All failover options described below are configured per Job. If you want all Jobs protected by failover recovery, you must set each individually.
|
All failover options described below are configured per Job. If you want all Jobs protected by failover recovery, you must set each individually. |
To set up a file CDP failover system
You should configure a second slave Agent at your backup site. It will be kept updated with a read-only copy of the Job's data; if your master Agent is disconnected, you can then use this second Agent as a temporary master so that users can continue to read and write files as usual. When the original Agent comes back online, it will replicate any changes that were made on the temporary master, and you can then redesignate it as master. To automate this process, see Automating CDP failover.
Setting up the second, slave Agent
Install the Agent on the computer on which the slave will run.
Link to an Existing Job on the master Agent.
It is common to use a cohosted topology for this. That is, the second, slave Agent is on the same computer as the backup Vault. The slave Agent does not have to be a cohosted with the Vault, but if it is not, the Vault will have to access the slave Agent remotely in order to automatically become the new master.
|
If you plan to set up a cohosted slave Agent, when you first connect it to the Job, you must specify the Vault name using its IP address instead of its computer name. If you specify the Vault using the computer name, you will not be able to switch the Agent's mode from slave to master. |
See Example: Setting up CDP Failover for an example configuration.