On Feb 26, 2012, at 5:56 PM, Randy Carpenter wrote: > We require 1 or 2 very small virtual hosts to host some remote services to > serve as backup to our main datacenter. One of these services is a DNS > server, so it is important that it is up all the time. > > We have been using Rackspace Cloud Servers. We just realized that they have > absolutely no redundancy or failover after experiencing a outage that lasted > more than 6 hours yesterday. I am appalled that they would offer something > called "cloud" without having any failover at all.
Pardon the weird question: Is the DNS service authoritative or recursive? If auth, you can solve this a few ways, either by giving the DNS name people point to multiple AAAA (and A) records pointing at a diverse set of instances. DNS is designed to work around a host being down. Same goes for MX and several other services. While it may make the service slightly slower, it's certainly not the end of the world. Taking a mesh of services from Rackspace, EC2, The Planet, or any other number of hosting providers will allow you to roll-your-own. The other solution is to go to a professional DNS service provider, e.g.: Dyn, Verisign, EveryDNS or NeuStar. While you can run your own infrastructure, the barrier for operating it properly is getting a bit higher each year in doing it "right". I was recently shown an attack graph of a ~200Gb/s attack against a DNS server. *ouch*. Sometimes being professional is knowing when to say "I can't do this justice myself, perhaps it's better/easier/cheaper to pay someone to do it right". - Jared (Disclosure: I work for one of the above named companies, but not in a capacity related to anything in this email).
