Consider the following Semi-Definite Feasibility problem
\begin{align}
\max_{\mathbf{Z}}~0 \\\
\mathrm{trace}(\mathbf{Z})\leq \rho \\\
\mathrm{trace}(\mathbf{S}_1\mathbf{Z}) \geq \alpha \\\
\mathrm{trace}(\mathbf{S}_2\mathbf{Z}) \geq \alpha \\\
\mathbf{Z} \geq 0
\end{align}
where $\mathbf{S}_i$ are hermitian matrices of appropriate dimensions. $\rho$,$\alpha$ are known positive quantities. This comes from a quadratic feasibility problem as its semi definite relaxation, so that $\mathbf{Z}=\mathbf{z}\mathbf{z}^H$ in the orginal problem. CVX never returned a rank-one solution for this. Does it mean, semi-definite relaxation is not optimal in this case, is there a theoretical way of arguing this?

 (Note: I decide the matrix to be rank one, if it has only one singular value above a particular threshold which is set very low as $10^{-6}$).