Scribd
Upload
231 views

Distributed Systems

1. The document discusses system availability when servers can fail. If one of four servers needs to be available, availability is 99.99%. If all four servers must be available, availability is 65.61%. 2. Striving for the highest degree of distribution transparency can negatively impact performance in a way users find unacceptable. For example, ensuring full location transparency could result in unreliable network connections and poor system performance. 3. Active standby systems have two servers that each service clients independently until one fails, while passive standby only activates the backup server during a failure and does not service clients regularly.

Uploaded by

chelsea oloo
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
231 views

Distributed Systems

1. The document discusses system availability when servers can fail. If one of four servers needs to be available, availability is 99.99%. If all four servers must be available, availability is 65.61%. 2. Striving for the highest degree of distribution transparency can negatively impact performance in a way users find unacceptable. For example, ensuring full location transparency could result in unreliable network connections and poor system performance. 3. Active standby systems have two servers that each service clients independently until one fails, while passive standby only activates the backup server during a failure and does not service clients regularly.

Uploaded by

chelsea oloo
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Assignment One

1. Consider a distributed system consisting of four replicated servers. Each of the servers is
available at any instant with a probability of 90%. What is the overall system availability
If
a) The system is designed so that it can be operational if any of the four servers is
operational.
- The probability of one server being down at a particular time is 0.1
- The probability of all four servers being down at the same time will be 0.1 /4=
0.0001. So the probability of at least one being available is 1-0.0001=0.9999.

b) The system is designed such that all four servers have to be available for the entire
system to be available.
- The probability of all four servers being available at the same time is 0.9/4=
0.6561.
2. Discuss why it is not always good idea to aim at implementing the highest degree of
distribution transparency possible? Provide an example.
- Aiming at the highest degree of transparency can cause a great amount of
performance loss that the end users will not be willing to accept. For example,
when full location transparency is given, the networks connecting the distributed
systems may end up being unreliable and have poor connection. This may result
in users losing trust in the operating system and also lead to poor performance in
the system due to less interaction by system users.

3. Explain two differences between active standby and passive standby systems.
- In active standby, two servers are configured as companion servers, each with its
own independent workload and runs as an individual server until one fails over. In
passive standby, failback is a planned fail-over or relocation of the server and its
resources to the primary node. Failback is not required.
- Active standby actively services clients on both virtual IP addresses while
arranging for them to migrate to one zone. Passive standby produces no functional
output but can be switched into another mode with a remotely controlled unit or
an internal signal.
Task One 1. Consider a disseminated framework comprising of
four duplicated servers. Each of the servers is accessible at any moment with
a likelihood of 90%. What is the generally framework accessibility If a)
The framework is outlined so that it can be operational in the event that any
of the four servers is operational. - The likelihood of one server being down at
a specific time is 0.1 - The likelihood of all four servers being down at the
same time will be 0.1 /4= 0.0001. So the likelihood of at slightest one
being accessible is 1-0.0001=0.9999. b) The framework is planned such that
all four servers got to be accessible for the whole framework to be available. -
The likelihood of all four servers being accessible at the same time is 0.9/4=
0.6561. 2. Discuss why
it isn't continuously great thought to point at executing the most
noteworthy degree of dissemination straightforwardness conceivable? Give an
example. - Aiming at the most noteworthy degree of straightforwardness can
cause a awesome sum of execution misfortune that the conclusion clients will

You might also like