Data replication based on compression ratio history

Data is replicated from a first data processing system to a second data processing system. The first system compresses data that is replicated to the second system and maintains a history of the compression ratio of the replicated data. When further data needs to be replicated, the first system consults the history to determine whether to compress the data. If the average of the compression ratios in the history indicates greater compression than a threshold, then the data is compressed. The replicated data may be checked on a file by file basis, with the compression ratio history for each file being stored in an extended file attribute. An integer value between 1 and 100 may be used for each history entry. An additional time parameter may be used to force the data to be compressed at predetermined intervals.