What characteristic defines a Rainbow attack?

The defining characteristic of a Rainbow attack is that it relies on precomputed tables of hashes. In this context, a Rainbow table is essentially a large set of precomputed hash values, allowing an attacker to reverse-engineer passwords from their corresponding hash values without needing to compute the hash during the attack phase. This is achieved by using a time-memory trade-off technique, which enables attackers to use massive tables of hashes to quickly look up and find the original plaintext passwords rather than recalculating hashes from scratch for each guess. The use of precomputed tables significantly speeds up the process of cracking hashed passwords compared to brute-force methods.

In contrast, a Rainbow attack does not typically involve real-time activities without preparation, as it necessitates extensive groundwork in generating and storing these tables. Additionally, it is not a form of social engineering because it does not rely on manipulating individuals to acquire sensitive information. Lastly, while some attacks may be performed offline (such as those that do not involve interaction with the server), Rainbow attacks can also be conducted online, depending on the situation and the availability of the hashing function's output. Thus, the reliance on precomputed tables to facilitate the reduction of computation time is what distinctly characterizes a Rainbow attack.