A method of storing data records in redundancy groups in a data storage system which is connected to a host processor via a processor interface and having a plurality of data storage elements for storing data thereon, a number of said data storage elements being configured into at least two redundancy groups, each redundancy group including n+m of said plurality of data storage elements, where n and m are both positive integers with n greater than 1 and m greater than 0, wherein each redundancy group is divided into a plurality of logical cylinders for storing data, and further including a cache memory connected to said processor interface and said redundancy groups, said method comprising the steps of:
- maintaining a plurality of said logical cylinders concurrently open to write data from said cache memory to a selected one of the plurality of concurrently open logical cylinders;
- selecting, in response to receipt of data records from said host processor, independent of said host processor, available memory space in one of said plurality of concurrently open logical cylinders to store the received data records;
- and
- writing the received data records and data redundancy information for said received data records in the selected available memory space.