When a system crashes, the crash dump files can be preserved in memory until after the OS reboots. When the OS is rebooting, the crash dump files are extracted from memory to the filesystem that is defined in the dump configuration. After these files are written, the OS automatically reboots to the normal multiuser configuration. This process is referred to as a deferred dump. Deferred dumps enable the OS to return to a running state more quickly after a kernel panic.
In addition, deferred dumps specifically benefit systems such as the M8 and M7 systems that ship without a local disk. With deferred dumps, you can instead use network-connected boot devices, such as iSCSI targets, which normally cannot store crash dump data on a disk during a panic. Deferred dump enables the crash dump data to be stored in memory instead of on a disk while the OS is crashing and rebooting. After the OS reboots, the crash dump data is extracted from memory and put into dump files on a network-connected boot device. Thus, deferred dump enables systems with network-connected boot devices to save a crash dump in the event of a panic.