Being Dense here I guess as I don't follow how to set this up. I have GUiminer setup for a half dozen pools but due to recent DOS attacks I assume my primary is down a lot more than I like so I would like to backup to Deep.
I used to do the same thing. Juggle multiple miners for each card with different -f values and have them running. Was a pain in arse and spun up a ton of extra processes that just bogged down the machine. I've been using the new --backup setup has been working like a champ.
Just add that --backup with the correct settings to your extra flags section of your main miner. So for deepbit just add this onto the list of the current flags you are using:
--backup=YourWorkerName:YourWorkerNamePassword@deepbit.net:8332
Leave the rest of your flags the same and by default, if the miner can't talk to the main pool for 2 consecutive cycles, it'll redirect that miner to the pool you have listed in the --backup flag. Additionally, it'll attempt to reconnect to the primary pool every 2 getwork cycles. You can tweak those values with those --failback and --tolerance flags.
You'll see the failover/failback attempts in your Console log.
Of course, if both your primary and secondary pools are down, you're out of luck without doing the old multiple miner -f setup.