Abstract: A Read-Modify-Write operation in a RAID system is done by separating the writing of new data and the writing of new parity, so that they are not done in parallel. This allows recovery to be performed at all stages, without requiring the excessive use of non-volatile cache, and minimizing the amount of time a span has to be locked. The invention accomplishes this by allowing one of the recoveries to be a recovery to the old data, before the write, with a signal to the host that the write operation failed in such a recovery situation, requiring the host to resend the data. This speeds up the entire operation while minimizing the use of resources while only requiring that in the rare instances of a failure during a particular part of the Read-Modify-Write, the host needs to resend the data.

Abstract: A method for converting between logical and physical memory space which adapts to different RAID types and configurations in a modular form. In particular, a module containing a standard set of conversion algorithms is used for all conversions. The standard set of conversion or translation algorithms operate on a pseudo representation of a RAID array that has all redundant components removed. For each RAID type and configuration, the standard algorithms can be used unchanged if a method is provided to convert the real RAID array representation into the pseudo version and back again. This simplifies both the programming and the debugging. This also makes the software more modular and more easily upgradeable.

Abstract: An improved method and apparatus for providing I/O operations. In particular, a particular I/O operation is programmed as a net in a matrix format. Upon initialization of the I/O subsystem, an intermediate representation of the net in matrix format is created. Subsequently, upon a running of the net operation, a complete version of the net matrix is created.