Well for one thing zookeeper is cheaper than a load balancer. For another, putting a load balancer in the middle of all of your internal flows is a pain in the posterior and can seriously mess with your throughput.
And finally, having a program update zookeeper is whackingly easier than writing a reliable program to update your load balancer. Sent from my iPad On Aug 12, 2011, at 9:49 PM, Rakesh Rajan <[email protected]> wrote: > I've been looking at Zookeeper recently and wondered how it compares with a > hardware loadbalancer for the following usecases: > > 1) Service discovery > - Using ZooKeeper, the server would register itself on a common znode which > can be looked up by the client to get a list of registered servers. > - Using H/W loadbalancer, I can have couple of server IPs behind a VIP name > and the client is only aware of the VIP address. > > 2) Load balancing > - Using ZooKeeper, the load-balancing happens at the client side. ( > connection is directly made from the client to the server ) > - Using H/W LB, the load-balancing happens at the load-balancer side ( and > hence the connection is routed via LB) > > So for the above 2 usescase, are there any specific advantages in using > ZooKeeper over a H/W load-balancer?
