Editing Cisco IOS (section)

== Architecture ==
Cisco IOS has a [[Monolithic kernel|monolithic]] architecture, owing to the limited hardware resources of routers and switches in the 1980s. This means that all processes have direct hardware access to conserve CPU processing time. There is no memory protection between processes and IOS has a [[Run to completion scheduling|run to completion scheduler]], which means that the kernel does not pre-empt a running [[Process (computing)|process]]. Instead the process must make a kernel call before other processes get a chance to run. IOS considers each process a single [[Thread (computing)|thread]] and assigns it a priority value, so that high priority processes are executed on the CPU before queued low priority processes, but high priority processes cannot interrupt running low priority processes.<ref name="auto"/>

The Cisco IOS monolithic kernel does not implement [[memory protection]] for the data of different processes. The entire physical memory is mapped into one virtual address space. The Cisco IOS kernel does not perform any memory [[paging]] or swapping. Therefore the addressable memory is limited to the physical memory of the network device on which the operating system is installed. IOS does however support aliasing of duplicated virtual memory contents to the same physical memory. This architecture was implemented by Cisco in order to ensure system performance and minimize the operational overheads of the operating system.<ref name="auto"/>

The disadvantage of the IOS architecture is that it increases the complexity of the operating system, data corruption is possible as one process can write over the data of another, and one process can destabilize the entire operating system or even cause a software-forced [[Crash (computing)|crash]]. In the event of an IOS crash, the operating system automatically reboots and reloads the saved configuration.<ref name="auto"/>