Scalability servers improve scalability and performance. The domain manager is bundled with a scalability server; additional scalability servers are optional.
The following illustration shows a typical scenario for using scalability servers in a network:
To avoid unnecessary network traffic over the WAN, scalability servers are deployed to each remote location or site in the network. By also using the software libraries of the scalability servers for permanent storage of software packages, only control data needs to be transferred across the WAN. The initial deployment of the software packages to the remote libraries can be done over the WAN, but local registration is also supported if the WAN links are very slow.
Geographical dispersion is one of the main reasons for using scalability servers, but other factors should be considered when planning the software delivery network, including the delivery window and library access methods:
The delivery window is defined by the maximum accepted time to deliver and execute a particular software package to all managed systems. The delivery window is defined by the software delivery user and not by CA Technologies.
Consider which library access method is to be used by the agents. Different access methods expose the scalability server and network to different loads. The scalability server can be configured by how many concurrent software jobs are allowed.
|
Copyright © 2013 CA.
All rights reserved.
|
|