Section 6: AWS EC2 Networking, Placement Groups, ENI & Hibernation

An EC2 instance needs to communicate with a database inside the same VPC. Which IP should normally be used?

The private IP, because communication stays inside the AWS network.

Why should AWS resources usually use private IP addresses when communicating inside a VPC?

It avoids unnecessary internet traffic and keeps communication private.

A user needs to access an EC2-hosted website from the internet. Which address can they use?

The public IPv4 address or public DNS name.

What can happen to a normal public IPv4 address after an EC2 instance is stopped and started?

It may change.

A company needs an EC2 server to keep the same public IP after stopping and starting. What should it use?

An Elastic IP.

What is the main difference between a public IP and an Elastic IP?

A public IP may change, while an Elastic IP remains fixed until it is released.

Can an Elastic IP be moved from one EC2 instance to another?

Yes, it can be reassociated with another resource.

Why should you avoid creating unnecessary Elastic IP addresses?

Public IPv4 addresses can create charges and should only be used when needed.

Which IP address is normally used only inside an AWS VPC?

A private IP address.

Which IP address allows direct communication from the internet?

A public IP address.

What is an EC2 placement group?

A setting that controls how multiple EC2 instances are physically placed on AWS hardware.

A scientific application needs extremely fast communication between EC2 instances. Which placement group should it use?

A cluster placement group.

Why does a cluster placement group provide high network performance?

It places instances physically close together within one Availability Zone.

What is the main disadvantage of a cluster placement group?

Instances are close together, so a failure may affect several of them.

A company has a few critical EC2 instances that should not share the same hardware. Which placement group should it use?

A spread placement group.

What is the goal of a spread placement group?

To reduce the chance that one hardware failure affects multiple critical instances.

Which placement group is best for a small number of important instances needing hardware separation?

Spread placement group.

A large Kafka, Hadoop, or Cassandra system needs groups of instances separated across hardware. Which placement group fits best?

A partition placement group.

How does a partition placement group organize instances?

It divides instances into separate partitions that do not share the same underlying hardware.

Which placement group is designed for large distributed applications?

Partition placement group.

Which placement group should you remember as “speed”?

Cluster.

Which placement group should you remember as “maximum separation”?

Spread.

Which placement group should you remember as “large distributed system”?

Partition.

Can a cluster placement group span multiple Availability Zones?

No, it is designed for instances placed close together in one Availability Zone.

Why is a spread placement group not normally used for hundreds of instances?

It is designed for a limited number of critical instances that need strong hardware separation.

What is an Elastic Network Interface, or ENI?

A virtual network card used by an EC2 instance.

What network information can an ENI contain?

Private IP addresses, security groups, a MAC address, and public or Elastic IP associations.

Does an EC2 instance receive an ENI when it launches?

Yes, it receives a primary ENI.

What is the simple way to remember an ENI?

ENI equals EC2’s virtual network card.

Can an EC2 instance have more than one ENI?

Yes, if the instance type supports it.

Why might an EC2 instance use multiple ENIs?

To connect to different network groups or separate different types of traffic.

A network interface must be moved from a failed instance to another instance. What AWS feature can help?

A secondary ENI can be detached and attached to another instance in the same Availability Zone.

What happens to the network settings when a secondary ENI is moved?

Its private IP addresses, security groups, and related network configuration move with it.

Can an ENI be moved freely between different Availability Zones?

No, an ENI belongs to a specific Availability Zone.

What is the difference between an EC2 instance and an ENI?

EC2 provides compute, while the ENI provides its network connection.

Which EC2 component holds security-group associations and private IP addresses?

The ENI.

A server needs separate management traffic and application traffic. What could be added?

A second ENI, if supported.

What does stopping an EC2 instance do to its RAM contents?

The RAM contents are lost.

What does hibernating an EC2 instance do to its RAM contents?

It saves the RAM contents to the encrypted EBS root volume.

What is the main purpose of EC2 hibernation?

To let an instance continue from its previous memory state after it starts again.

An application takes a long time to initialize and should continue where it stopped. Which option may help?

EC2 hibernation.

What is the simple difference between stopping and hibernating?

Stop loses RAM; hibernate saves RAM.

What is the difference between hibernating and terminating?

Hibernating preserves the instance and RAM state, while terminating permanently deletes the instance.

Where does AWS save RAM when an instance is hibernated?

On the encrypted EBS root volume.

Why must the EBS root volume have enough free space for hibernation?

It must store the instance’s RAM contents.

When must hibernation normally be enabled?

When the EC2 instance is launched.

Can every EC2 instance type and operating system use hibernation?

No, the instance type, AMI, and configuration must support it.

Do you pay for EC2 compute while the instance is hibernated?

No compute charge applies while stopped, but storage charges continue.

What costs can continue while an instance is hibernated?

EBS storage and other attached resources may still create charges.

Which action should be used when you want to save open applications and memory?

Hibernate.

Which action should be used when you only need to turn the instance off and do not need the RAM contents?

Stop.

Which action permanently removes the instance?

Terminate.

A company needs a fixed internet-facing IP for an EC2 server. What should it choose?

Elastic IP.

Two EC2 instances need the lowest possible latency between them. What should they use?

A cluster placement group.

Three critical EC2 instances must be kept on separate hardware. What should they use?

A spread placement group.

A large distributed database needs hardware failure isolation by groups. What should it use?

A partition placement group.

What acts like the virtual network card for an EC2 instance?

An ENI.

An application must restart quickly with its previous RAM state. What EC2 feature should be used?

Hibernation.

An EC2 instance is stopped and started, and its public IP changes. How can this be prevented?

Associate an Elastic IP.

An instance needs a secondary network connection that can later move to another instance. What should be used?

A secondary ENI.

Which placement option prioritizes network speed over hardware separation?

Cluster placement group.

Which placement option prioritizes hardware separation for a few critical servers?

Spread placement group.

Which placement option separates large distributed workloads into hardware groups?

Partition placement group.