How To:Use Amanda Applications on a Client
This article is a part of the How Tos collection.
The Application API provides for pluggable applications that are designed to back up specific kinds of data. Applications are executables that live on the Amanda client. Note that, for security reasons, you cannot provide a full path to the executable; it must reside in the application directory.
Configuration
Define the my_app_engineering application using the myapplication binary.
define application-tool my_app_engineering {
comment "a comment"
"my_app" # inherit config of the my_app application
plugin "myapplication" # name of the application
property "mailto" "engineering" # can set property
}
Now you have an application named my_app_engineering. Next, define the engineering_app dumptype using the my_app_engineering application-tool
define dumptype engineering_app {
program "APPLICATION"
application "my_app_engineering"
}
Any DLE with dumptype engineering-app will then use myapplication.
You can get a bit fancier than this, using nested definitions: Define the my_dumptype_root using a modified my_app application
define dumptype my_dumptype_root {
program "APPLICATION"
application { # define a custom application
"my_app" # inherit setting from another application
property "mailto" "root" # override property
}
}
DLEs using my_dumptype_root will use the myapplication plugin to back up the client.
Available Applications
For the most up-to-date list, see amanda-applications(7).