Sequential-access device model Data storage characteristics Logical elements within a partition
SCSI-2

10.1.3 Partitions within a volume

Another attribute of a volume is called a partition. Partitions consist of one or more non-overlapped mini-volumes, each with its own beginning and ending points, contained within single physical volume. Each partition (x) within a volume has a defined beginning-of-partition (BOP x), an early-warning position (EW x), and an end-of-partition (EOP x).

All volumes have a minimum of one partition called partition 0, the default data partition. For devices which support only one partition, the beginning-of-partition zero (BOP 0) may be equivalent to the beginning-of- medium and the end-of-partition zero (EOP 0) may be equivalent to the end- of-medium.

When a volume is mounted, it is logically positioned to beginning of the default data partition (BOP 0). When a REWIND command is received in any partition (x), the device positions to the beginning-of-partition (BOP x).

Partitions on a volume do not need to be recorded in any defined order, nor do all partition numbers in a sequence need to be present on a volume. It is sufficient for a device to be able to locate a partition, given its code value, or determine that it does or does not exist on the volume. For interchange, information about which partitions are present on a volume may be stored on the volume in a device-defined area (possibly unavailable to the initiator) or the information may be an intrinsic attribute of the device implementation.

Figure 24 shows a possible partition implementation for a four-track serpentine recording device, assuming that each track group defines a partition.

Figure 24 - Partitioning example - one partition per track group

Another possible partition implementation for this four-track serpentine recording device is shown in figure 25, using two track groups to define each partition.

Figure 25 - Partitioning example - one partition per two track groups

The previous examples show the beginning and ending points for a partition aligned with physical bounds of the medium. This is not a mandatory requirement for partitioning; it is sufficient for a device to be able to locate to and stay in any partition bounded by a BOP x and EOP x. In this case, a recorded mark or some other device-recognizable attribute could be used to delineate the partitions. Figure 26 shows a possible two- partition implementation for a device with only one track group.

Figure 26 - Partitioning example - two partitions per track group

Three methods are defined in the MODE SENSE and MODE SELECT commands for managing partitions; each method is progressively more difficult to implement and manage:

  1. device-defined fixed locations;
  2. device-defined based on an initiator supplied number of partitions and a device specific allocation algorithm; and,
  3. definition by name and capacity by an initiator.