SpamAssassin and MagicSpam are anti-spam solutions in Plesk. Are they compatible with each other?
MagicSpam works very well in combination with SpamAssassin because they work at different levels: MagicSpam checks mails in RBLs while SpamAssassin checks only the content of the email.
MagicSpam reduces load on a server by blocking messages directly in the SMTP layer based on IP reputation and SMTP best practices: it operates based on the characteristics of the sending server that determine if said server is displaying signs of non compliance or other behavior that would indicate a spam source (including a listing on a blocklist). So MagicSpam works before mail reached the delivery system, knocking the worst out before it uses up bandwidth, I/O or overhead.
This lowers a load because it will block messages instead of passing them on to be processed further by software such as SpamAssassin, it also subsequently reduces network usage and disk usage for the same reasons.
SpamAssassin, on the other hand, is a second level of anti-spam protection with content-based scanning. Using it is an effective way to handle messages which get past MagicSpam. For email coming from large webmail providers or from determined spammers with properly configured mail servers often the only remaining option is to filter by content using a tool such as Spam Assasin.