droplet takes over.
Had to put a 3fer on this one. This would make DO less "platform for prototyping" and more "I might run my critical LOB application with these guys".
would be awesome!
I like this
Definitively a MUST. Would require an extra address range which is routed to all datacenters via anycast or something..
The last maintenance on NY3 datacenter kept my webserver constantly timing out during one hour. All the high availability solutions with the current floating IP implementation wouldn't work because today it's only possible to reassign the IP to droplets in the same datacenter, which in this case were entirely compromised.
It would be great to have the ability to float IPs between datacenters. It would then be possible to build a really robust high availability infra.
Digital Ocean generously sponsors rails-assets.org, and I wanted to use the floating IP feature but our only downtime so far was due to a hardware failure on one of Digital Ocean's nodes. Because there's no way to guarantee that two droplets in the same region are on different hypervisors, then we have to deploy our failover machine in a different region. So that makes the floating IP functionality pretty much useless for us. Switching our DNS to AWS's Route 53 was the best option for us.
You won't be notified about changes to this idea.