A key with parent and child servers
is keeping them replicated in order for individual targets to
be deployed with the appropriate content.
When building a multiple server architecture, there are two main
design choices
- Several OS deployment servers connected
to a single, centralized database.
- Several OS deployment servers connected
to several databases.
Setting up the architecture and then replicating the information
from one server to another depends on these options. More details
on the two different architectures and how to set them can be found
in
the installation guide, under
the section on multiple server architecture.
It is important to remember the following points when performing
replication online:
- Each subordinate server needs to download files from its parent server.
This means that the parent server
must be up and running during the whole replication process.
- Replication can be scheduled for a specific time and repeated
at an interval specified in days. These settings are set on the subordinate
server. When set, the replication process becomes autonomous and can
be performed without human intervention. Drawbacks to this are the
relative loss of control over the process, and network and processor
usage.
- Server replication is performed by copying files from the parent server
to the subordinate server. A selection can be performed on the kind
of information that must be replicated. Files that have been modified
are copied over.
Note: If you use the config.csv configuration
file to setup your server architecture and its options, do not use
the links provided on the page to create your architecture. Contradicting information
provided on the web interface and
through the config.csv file leads to unexpected
behavior.
from parent to child servers
.
It can be performed through several mechanisms.