Proxy timeouts in Nginx - Time & Space Complexity

Time Complexity: Proxy timeouts

O(n)

Understanding Time Complexity

When nginx acts as a proxy, it waits for responses from other servers. Analyzing time complexity helps us understand how waiting times grow when many requests happen.

We want to know how the time nginx spends waiting changes as the number of requests increases.

Scenario Under Consideration

Analyze the time complexity of the following nginx proxy timeout settings.

proxy_connect_timeout 10s;
proxy_send_timeout 30s;
proxy_read_timeout 30s;

location /api/ {
    proxy_pass http://backend_server;
}
    

This snippet sets how long nginx waits to connect, send, and read from the backend server before giving up.

Identify Repeating Operations

Identify the loops, recursion, array traversals that repeat.

• Primary operation: Handling each incoming request and waiting for backend response within timeout limits.
• How many times: Once per request, repeated for every request nginx receives.

How Execution Grows With Input

As the number of requests increases, nginx handles each request independently, waiting up to the timeout for each.

Input Size (n)	Approx. Operations
10	Waits for up to 10-30 seconds per request, total waits scale roughly with 10 times.
100	Waits scale roughly with 100 times, as each request waits independently.
1000	Waits scale roughly with 1000 times, each request handled separately.

Pattern observation: Total waiting time grows linearly with the number of requests.

Final Time Complexity

Time Complexity: O(n)

This means the total waiting time grows directly in proportion to the number of requests nginx handles.

Common Mistake

[X] Wrong: "Timeouts cause nginx to wait only once, no matter how many requests come in."

[OK] Correct: Each request is handled separately, so nginx waits for each one up to the timeout, making total wait time grow with request count.

Interview Connect

Understanding how proxy timeouts affect request handling helps you explain server behavior under load, a useful skill for real-world system design and troubleshooting.

Self-Check

What if we changed proxy_read_timeout to a much smaller value? How would the time complexity change?

Practice

1. What does the proxy_connect_timeout directive control in Nginx?

easy

A. The time Nginx waits to receive the full response from the backend

B. The time Nginx waits to send the full request to the backend

C. The time Nginx waits to establish a connection with the backend server

D. The time Nginx caches the response from the backend

Solution

1. Step 1: Understand proxy_connect_timeout purpose

   This directive sets how long Nginx waits to connect to the backend server before giving up.

2. Step 2: Differentiate from other timeouts

   proxy_send_timeout controls sending request time, proxy_read_timeout controls response wait time, so they are different.

3. Final Answer:

   The time Nginx waits to establish a connection with the backend server -> Option C

4. Quick Check:

   proxy_connect_timeout = connection wait time [OK]

Hint: Connect timeout is for establishing backend connection only [OK]

Common Mistakes:

• Confusing connect timeout with read or send timeout
• Thinking it controls response wait time
• Assuming it caches responses

2. Which of the following is the correct syntax to set a 10-second proxy read timeout in Nginx?

easy

A. proxy_read_timeout 10;

B. proxy_read_timeout 10s;

C. proxy_read_timeout = 10s;

D. proxy_read_timeout: 10s;

Solution

1. Step 1: Recall Nginx timeout syntax

   Timeout values are set as numbers with optional time units, but seconds are default so '10' means 10 seconds.

2. Step 2: Check syntax correctness

   Correct syntax is proxy_read_timeout 10; without '=' or ':' and without 's' suffix.

3. Final Answer:

   proxy_read_timeout 10; -> Option A

4. Quick Check:

   Timeouts use number only, no '=' or ':' [OK]

Hint: Use number only for seconds, no '=' or ':' in directives [OK]

Common Mistakes:

• Adding '=' sign which is invalid
• Using ':' instead of space
• Adding 's' suffix which is not required

3. Given this Nginx config snippet:

location /api/ {
  proxy_pass http://backend;
  proxy_connect_timeout 5;
  proxy_send_timeout 10;
  proxy_read_timeout 15;
}

What happens if the backend server takes 20 seconds to send the full response?

medium

A. Nginx closes connection after 10 seconds waiting to send request

B. Nginx closes connection after 5 seconds trying to connect

C. Nginx waits indefinitely for the backend response

D. Nginx closes connection after 15 seconds waiting for response

Solution

1. Step 1: Identify relevant timeout for response wait

   proxy_read_timeout controls how long Nginx waits for the backend to send the full response.

2. Step 2: Compare backend response time with proxy_read_timeout

   Backend takes 20 seconds, proxy_read_timeout is 15 seconds, so Nginx closes the connection after 15 seconds.

3. Final Answer:

   Nginx closes connection after 15 seconds waiting for response -> Option D

4. Quick Check:

   Response wait = proxy_read_timeout = 15 [OK]

Hint: Response wait time controlled by proxy_read_timeout [OK]

Common Mistakes:

• Confusing send timeout with read timeout
• Assuming connect timeout applies to response wait
• Thinking Nginx waits forever

4. You set proxy_connect_timeout 10; but Nginx still times out connecting to backend after 5 seconds. What is the likely cause?

medium

A. The backend server is closing connections early

B. There is a lower timeout set by proxy_connect_timeout in a lower-level block

C. Nginx ignores proxy_connect_timeout and uses default 5 seconds

D. The proxy_connect_timeout directive is misspelled

Solution

1. Step 1: Understand directive inheritance in Nginx

   Timeout directives can be set at multiple levels; lower-level settings override higher-level ones.

2. Step 2: Identify cause of shorter timeout

   If a lower-level block has a 5-second timeout, it overrides the 10-second setting.

3. Final Answer:

   There is a lower timeout set by proxy_connect_timeout in a lower-level block -> Option B

4. Quick Check:

   Timeouts can be overridden by lower-level config [OK]

Hint: Check for overriding timeouts in child blocks [OK]

Common Mistakes:

• Assuming Nginx ignores the directive
• Blaming backend without checking config
• Misspelling directive (would cause error, not timeout)

5. You want Nginx to wait up to 20 seconds to connect to backend, 30 seconds to send the request, and 40 seconds to receive the response. Which configuration snippet correctly sets these timeouts?

hard

A. proxy_connect_timeout 20;
proxy_send_timeout 30;
proxy_read_timeout 40;

B. proxy_connect_timeout 20s;
proxy_send_timeout 30s;
proxy_read_timeout 40s;

C. proxy_connect_timeout=20;
proxy_send_timeout=30;
proxy_read_timeout=40;

D. proxy_connect_timeout: 20;
proxy_send_timeout: 30;
proxy_read_timeout: 40;

Solution

1. Step 1: Recall correct syntax for timeout directives

   Timeouts are set as directive name followed by a space and number (seconds), ending with semicolon; no '=' or ':' or 's' suffix needed.

2. Step 2: Validate each option

   proxy_connect_timeout 20; proxy_send_timeout 30; proxy_read_timeout 40; uses correct syntax; A uses ':'; B adds unnecessary 's'; C uses '='.

3. Final Answer:

   proxy_connect_timeout 20; proxy_send_timeout 30; proxy_read_timeout 40; -> Option A

4. Quick Check:

   Use number only with semicolon, no '=' or ':' [OK]

Hint: Timeouts use number only, no '=' or ':' [OK]

Common Mistakes:

• Adding 's' suffix to numbers
• Using '=' or ':' instead of space
• Missing semicolon at line end