Topic 19 – The operating system zoo Flashcards
disadvantages include:
- This is a slow method for communicating and is not suited to all types of processing especially if data must be repeatedly communicated to complete the task
- Total processing power can fall and drop as users join and leave the task or computer processes start and end
name 2 disadvantages of:
Distributed systems
In this setup there will be two or more independent computers that have there own OS and hardware such as processor memory and storage
these will have middleware installed on them known as a node
by communicating with the nodes you can take advantage of all of the computers within the clusters to achieve a task and control the cluster as if it were a single system
describe a
cluster
this is mainly found on personal computers
it is implemented as multiple cores inside a single processor, each core will have its own cache and the cores can communicate with each other using a memory bus built inside the processor
describe
multicore processing
describe the structure of a
monolithic kernel vs microkernel
describe the
Graphics processing unit (GPU)
this is a dedicated processor that is used to render images.
It carries out its work in a parralell fashion as each frame can be calculated independently.
When a frame is being rendered by this it is calculating for example the colour and brightness of every pixel.
these have become useful in other areas aswell for applications such as bitcoin mining and neural networks.
advanatges include:
- Processing power from idle computers can be harnessed via the internet using volunteers or by paying participants
name an advantage of
Distributed systems
typically found on server hardware
this features processors that are independent of each other, again each one will have access to its own cache,
within this sytem the operating system must handle carefully how the execution process is carried out since all processors will be competing for the system resources
describe
multiprocessor processing
give a definition for the term
Distributed systems
a parallel processing method used to increase processing power by utilising the spare processing power of network connected computers
describe a
hard Real Time System (RTS)
this is a system that must compute an answer within a given time period. Failing to do so means that the system has failed
within this system an answer for a computation should be given before a time period however not meeting this limit does not necessarily mean the system has failed as long as it can still serve its purpose
describe a
soft Real Time System (RTS)
these strive for:
- Robustness - these types of systems should be able to run for years without any problems. It is inherent that updating these systems may not be practical and so they must be built to last
- Efficiency - these systems are typically used on low spec hardware in order to keep costs low so the OS and software must be lightweight and must be able use resources as efficiently as possible
what are two
key criterias that embedded systems strive for
describe a
cluster
In this setup there will be two or more independent computers that have there own OS and hardware such as processor memory and storage
these will have middleware installed on them known as a node
by communicating with the nodes you can take advantage of all of the computers within the clusters to achieve a task and control the cluster as if it were a single system
points for this include
- structure - All OS modules exist within the kernel and operate in the processors kernel mode
- code in kernel - A kernel can contain millions of lines of code
- communication between modules - OS modules can interact directly with each other
- robustness - Failure in one module or a rouge device driver can cause havoc on the system and ultimately crash it
- performance - This design has great performance since all modules run in kernel mode context switching does not occur
using the points below give 5 points about a
monolithic kernel
- structure
- code in kernel
- communication between modules
- robustness
- performance
describe
multicore processing
this is mainly found on personal computers
it is implemented as multiple cores inside a single processor, each core will have its own cache and the cores can communicate with each other using a memory bus built inside the processor
describe a
soft Real Time System (RTS)
within this system an answer for a computation should be given before a time period however not meeting this limit does not necessarily mean the system has failed as long as it can still serve its purpose
a computer system that is used for time critical tasks. The ability to perform and complete tasks within a timed deadline defines this type of system
define the term
Real time systems (RTS)