Boston Linux & Unix (BLU) Home | Calendar | Mail Lists | List Archives | Desktop SIG | Hardware Hacking SIG
Wiki | Flickr | PicasaWeb | Video | Maps & Directions | Installfests | Keysignings
Linux Cafe | Meeting Notes | Blog | Linux Links | Bling | About BLU

BLU Discuss list archive


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Discuss] cluster DNS servers



--On Friday, March 22, 2013 1:31 PM -0400 Dan Ritter <dsr at randomstring.org> 
wrote:

> The next best method would be to use a group of DNS servers that
> had failover for the service IP address. There's no need for

You don't need or want failover for caching name servers. DNS clients have 
fault tolerance built in.

If you have a hot/cold cluster and one node fails then the other node will 
have an empty cache when it spins up. You also have no DNS at all for 
however long it takes for the cold node to spin up.

If you have a hot/hot pair with no failover then each node will have some 
cached data. If a node fails then you still have the other node's cache. 
There is no down time waiting for a cold node to spin up. The DNS clients 
will automatically rotate through their resolver lists when nodes are 
unreachable.

If you shuffle the resolver lists on each client node then you get the 
benefits of a load balancer for zero cost.

-- 
Rich P.



BLU is a member of BostonUserGroups
BLU is a member of BostonUserGroups
We also thank MIT for the use of their facilities.

Valid HTML 4.01! Valid CSS!



Boston Linux & Unix / webmaster@blu.org