In order to use Double-Take Availability replication, data that will be replicated on a block device must be accessed through a loop device, which is specially attached using the Double-Take Availability loop driver (DTLOOP). DTLOOP allows the loop device to serve as an access point for operations performed on the block device so that data changes can be captured. Existing block devices may be available for replication, but the data on those block devices can only be replicated if they are accessed through the DTLOOP loop device. It is important that operations on the block device be made through the loop device only, or the operations will not be replicated. Failure to do so will result in corrupted data on the target.
DTSetup allows you to configure entries in /etc/DT/dtloop_devices to attach block devices as DTLOOP when Double-Take Availability is started.
If your block device being protected with DTLOOP has a file system on it, do not mount them from /etc/fstab. They should be mounted from an init script. DTMount must be started in the boot sequence before the script to mount the loop devices is executed in order to ensure that the loop devices have the replicated block devices associated with them. The script should then mount the loop device, not on the native block device.
When making replication configuration changes, stop any applications that may be running and restart them after the replication changes have been made. Double-Take Availability needs to be loaded on the file system before any applications, otherwise some data may not be replicated.
After the block device replication configuration is complete, applications must read and write through the /dev/loop# device in order for replication to work.
You can also attach and detach DTLOOP manually using the Setup Tasks, Configure File System or Block Device Replication, (DEPRECATED) Manual Replication Configuration menu option. Changes made from this menu are not persisted between reboots/restarts.