With the DFM backup command you are limited to two options. The first is the traditional long running dump style backup which at any scale I've never seen go less than about half an hour. About the only thing you can do is just wait. Using automatiion, I've always worked it into a batch file that would later send email status or check back into the automation system when it's done, since during the process there is nothing to do but wait.
The other options is to deploy DFM on LUNs attached to the host. There are options on the DFM backup command to then use Filer based snapshots to take a consistent database backup in very short times as expected. The kicker of course is that you have to be using block protocols on your Filer directly mapped to the DFM host - something I see in practice less and less (either one isn't using block procotols at all - fairly common - or if using block it's behind an ESX infrastructure where it is hard to impossible to establish an RDM directly from the VM Guest to the Filer because that is just not the way we do it").