but if u are stuck with a switcher, I would keep the switcher ground plane separate from the OCXO ground plane. I would use a pi network of shunt (470 uf || 1 uF ceramic), series 10 uH (minimum), shunt (470 uf || 1 uF ceramic || 0.1 uF ceramic). And keep the inductor way far away from the switcher's inductor.
I use a lot of SMPS for RF circuits in low noise environment. My experience is some oscillates!!! Just because you copy the application schematic doesn't mean it's stable. Ha ha, I claimed my fame in one company solving a camera image noise problem where few engineers spent a year trying to fix the problem. 3 SMPS chip oscillated!!! They looked everywhere except the simple chip supply. It was Linear Technology. My experience is they are not very good in debugging their products. I actually went to their facility to demo one of their chip lock up on power up if I do it in certain way. They were red faced!!!!
A passive filter will work to remove AC noise and switching transients, but it will not remove slow varying DC that may result in very low frequency phase noise.
So if you want to avoid kHz or more off-carrier phase noise due to a noisy power supply, a passive well designed filter may help, but if your trouble is in very low frequency variations (Hz range or less), then a passive filter is not a practical solution. You may consider a LDO linear regulator in between the OCXO and the SMPS. Due to the low voltage drop, dissipation isn't a problem mostly.
Before implementing any solution, why not first determine whether phase noise is a problem for your application? If it is a problem, make sure you can determine whether a proposed solution does help.