Here are some of the common problems you may encounter, and suggested solutions:

Problem: Glop.
Glop is a tank full of jelly rather than biodiesel. I've only had glop once, and it was when I tried to make biodiesel using really acidic oil. If you have to add more than 3 grams of lye to neutralize the acid in the titration, then I wouldn't recommend using the one-stage method (see this discussion of two stage acid method for high free fatty acid oil). I have found that the oil is much more acidic in the summer, probably because it's sitting out in the hot sun. So in the summer, I'd recommend taking your titration kit with you to the restaurant to test the oil before you bother to take it.

Solution?  I tried rescuing my glop by reprocessing it, but it didn't really work. I tried several one-liter test batches using various amounts of lye, or no lye at all. The no-lye redo worked best in the test-batch, and although it did give me two layers with the larger batch, the biodiesel was not good, and plugged up my fuel filter. So if you get glop, you may have to dispose of it.

Problem: Fluffy middle layer.
This usually indicates that you had water in your oil.

Solution?  I'm not sure you can rescue the biodiesel on top of the fluffy layer.  You can try reprocessing some with different amounts of lye, to see if any more glycerin forms. (Girl Mark Alovert recommends reprocessing 1 liter of Biodiesel with 50 ml methanol and 1 g of lye).  Next time, before you add methoxide to your oil (after getting the oil up to 130 degrees) drain off some from the bottom and examine it visually and by feel to determine if it seems more like water than oil. Keep draining until you get to the oil! And make sure the restaurant has a good lid on its oil drum, so no water is getting in.

Problem: Not getting two layers.
Your amount of lye is probably off.

Solution? Do the titration carefully with each batch of oil and scale up to the amount of oil in your processor. Trust the titration; with decent oil that has no water, the reaction should work every time. Heating the oil and using a pump that provides vigorous mixing also helps produce a good reaction. Also, be sure you're using enough methanol. The recommended amount is 22%; the extra will help push the chemical reaction to completion.

Problem: Emulsification.
This usually occurs when you bubblewash too vigorously or too soon after processing.

Solution.  There is a great section on breaking emulsions using waste glycerin here.  I have used this method and it works!    However, you will need to let the glycerin settle out for several days again. To prevent future emulsions, be sure to: heat the wash water, bubble very gently for the first couple of washes, and add a little vinegar to the first wash or two.

Problem: Fuel filter plugging.
This happens either when your biodiesel wasn't made properly or if you haven't let it sit for enough time to allow the glycerin to fall out of solution (or haven't filtered it properly). This has only happended to me a couple times- once when I used biodiesel that had only sat for two days (before I knew about the importance of settling time); another time when I tried using biodiesel that I had tried to make by rescuing glop; and a third time after letting my car sit for 5 days in pretty cold weather. The last time, my car ran OK, but my engine malfunction light came on, and this went away after changing the fuel filter.

Solution. Use decent oil (not too acidic) with no water, trust your titration, use the proper amounts of methanol and lye, heat at 130 degrees F, and stir vigorously for at least an hour and you should get good biodiesel. Then be sure to let it sit for at least 5 days before washing, filtering to 5 microns, and putting it in your car. 

Problem: Cold weather operation
Soy-based biodiesel has a gel point around 32 deg F.

Solution: If you drive your car every day, the fuel will be warmed a bit. If the temperature gets into the 20's, you should probably blend in 20% petrodiesel, although my car has done fine with an overnight temp of 26 deg F with biodiesel that had a bit of anti-gel additive added to my homebrew. (However, my homebrew is made with some canola oil which has a lower gel point). Apparently, the anti-gel additives made for diesel don't work very well for B100, although they do work for blends. They may suppress the gel point of B100 a few degrees though, which may be helpful. There is a nice calculator here that will tell you approximately how much diesel to add at a given temperature, or what temp a certain blend is good to. Of course this is approximate since it depends on the source of the biodiesel, how long it's at the given temp.