What is the primary technical term for adding more servers to handle increased traffic, rather than upgrading an existing server?

Horizontal Scaling. Horizontal scaling involves adding more machines (scaling out). Vertical scaling involves upgrading a single machine's resources (scaling up).

If a large website relied solely on one extremely powerful server, what risk increases dramatically?

Single Point of Failure (SPOF). A single point of failure means if that one server fails (due to hardware error, power outage, etc.), the entire website goes offline.

What is the specialized network device or software used to efficiently divide incoming web traffic across multiple backend servers?

Load Balancer. Load balancers distribute user requests across the server cluster, ensuring no single server is overwhelmed and providing redundancy.

Beyond traffic handling, why do multiple servers improve website reliability?

They provide redundancy and fault tolerance.. If one server fails, the others in the cluster automatically pick up the load, ensuring service continuity (fault tolerance).

The concept that increasing the capacity of a single machine (vertical scaling) yields progressively smaller performance gains for exponentially higher costs is known as:

The Law of Diminishing Returns. The Law of Diminishing Returns applies because hardware prices escalate sharply for the absolute highest performance components, making adding many smaller servers cheaper and more efficient.

Which term describes the ability of a system to continue functioning, even if one or more components fail?

Fault Tolerance. Fault tolerance is the design principle that allows a system to withstand partial failures without total shutdown.

Which Load Balancer algorithm sends new traffic to the server that currently has the fewest active connections?

Least Connection. The Least Connection method is dynamic and ensures that the workload is balanced based on current server utilization rather than just server count.

What is the technique called when a large database is partitioned into smaller, independent databases (e.g., storing user data in different geographic zones)?

Sharding. Sharding is a method of horizontal partitioning of data, allowing the database load to be spread across multiple machines.

When scaling horizontally, engineers prefer many commodity servers over specialized high-end servers primarily because of:

Lower Total Cost of Ownership (TCO) and easier replacement.. Commodity hardware is cheap, standardized, and easy to replace. High-end proprietary hardware is extremely expensive to acquire and maintain.

The main performance challenge introduced when transitioning from a single server to multiple distributed servers is:

The complexity of state and data synchronization.. Ensuring all servers have the latest, consistent copy of user data and session information is the central challenge of distributed systems.

Why is distributing servers geographically vital for large international websites?

To reduce network latency for distant users.. Servers closer to the end-user reduce the physical distance data must travel, significantly decreasing network latency.

What architectural style breaks a single, large application into a collection of smaller, independently deployable services?

Microservices Architecture. Microservices allow different parts of the application (e.g., payment processing vs. user authentication) to scale independently across different servers.

In a distributed web setup, what is the role of a 'Reverse Proxy'?

It acts as a gateway, receiving client requests and forwarding them to one or more backend servers.. The reverse proxy (often doubling as a load balancer) shields the backend servers and directs incoming traffic.

What does a Load Balancer perform, specifically to determine if a server in the pool is still operational and ready to receive traffic?

Health Checks. Health checks are automated pings or requests sent to backend servers; if the server doesn't respond correctly, the load balancer temporarily removes it from the pool.

What is 'Session Affinity' (or 'Sticky Sessions') used for in load balancing?

To ensure a user's subsequent requests always return to the same backend server.. If session data (like an uncompleted shopping cart) is stored locally on a server, Session Affinity ensures the user remains connected to that specific server to maintain state.

What is the key disadvantage of vertical scaling compared to horizontal scaling?

It eventually hits physical limits and high costs.. There's a physical limit to how much RAM or how many CPUs you can put into one chassis, and the cost curve for high-end parts is steep.

When a web service uses multiple servers, a failure in one server might overload and cause failures in adjacent dependent servers. This is called a:

Cascading Failure. A cascading failure happens when the initial failure is passed along or amplified by dependencies within the system.

The percentage of time a service or system is operational and available to users is referred to as:

Uptime. Uptime is the critical metric for measuring service availability, often expressed in 'nines' (e.g., 99.999%).

Why is scaling stateless web servers easier than scaling databases?

Web servers don't store persistent data that requires synchronization.. Stateless servers (which don't store user session data locally) can simply be duplicated. Databases, which store state, require complex synchronization and consistency protocols.

The process where a distributed system automatically adds or removes computing resources based on fluctuations in demand is known as:

Elasticity or Autoscaling. Elasticity allows resources to scale in or out instantly, minimizing wasted resources during low traffic and preventing outages during peak traffic.

Content Delivery Networks (CDNs) relate to distributed systems by:

Creating global server replicas to cache static content closer to users.. CDNs distribute static assets (images, videos, scripts) globally, enhancing performance by reducing latency for users everywhere.

Which of these is NOT a core reason for large websites to use distributed server architecture?

Elimination of all software bugs.. Distributed systems improve reliability and performance, but they cannot eliminate software bugs; in fact, they often make debugging more complex.

Why is 'Loose Coupling' favored in microservices architecture over tightly coupled monolithic architecture?

Loosely coupled components can be updated and scaled independently.. Loose coupling means services interact but are not heavily dependent; the failure or update of one service doesn't require updating or restarting the entire system.

Which type of server usually sits closest to the end-user in a geographically distributed system, often handling caching and security checks?

Edge Server (or Proxy/CDN point). Edge servers are geographically distributed points of presence that handle initial user contact, serving cached content, and forwarding API calls.

What is the primary drawback of using a simple 'Round Robin' load balancing strategy?

It ignores the current load or performance differences between servers.. Round Robin blindly sends traffic sequentially. If one server is much weaker or is currently busy processing a heavy task, Round Robin will still send it a new request, potentially leading to slow performance on that server.

QuizKnowledge

Scaling Giants: Why Big Websites Go Distributed

Test your knowledge on why large websites use horizontal scaling, load balancers, and microservices architecture for maximum performance and reliability.

25 Questions•20 Mins•Medium