Junior Load Balancing & Networking Interview Questions (50 MCQs)

Question 1

What is the main purpose of a load balancer?

Accepted Answer

Distributing incoming traffic across multiple backend servers

Answer

Storing responses to reduce database queries

Answer

Compressing data before sending it to the client

Answer

Translating domain names to IP addresses for clients

Question 2

How does the round robin algorithm distribute requests?

Accepted Answer

In order, taking turns through each server in the list

Answer

To the server with the fewest active connections

Answer

Based on a hash of the client's IP address

Answer

To the server that has responded the fastest

Question 3

When is weighted round robin a better fit than plain round robin?

Accepted Answer

When servers have different capacities or hardware power

Answer

When all servers have exactly identical specifications

Answer

When each user must always reach the same server

Answer

When there is only a single backend server

Question 4

Where does the least connections algorithm send a new request?

Accepted Answer

The server with the fewest active connections at that moment

Answer

The server that has received the fewest requests all day

Answer

The server with the lowest disk usage

Answer

The server that is geographically closest

Question 5

What is the main benefit of the IP hash algorithm?

Accepted Answer

A given client usually always reaches the same server

Answer

Load is always spread perfectly evenly across servers

Answer

Slow servers are automatically taken out of rotation

Answer

It completely removes the need for health checks

Question 6

What characterizes the random load balancing algorithm?

Accepted Answer

With many requests, it gives a fairly even distribution on average

Answer

It guarantees every server gets exactly the same request count

Answer

It always picks the least loaded server

Answer

It keeps each client pinned to one fixed server

Question 7

On what basis does the least response time algorithm pick a server?

Accepted Answer

The server's recent response speed, often combined with connection count

Answer

A fixed order defined in the configuration file

Answer

The number of CPU cores on each server

Answer

A hash of the request URL path

Question 8

What is a health check used for in a load balancer?

Accepted Answer

Detecting failed servers so traffic is not sent to them

Answer

Measuring the internet speed of end users

Answer

Verifying code correctness before deployment

Answer

Encrypting traffic between client and server

Question 9

What does an endpoint like `/health` typically return?

Accepted Answer

A `200` when the service is ready, and an error code when it is not

Answer

The full list of active users in the system

Answer

Always a `500` so the load balancer stays alert

Answer

The server's recent log file for debugging

Question 10

What is the difference between active and passive health checks?

Accepted Answer

Active sends probe requests; passive watches errors in real traffic

Answer

Active runs only at night; passive runs only during the day

Answer

Active runs on the database; passive runs on the client

Answer

Active restarts the server; passive only writes logs

Question 11

What does a Layer 4 (L4) load balancer base its decisions on?

Accepted Answer

IP addresses and ports at the TCP/UDP level

Answer

The URL path and HTTP headers

Answer

The contents of browser cookies

Answer

The user's type and role in the application

Question 12

Which capability is only possible with a Layer 7 (L7) load balancer?

Accepted Answer

Sending `/api` to one server group and `/images` to another

Answer

Spreading TCP connections across several servers

Answer

Removing a server that does not respond on its port

Answer

Distributing UDP traffic across several instances

Question 13

Why is a Layer 4 load balancer usually faster and lighter than a Layer 7 one?

Accepted Answer

Because it does not parse HTTP and just forwards packets

Answer

Because it always runs on more expensive hardware

Answer

Because it caches the response to every request

Answer

Because it connects to only one backend server

Question 14

What does sticky session (session affinity) mean?

Accepted Answer

A user's requests are always sent to the same server

Answer

The user's session is stored in the browser forever

Answer

The user is never logged out after logging in

Answer

All servers share one session in memory

Question 15

What is a major downside of sticky sessions?

Accepted Answer

If that server goes down, the user's session is lost

Answer

It causes HTTPS to stop working

Answer

It makes every request dramatically slower

Answer

It does not allow more than two servers

Question 16

What is the most common alternative to sticky sessions for storing session state?

Accepted Answer

Keeping sessions in a shared store such as Redis

Answer

Keeping sessions in each server's memory separately

Answer

Disabling login entirely in the application

Answer

Manually copying sessions between servers every night

Question 17

Why are stateless servers preferred behind a load balancer?

Accepted Answer

Any request can go to any server without losing data

Answer

Stateless servers never need any database

Answer

Stateless servers never crash

Answer

Being stateless removes the need for a load balancer

Question 18

What role does a load balancer play in horizontal scaling?

Accepted Answer

It provides a single entry point in front of many identical servers

Answer

It automatically increases a single server's CPU and RAM

Answer

It splits the database into several shards

Answer

It compiles application code for higher speed

Question 19

What does TLS termination at the load balancer mean?

Accepted Answer

The load balancer decrypts HTTPS traffic and forwards it to backends

Answer

The load balancer blocks HTTPS connections entirely

Answer

A certificate is generated per user on the load balancer

Answer

The load balancer expires the certificate after a while

Question 20

What is a reverse proxy?

Accepted Answer

A server in front of backends that receives client requests on their behalf

Answer

A server in front of clients that connects to the internet on their behalf

Answer

A tool for rolling back database changes

Answer

A server that returns responses in reverse order

Question 21

On whose behalf does a forward proxy usually act?

Accepted Answer

Clients, such as employees on a corporate network

Answer

The backend servers of a website

Answer

The internet's DNS servers

Answer

The application's primary database

Question 22

Which option is an example of a software load balancer?

Accepted Answer

NGINX or HAProxy

Answer

PostgreSQL or MySQL

Answer

Redis or Memcached

Answer

Webpack or Vite

Question 23

In AWS, what is the general difference between ALB and NLB?

Accepted Answer

ALB works at Layer 7 (HTTP) and NLB at Layer 4 (TCP/UDP)

Answer

ALB is only for databases and NLB is only for web

Answer

ALB is free and NLB is always paid

Answer

ALB supports only IPv6 and NLB only IPv4

Question 24

What information does the `X-Forwarded-For` header give the backend?

Accepted Answer

The original client IP address before passing through the proxy

Answer

The IP address of the target backend server

Answer

The domain name the user previously visited

Answer

The load balancer's internal port number

Question 25

What does an API gateway typically do?

Accepted Answer

A single API entry point with routing, authentication, and rate limiting

Answer

Storing static files such as images and videos

Answer

Running database migrations during deployment

Answer

Compiling frontend code for browsers

Question 26

Why is rate limiting often done at the API gateway or reverse proxy?

Accepted Answer

So excess requests are rejected before reaching the services

Answer

Because backends technically cannot reject requests

Answer

To increase response speed for every user

Answer

Because rate limiting only works with UDP

Question 27

What is the main job of DNS?

Accepted Answer

Translating domain names into IP addresses

Answer

Encrypting traffic between browser and server

Answer

Distributing large files among users

Answer

Managing a website's user sessions

Question 28

What does a DNS `A` record specify?

Accepted Answer

The IPv4 address of a domain name

Answer

The IPv6 address of a domain name

Answer

The mail server of a domain

Answer

An alias for another domain name

Question 29

Which DNS record maps a domain name to an IPv6 address?

Accepted Answer

`AAAA`

Answer

`A`

Answer

`MX`

Answer

`TXT`

Question 30

What does a DNS `CNAME` record do?

Accepted Answer

Defines one name as an alias of another name

Answer

Maps a name directly to an IPv4 address

Answer

Specifies the domain's mail server

Answer

Specifies the domain's authoritative servers

Question 31

What is a DNS `MX` record for?

Accepted Answer

Specifying the mail servers that receive a domain's email

Answer

Specifying the maximum number of allowed requests

Answer

Mapping a domain to an IPv6 address

Answer

Defining the default TTL for all records

Question 32

What does a DNS `NS` record specify?

Accepted Answer

The authoritative nameservers for a domain

Answer

The IP address of the main web server

Answer

The domain's HTTPS certificate

Answer

The list of disabled subdomains

Question 33

What does TTL mean on a DNS record?

Accepted Answer

How long resolvers may cache the answer

Answer

How long the server takes to respond

Answer

How many times a domain can be changed

Answer

The domain registration's expiry date

Question 34

How does DNS round robin work?

Accepted Answer

Several IPs are returned for one name, rotating their order

Answer

DNS directly proxies HTTP requests to servers

Answer

Each user receives a different domain every time

Answer

The DNS server checks each backend's health over HTTP

Question 35

What does GeoDNS do?

Accepted Answer

Returns the IP of the nearest datacenter based on user location

Answer

Stores users' precise GPS location

Answer

Categorizes domains by the registrant's country

Answer

Sends every country's traffic to one central server

Question 36

What does anycast mean in networking?

Accepted Answer

One IP is announced from many locations and traffic reaches the nearest

Answer

One message is sent to every device on the network

Answer

Every server has one unique, fixed IP

Answer

One message is sent to a specific group of receivers

Question 37

What is the risk of having only one load balancer?

Accepted Answer

The load balancer itself becomes a single point of failure

Answer

Backend servers can no longer connect to the database

Answer

The domain's DNS is deleted automatically

Answer

There is no risk because load balancers never fail

Question 38

In an active-passive load balancer setup, what does the passive node do?

Accepted Answer

Stands by and takes over if the active node fails

Answer

Always processes half of the traffic

Answer

Only stores the active node's logs

Answer

Processes HTTPS requests separately from HTTP

Question 39

What is a Virtual IP (VIP) in the context of load balancers?

Accepted Answer

An address clients connect to that can move between nodes

Answer

The private IP address of each backend server

Answer

An address used only in development environments

Answer

A special IP for high-priority VIP users

Question 40

What is a common advantage of software load balancers over hardware ones?

Accepted Answer

Lower cost and flexibility to run on commodity servers

Answer

They always have higher throughput than any hardware device

Answer

They never need configuration or maintenance

Answer

Only they can distribute TCP traffic

Question 41

What does connection draining do when removing a server from a load balancer?

Accepted Answer

Stops new requests but lets in-flight requests finish

Answer

Immediately cuts all active connections

Answer

Moves connections into a file on disk

Answer

Replaces the server with a new version

Question 42

What does a `502 Bad Gateway` from a load balancer usually mean?

Accepted Answer

The backend returned an invalid response or closed the connection

Answer

The client sent a request with invalid syntax

Answer

The user is not allowed to access this resource

Answer

The requested resource does not exist on the server

Question 43

When is a `504 Gateway Timeout` returned?

Accepted Answer

When the backend does not answer the proxy within the allowed time

Answer

When the client has sent too many requests

Answer

When the server's certificate has expired

Answer

When the HTTP method is not allowed for that path

Question 44

What must a reverse proxy support to pass WebSocket traffic through?

Accepted Answer

The `Upgrade` header and keeping a long-lived connection open

Answer

Only `GET` requests without a body

Answer

gzip compression for every response

Answer

Caching every message for 24 hours

Question 45

How does a load balancer improve service availability?

Accepted Answer

By routing traffic only to healthy servers when one fails

Answer

By completely preventing servers from crashing

Answer

By automatically backing up the database

Answer

By increasing users' internet bandwidth

Question 46

What role does a load balancer play alongside autoscaling?

Accepted Answer

It adds new instances to rotation and removes terminated ones

Answer

It decides which code is deployed on instances

Answer

It calculates and pays for each instance

Answer

It limits the number of allowed users in the system

Question 47

What problem does service discovery solve?

Accepted Answer

Finding current addresses of service instances that keep changing

Answer

Finding security bugs in services' code

Answer

Discovering new users for the application

Answer

Finding slow database queries

Question 48

On which default port does a load balancer usually accept HTTPS traffic?

Accepted Answer

`443`

Answer

`80`

Answer

`22`

Answer

`3306`

Question 49

What is the benefit of caching responses at the reverse proxy?

Accepted Answer

Repeated responses are served without hitting the backend

Answer

Database data stays in sync forever

Answer

The need for HTTPS disappears

Answer

Users no longer need to log in

Question 50

What is host-based routing on a Layer 7 load balancer?

Accepted Answer

Routing requests to different services based on the `Host` header

Answer

Sending requests to the server with the shorter hostname

Answer

Choosing a server based on the client's operating system

Answer

Keeping each website on a separate load balancer