Instance types and families in AWS - Time & Space Complexity

Time Complexity: Instance types and families

O(n)

Understanding Time Complexity

When choosing instance types and families, it's important to understand how the number of instances affects the time to launch and manage them.

We want to know how the time to provision and handle instances grows as we increase their count.

Scenario Under Consideration

Analyze the time complexity of launching multiple EC2 instances from a chosen instance family.

for i in range(n):
    ec2.run_instances(
        ImageId='ami-12345678',
        InstanceType='t3.micro',
        MinCount=1,
        MaxCount=1
    )
    

This sequence launches n EC2 instances one by one using the same instance type from a family.

Identify Repeating Operations

Identify the API calls, resource provisioning, data transfers that repeat.

• Primary operation: The run_instances API call to launch one instance.
• How many times: This call repeats n times, once per instance.

How Execution Grows With Input

Each instance requires a separate API call, so the total calls grow directly with the number of instances.

Input Size (n)	Approx. Api Calls/Operations
10	10
100	100
1000	1000

Pattern observation: The number of API calls increases one-to-one with the number of instances.

Final Time Complexity

Time Complexity: O(n)

This means the time to launch instances grows linearly as you add more instances.

Common Mistake

[X] Wrong: "Launching multiple instances at once takes the same time as launching one instance."

[OK] Correct: Each instance requires its own setup and API call, so total time grows with the number of instances.

Interview Connect

Understanding how instance provisioning scales helps you design systems that handle growth smoothly and predict resource needs.

Self-Check

"What if we used a single API call to launch multiple instances at once? How would the time complexity change?"

Practice

1. Which AWS instance family is best suited for general purpose workloads like web servers and small databases?

easy

A. C family (e.g., c5, c6g)

B. P family (e.g., p3, p4)

C. R family (e.g., r5, r6g)

D. T family (e.g., t3, t4g)

Solution

1. Step 1: Understand instance family purposes

   The T family is designed for general purpose workloads with balanced CPU and memory.

2. Step 2: Match workload type to instance family

   Web servers and small databases typically need balanced resources, fitting T family well.

3. Final Answer:

   T family (e.g., t3, t4g) -> Option D

4. Quick Check:

   General purpose = T family [OK]

Hint: General purpose = T family instances [OK]

Common Mistakes:

• Confusing compute-optimized C family for general purpose
• Choosing memory-optimized R family for small databases
• Selecting GPU P family for simple web servers

2. Which of the following is the correct syntax to specify a compute-optimized instance type in AWS?

easy

A. t3.large

B. c5.large

C. r5.large

D. m5.large

Solution

1. Step 1: Identify compute-optimized prefix

   Compute-optimized instances start with 'c' (e.g., c5, c6g).

2. Step 2: Check instance type format

   Correct format is family + size, like 'c5.large'.

3. Final Answer:

   c5.large -> Option B

4. Quick Check:

   Compute-optimized = c5.large [OK]

Hint: Compute-optimized instances start with 'c' [OK]

Common Mistakes:

• Choosing m5.large which is general purpose
• Selecting r5.large which is memory optimized
• Using t3.large which is burstable general purpose

3. Given the following instance types, which one provides the highest memory capacity per vCPU?

A) m5.large
B) c5.large
C) r5.large
D) t3.large

medium

A. r5.large

B. c5.large

C. m5.large

D. t3.large

Solution

1. Step 1: Understand instance family memory focus

   R family is memory optimized, offering more memory per vCPU than others.

2. Step 2: Compare memory per vCPU for given types

   r5.large has 16 GiB RAM and 2 vCPUs (8 GiB/vCPU), higher than m5.large (8 GiB/2 vCPU = 4 GiB/vCPU), c5.large (4 GiB/2 vCPU = 2 GiB/vCPU), t3.large (8 GiB/2 vCPU = 4 GiB/vCPU).

3. Final Answer:

   r5.large -> Option A

4. Quick Check:

   Memory optimized = r5.large [OK]

Hint: Memory optimized = R family (r5, r6g) [OK]

Common Mistakes:

• Assuming m5.large has more memory per vCPU
• Confusing compute-optimized c5.large as memory heavy
• Thinking burstable t3.large has highest memory per vCPU

4. You launched an instance with type c5.large but your application needs more memory. Which instance type should you choose to fix this issue without changing the CPU count?

medium

A. r5.large

B. c5.xlarge

C. t3.large

D. m5.large

Solution

1. Step 1: Identify current instance specs

   c5.large has 2 vCPUs and 4 GiB memory.

2. Step 2: Find instance with same vCPU but more memory

   r5.large has 2 vCPUs and 16 GiB memory, increasing memory without changing CPU count.

3. Final Answer:

   r5.large -> Option A

4. Quick Check:

   More memory, same CPU = r5.large [OK]

Hint: Keep vCPU same, pick memory-optimized with same size [OK]

Common Mistakes:

• Choosing c5.xlarge which doubles CPU and memory
• Picking t3.large which is burstable, not memory optimized
• Selecting m5.large which has less memory (8 GiB) than r5.large

5. You need to run a machine learning training job that requires high GPU performance and large memory. Which instance family and size combination is the best fit?

hard

A. r5.12xlarge

B. c5.4xlarge

C. p4d.24xlarge

D. t3.2xlarge

Solution

1. Step 1: Identify instance families for GPU workloads

   P family instances are designed for GPU-intensive tasks like machine learning.

2. Step 2: Choose size with high GPU and memory

   p4d.24xlarge offers multiple GPUs and large memory, ideal for training jobs.

3. Final Answer:

   p4d.24xlarge -> Option C

4. Quick Check:

   GPU + large memory = p4d.24xlarge [OK]

Hint: GPU needs = P family, largest size for training [OK]

Common Mistakes:

• Picking c5.4xlarge which lacks GPUs
• Choosing r5.12xlarge which is memory optimized but no GPU
• Selecting t3.2xlarge which is burstable general purpose