In computing, rebooting is the process by which a running computer system is restarted, either intentionally or unintentionally.
Reboots can be either hard (alternatively known as cold) whereby the power to the system is physically turned off, or soft ( alternatively known as warm) where the system restarts without the need to interrupt the power or trigger a reset line.
Contents |
A hard reboot (also known as a cold reboot, cold boot or cold start) is when power to a computer is abruptly turned off and then turned back on. Since the operating system does not have the opportunity to perform any shutdown procedures, data loss or corruption may occur if transactions in disk caches are not written to the filesystem. After the computer starts again, the filesystem may be in a corrupted state, requiring an integrity check of on-disk filesystem structures to be performed. In a worst case scenario, corruption may affect files that are required for the operating system to start, thereby preventing it from booting again.
A hard reboot may be caused by power failure, by accident or deliberately as a last resort to reset an unresponsive system or critical error. It can also be used by intruders to access cryptographic keys from RAM, in which case it is called a cold boot attack.[1] The cold boot attack relies on the data remanence property of DRAM and SRAM to retrieve memory contents which remain readable in the seconds to minutes after power has been removed.[1]
However, other types of reset can also disrupt the operating system in the same manner as power loss, for example
An uninteruptible or redundant power supply may be used to prevent a critical system from shutting down and rebooting following an unexpected loss of power.
A soft reboot (also known as a warm reboot) involves restarting a computer "normally" under software control, without removing power or (directly) triggering a hardware-based reset. It usually, though not always, refers to an orderly shutdown and restart of the machine. On some machines, the BIOS may perform a shorter initialisation sequence than it would for a cold start, for example by skipping a power-on self-test.
The Control-Alt-Delete combination initiated a warm reboot on many legacy operating systems, including DOS and Windows 95. This key sequence was subsequently replaced on later operating systems, for example on Windows NT, where it has now become a secure attention sequence.
Special configurations may allow for part of the system state, like a RAM disk, to be preserved through the reboot.[2]
The Linux kernel has optional support for kexec, a system call which transfers execution to a new kernel and skips hardware or firmware reset. The entire process occurs independently of the system firmware. The kernel being executed does not have to be a Linux kernel.
Random reboot is a non-technical term referring to an unintended (and often undesired) reboot for which the cause is not immediately evident to the user. Such reboots may occur due to a multitude of software and hardware problems, such as triple faults. They are generally symptomatic of an error in ring 0 that is not trapped by an error handler in an operating system or a hardware-triggered non-maskable interrupt.
Systems may be configured to reboot automatically at a scheduled time, after a power failure, or a fatal system error or kernel panic. The method by which this is done varies depending whether the reboot can be handled in software, or must be handled at the firmware or hardware level. Operating systems in the Windows NT family (from Windows NT 3.1 through Windows 7) have an option to modify the behaviour of the error handler so that a computer immediately retarts rather than displaying a Blue Screen of Death (BSOD) error message.
Rebooting may be used by users, support staff or system administrators as a technique to workaround bugs in software, for example memory leaks or processes that hog resources to the detriment of the overall system, or to terminate malware. While this approach does not address the root cause of the issue, resetting a system back to a good, known state may allow it to be used again for some period until the issue next occurs.
On a multi boot system, a reboot is required to switch between installed operating systems.
The introduction of advanced power management support in system BIOS'es allowed operating systems greater control of hardware power management features. On subsequent ACPI BIOS'es, newer operating systems are able to manage different power states and thereby sleep and/or hibernate. While hibernation also involves turning a system off then subsequently back on again, the operating system does not re-initialise from scratch, therefore differentiating this process from rebooting.
A reboot may be simulated by software running on an operating system. For example, the Sysinternals BlueScreen utility, which is used for pranking. Malware may also simulate a reboot, and thereby deceive a computer user for some nefarious purpose.