Load Balancers
Summary
PDF p.140A load balancer distributes client requests across multiple server nodes to manage varying loads, provide fault tolerance, and mitigate denial of service attacks. There are two main types: Layer 4 load balancers, which make decisions based on IP addresses and ports, and Layer 7 load balancers, which use application-level data for more complex routing.
In plain words
Supplementary — not from your PDFA load balancer spreads requests across a pool of servers. That improves scalability and fault tolerance and helps absorb denial of service traffic. Layer 4 balancers decide by IP address and port; layer 7 balancers (content switches) decide by application data such as the URL. Scheduling can be round robin, fewest connections or best response time, and health checks keep failed nodes out of the pool.
Detailed explanation
PDF p.140-
Function
- Distribution: Spreads client requests across server nodes in a farm or pool.
- Scalability: Handles light to heavy loads.
- Fault Tolerance: Forwards requests to another server if one fails.
- Use Cases: Web servers, email servers, web conferencing, video conferencing, streaming media servers.
-
Types of Load Balancers
-
Layer 4 Load Balancer
- Operation: Makes forwarding decisions based on IP address and TCP/UDP port values.
- OSI Layer: Transport layer.
-
Layer 7 Load Balancer (Content Switch)
- Operation: Makes forwarding decisions based on application-level data (e.g., URL, data types like video or audio streaming).
- OSI Layer: Application layer.
- Complex Logic: Requires more processing power.
-
Layer 4 Load Balancer
-
Scheduling
- Algorithm: Determines which node processes each incoming request.
-
Methods
- Round Robin: Picks the next node in sequence.
- Fewest Connections: Chooses the node with the fewest active connections.
- Best Response Time: Selects the node with the best response time.
- Weighting: Can use administrator-set preferences or dynamic load information.
- Health Checks: Uses heartbeat or health check probes to verify node availability and load.
-
Source IP Affinity and Session Persistence
-
Source IP Affinity
- Layer 4 Approach: Keeps client sessions connected to the initial node.
-
Session Persistence
- Application-Layer Load Balancer: Uses cookies to maintain session connections.
- Reliability: More reliable than source IP affinity but requires browser cookie acceptance.
-
Source IP Affinity
Important terms
taken from the text above- Distribution
- Spreads client requests across server nodes in a farm or pool.
- Fault Tolerance
- Forwards requests to another server if one fails.
- OSI Layer
- Transport layer.
- Complex Logic
- Requires more processing power.
- Algorithm
- Determines which node processes each incoming request.
- Round Robin
- Picks the next node in sequence.
- Fewest Connections
- Chooses the node with the fewest active connections.
- Best Response Time
- Selects the node with the best response time.
- Weighting
- Can use administrator-set preferences or dynamic load information.
- Health Checks
- Uses heartbeat or health check probes to verify node availability and load.
- Layer 4 Approach
- Keeps client sessions connected to the initial node.
- Application-Layer Load Balancer
- Uses cookies to maintain session connections.
- Reliability
- More reliable than source IP affinity but requires browser cookie acceptance.
Examples & real-world scenarios
Supplementary — not from your PDF- Round-robin distribution across three web servers.
- A layer 7 balancer sending /video requests to streaming servers.
- Cookie-based session persistence keeping a shopper on one node.
Scenario
During a flash sale one web server fails. Health checks detect it and the load balancer stops sending requests to it, so customers are served by the remaining nodes.
Common mistakes
Supplementary — not from your PDF- Mixing up source IP affinity (layer 4) and cookie-based session persistence (layer 7).
- Forgetting that a load balancer is also a corrective control against DoS.
Practical skills
Supplementary — not from your PDF- Choose a scheduling method.
- Explain session persistence options.
What I should remember
Key Points PDF p.140-
Function
- Distribution: Client requests across server nodes.
- Scalability: Light to heavy loads.
- Fault Tolerance: Server failure handling.
- Use Cases: Web, email, conferencing, streaming servers.
-
Types of Load Balancers
- Layer 4: IP address, TCP/UDP port values.
- Layer 7: Application-level data, complex logic.
-
Scheduling
- Algorithm: Node selection.
- Methods: Round robin, fewest connections, best response time.
- Health Checks: Node availability and load verification.
-
Source IP Affinity and Session Persistence
- Source IP Affinity: Layer 4, initial node connection.
- Session Persistence: Application-layer, cookie-based.