Gravity of photons in different reference frames I know that photons have gravity because they contribute to the stress energy tensor, but this means that observers in different reference frames experience a different gravitational force from the same photon. How is this resolved? Doesn't this mean that an object interacting with the photon in reference frame A would interact differently in reference frame B? 
 A: To summarize in a sentence: yes, the "interaction" is different in frames A and B, but by construction Einstein's equations of gravity are still obeyed in both frames.
This is a fundamental principle in relativity: while physical quantities may look different to different observers, the physical behavior measured by those observers must be consistent.
As a simple example in special (rather than general) relativity, consider an infinitely long rod with a uniform negative electric charge density.  An observer A who is stationary with respect to the rod releases a positive charge Q some distance away from the rod.  The charge Q will be attracted to the rod and accelerate towards it.
Now, consider an observer B moving at some speed $v$ in a direction parallel to the rod.  In this frame, the rod appears to be moving, and therefore will appear Lorentz contracted.  This will cause the charge density of the rod to appear larger in frame B than in frame A, and therefore you might expect that observer B sees Q accelerate towards the rod faster than it did in frame A.  But wait!  Since the rod is moving in frame B, it serves as a current and therefore produces a magnetic field.  This magnetic field conspires to decrease the acceleration of Q towards the rod, and therefore the physics described by the two observers A and B will be consistent.  This apparently magical consistency stems from the fact that Maxwell's equations are relativistic.
(By the way, I was a bit cavalier and ignored time dilation effects; the point is that although the observers A and B would apparently describe different physical systems, their different descriptions give the same prediction for the behavior of the charge Q.)
The story is much the same in general relativity: different observers in different frames may disagree on the exact distribution of energy (or how much "energy" is contained in a given system like a photon), but Einstein's equations are covariant, so that these different observers will agree that Einstein's equations are obeyed in both of their frames.
