The high-performance Angara interconnect supports up to 4D torus topology. The Angara interconnect system software requires that network traffic for user jobs must not overlap. Additionally, the allocated node set for a user job must be connected—that is, connected in terms of Angara routing. When allocating nodes in the Angara interconnect, transit nodes may be required to maintain the connectivity of a node set. Our approach to node allocation is routing-aware; for each node set (a solution to the node allocation problem), we execute a routing table generation algorithm. Thus, our approach is fine-grained, with the goal of minimizing the number of transit nodes. For these reasons, it is important to evaluate the utilization of high-performance computing (HPC) systems based on the Angara interconnect. In this work, we evaluate Angara routing table generation algorithms and node allocation algorithms.