First, I would rewrite it a little bit. Set $\alpha=\beta_1+\beta_2$ and $\gamma=\beta_1\beta_2$. Then you have
\begin{align}
a&=b\frac{\alpha^2-\gamma}{\alpha\gamma}\\
c&=b\frac{\gamma}{\alpha}
\end{align}
and your inequality becomes
$$
|a-\alpha|+|b-a\alpha+\gamma|+|c\gamma|\leq 2
$$
which becomes
\begin{align}
\left|b\frac{\alpha^2-\gamma}{\alpha\gamma}-\alpha\right|+
\left|b-b\frac{\alpha^2-\gamma}{\gamma}+\gamma\right|+
\left|b\frac{\gamma^2}{\alpha}\right|\leq 2\\
\left|\frac{\alpha^2-\gamma}{\alpha\gamma}\right|\left|b-\frac{\alpha^2\gamma}{\alpha^2-\gamma}\right|+
\left|\frac{2\gamma-\alpha^2}{\gamma}\right|\left|b-\frac{\gamma^2}{2\gamma-\alpha^2}\right|+
\left|\frac{\gamma^2}{\alpha}\right||b-0|\leq 2\\
\end{align}
On the left you have a sum of functions of the form $m|b-n|$, which is a modified absolute value function with a zero at $n$ and slope $\pm m$. It's pretty awkward to give a general solution for the interval of possible $b$'s depending on $\alpha$ and $\gamma$, but if you have concrete values given, it should be fairly easy to determine, since you have a convex piecewise linear function on the left.
EDIT:
As we've already seen, the function on the left hand side takes on the form:
$$f(b) = \sum_i m_i |b_i-n_i|$$
It's obvious that $f'(b)$ is non-decreasing, where it's defined (please plot an example to get a feel for it), so the function is convex. Further it's piecewise linear with kinks at the $n_i$. Hence, the minimum of this function is either at exactly at one of the $n_i$ or on a whole interval $[n_i, n_{i+1}]$. Thus it's enough to check the value of $f$ at each of the $n_i$ and take the minimum thereof. If $f(n_i)$ is smaller or equal to 2 for one $n_i$, than it's a solution, otherwise there is no solution at all. In your case you have to check three values:
\begin{align}
f\left(\frac{\alpha^2\gamma}{\alpha^2-\gamma}\right)
&=
\left|\frac{2\gamma-\alpha^2}{\gamma}\right|\left|\frac{\alpha^2\gamma}{\alpha^2-\gamma}-\frac{\gamma^2}{2\gamma-\alpha^2}\right|+
\left|\frac{\gamma^2}{\alpha}\right|\left|\frac{\alpha^2\gamma}{\alpha^2-\gamma}-0\right|\\
&=
\left|\frac{\alpha^2(2\gamma-\alpha^2)}{\alpha^2-\gamma}-\gamma\right|+
\left|\frac{\alpha\gamma^3}{\alpha^2-\gamma}\right|\\
f\left(\frac{\gamma^2}{2\gamma-\alpha^2}\right)
&=
\left|\frac{\alpha^2-\gamma}{\alpha\gamma}\right|\left|\frac{\gamma^2}{2\gamma-\alpha^2}-\frac{\alpha^2\gamma}{\alpha^2-\gamma}\right|+
\left|\frac{\gamma^2}{\alpha}\right|\left|\frac{\gamma^2}{2\gamma-\alpha^2}-0\right|\\
&=
\left|\frac{(\alpha^2-\gamma)\gamma}{\alpha(2\gamma-\alpha^2)}-\alpha\right|+
\left|\frac{\gamma^4}{\alpha(2\gamma-\alpha^2)}\right|\\
f\left(0\right)
&=
\left|\frac{\alpha^2-\gamma}{\alpha\gamma}\right|\left|0-\frac{\alpha^2\gamma}{\alpha^2-\gamma}\right|+
\left|\frac{2\gamma-\alpha^2}{\gamma}\right|\left|0-\frac{\gamma^2}{2\gamma-\alpha^2}\right|\\
&=
\left|\alpha\right|+
\left|\gamma\right|\\
\end{align}
If you have some $\beta_1$ and $\beta_2$ given, it should be fairly easy to find the minimum now. If you want to have a general idea, where the inequality is satisfied I would suggest plotting $\min(f(n_1), \ldots, f(n_3))$ over the allowed range of $\beta_1$ and $\beta_2$. Solving the inequality directly for those $\beta_1$ and $\beta_2$ that have a solution for $b$ seems way too complicated to me.