DNS64 and NAT64 IPv6 transition mechanisms are expected to play an important role in the near future to solve the problem that some of the new clients will not be able to get public IPv4 addresses and thus having only IPv6 addresses they still should be able to reach servers that have only IPv4 addresses. In our earlier experiments, the TOTD DNS64 implementation showed significantly better average performance than BIND, however TOTD was not stable, therefore now it was carefully tested to find the reason for its experienced strange behavior. Besides the detailed description of the testing method, the bug and the correction, a security vulnerability is disclosed and a patch is provided. The performance and the stability of the modified versions of TOTD are analyzed and compared to that of the original TOTD and BIND.