What I understood here, is that they were redundant systems, but running the same software, with the same bug present, so both went down.

a.k.a. "flailover"