Using A Pool Of Static Ip Addresses; Using A Single Static Ip Address - Dell DX6000 Application Manual

This method is supported and recommended for all deployments.
Using some DNS servers, it is even possible to move the maintenance of the PAN addresses out
of the applications and into the DNS itself. BIND, the most commonly used DNS server, allows an
administrator to enter multiple A records that map a single DNS name to more than one IP address.
This again requires static IP addresses, but it enables the application to use a single DNS name
for the entire cluster. The DNS server will select one of the defined IP addresses on a round-robin
basis. If one of them does not respond, the application must resolve the host name again.

2.2.4. Using a Pool of Static IP Addresses

A less desirable approach than either of the preceding is for your application to address a DX
Storage cluster using a stored list of several, perhaps all, of the static IP addresses for the nodes in
the cluster.
This method is supported but not recommended for a production environment.
This application's stored list of IP addresses must be accessible programmatically from the
application. If one of the nodes fails to respond to a request, then the application can simply try
another one.
If a redirect response reveals the presence of a storage node not in the original list, the application
should be able to add the new IP address to the list. If the cluster is relatively stable with respect
to static node IP addresses, this may be a good approach. If, on the other hand, nodes are being
added to and removed from the cluster often, you should not use this method.

2.2.5. Using a Single Static IP Address

The simplest but least recommended (and least supported) way for an application to address a DX
Storage cluster is to assign a static IP address to at least one of the nodes of the cluster and then
use that IP address in every request.
This method should be used only in a development environment. It can be set up quickly but is not
maintainable in a larger system.
The simplicity of this approach is balanced by the significant disadvantage that, if the sole PAN is
taken out of service or fails completely, the application cannot send requests to the cluster, even
though other nodes might still be functioning and all desired content still available.
Copyright © 2010 Caringo, Inc.
All rights reserved 11
Version 5.0
December 2010