Is it possible that action-reaction pairs of friction can be exerted opposite to same source of force on a body so that it pays 2x the friction force? This is going to be a long post-like question so make sure you have enough time before reading this. And this I got after thinking continuously for about 4 hours on the same question and I want to verify that if my ideology about this question is correct. First, what the question actually is?
Question
Please refer to the diagram in image
Explanation of the Image (You can skip this)
The image of the question is just below, please refer to it for better visualization. There are two blocks A and B. Block A has a mass of $m$ while block two has a mass of $2m$. They are placed on top of each other. The surface between them (Block A and B) has a coefficient of friction $\mu$ and that between Block B and ground is $2\mu$. In front of these two blocks is a wall that has a pulley attached to it (the wall). Now the block A is attached to a string which then runs over the pulley which is attached to the wall. And from that pulley, the string runs over another pulley which is attached to Block B and this string finally terminates getting attached back to the wall ( refer image for better visualization ). Now a force $\overleftarrow{F}$ is applied on the Block A.

Actual Question
What would be the minimum value of $F$ so that the blocks start to move?
My Solution
One Note Before Starting This
Whenever I say a force being "paid by a system", I mean either it gets canceled by another force on the system or it gets exerted on the system. Also, the source of the force mentioned above in the main title of the question refers to the main force $\overleftarrow{F}$ acting on the system.
Now the solution
The most shocking thing I learned while solving this question was that the friction between block A and B will be paid two times all differently by the same force $\overleftarrow{F}$ acting in disguise on the two blocks. This was shocking as I first thought that the friction will be the usual force paid by the two blocks in some ration and not two times. But what I came to figure out was that it will be paid by each block separately and the amount of force of friction paid will be the same for two blocks and not in some ratio and those two will actually be the action-reaction pairs. One way we can thing about this is that the first block let's say starts moving in the left direction then it will exert a force of friction (which would be the reaction pair of the friction on block A) on the second block. Now to overcome this force the second block will pay the same amount of force in opposite direction. Hence the two pairs of force which are action-reaction pairs will be paid differently by each block. Now, these two pairs will be equal to $\mu N$ where $N$ is the Normal force by Block A on B. And $N = mg$. Hence
Friction b/w Block A and B $ = \mu N \\ \rightarrow  \mu mg$
Friction b/w Block B and ground $= 2\mu N \\ \rightarrow 2\mu(3m)g \\  \rightarrow 6\mu mg$
For Block A
Now on block A, there is a force $\overleftarrow{F}$ and a friction $\mu mg$ which would be paid completely by this block so the tension $T$ in the string would be
$$ T = F - \mu mg $$
Let's say this equation one.
For Block B
Now there is $2T$ on block B and friction from the ground and surface above and remember that this block will pay its reaction pair separately. Hence
$$ 6\mu mg + \mu mg = 2T $$
Let's say this equation two.
Analyzing equation 1 and 2 we get
$$7\mu mg = 2F - 2\mu mg \\ F = \frac{9\mu mg}{2}$$
Final Question
Now here I want to ask that was this way of thinking correct or was it just a coincidence that I reached this solution? And thank you for spending your precious time on this question to help me out. I appreciate your efforts.
Image with all forces labelled

 A: 
I am asking that both the pairs of friction force which are action-reaction pairs on body A and B are opposing the force F and is that actually what is happening?

Friction opposes relative motion$^*$ between surfaces, not other forces (at least not explicitly).
Pulling on block A to the left will tend to cause block A to slide to the left across block B. Therefore, static friction will act to the right on block A. By Newton's third law, this static friction force must also act on block B to the right, but you can look at it for the same reason. If block A would slide to the left relative to block B, then block B would be sliding to the right relative to block A. Hence, friction opposes this sliding.
The same is true between block B and the ground. Block A moving to the left would cause block B to move to the right, thus sliding to the right relative to the ground. Static friction between block B and the ground opposes this and acts to the left.
Of course force $F$ sets all of this up, but the friction forces themselves are not acting because of force $F$ directly. They are just opposing relative motion. An easy way this "force opposition" line of thinking can confuse you is to say, "Well why isn't friction on block A opposing the tension force? How can friction know which force to oppose?" And the answer to this is that friction isn't opposing forces, its opposing relative motion.

$^*$Of course with static friction it is impending relative motion rather than actual relative motion, but for sake of brevity I will say relative motion to refer to that case as well.
A: 
Now here I want to ask that was this way of thinking correct or was it
just a coincidence that I reached this solution?

Your thinking is seems OK and solution correct, but it could have reached more quickly with applicable free body diagrams and realizing that a constraint is that both blocks must move at the same time.  I will respond to your response to  @BioPhysicist asking you what your really asking, that is:

I am asking that both the pairs of friction force which are
action-reaction pairs on body A and B are opposing the force  and is
that actually what is happening?

What is really happening here is that linking the movement of both masses with the pulley system gives you the constraint that in order for either mass to move, both must necessarily move. With that knowledge you can start by treating both blocks as a combined system.

*

*Draw a free body diagram for the combined masses as a system showing all external forces acting on the combined masses. Those forces would be $F$ and $f_{2}=(2u)(3mg)$ acting to the left and $3T$ acting to the right. That gives you one equation and two unknowns $F$ and $T$. Note that for the combined system, $f_1$ is an internal force that is not included.


*Draw a free body diagram of block A alone. For it you have $F$ acting to the left and $T+f_{2}$ where $f_{2}=umg$ acting to the right. Solve that for $T$ and plug into the first equation gets you the answer.
Hope this helps.
A: I suppose that when the system was assembled, some small tension $T$ was applied on the string to attach it to the wall, so that it can be straight. That is before $F$ comes to play.
When some $F$ is applied, while $F - T < \mu mg$ nothing moves. And the tension in the string doesn't change because the string is not affected by $F$.
As soon as $F - T = \mu mg$ block A can move. Let's say it moves a small displacement $\Delta L$.
Now, the tension in the string is $T' = T + E\frac{\Delta L} {L}$, where E is the modulus of elasticity and $L$ the total length.
But the meaning of a string is being rigid, so we can translate this to an infinite $E$.
So, even for a very small displacement, $T'$ in the string becames big enough to move block B, what means: $2T' = (2\mu)3mg + \mu mg => T' = \frac{7}{2}\mu mg$.
Back to block A, when it starts to move $T -> T'$, so: $F - T' = \mu mg => F = \frac{9}{2}\mu mg$
