The load balancer
Spreading traffic across several servers: the algorithms, layer 4 versus layer 7, health checks, sessions, and what happens when a server or the balancer itself fails.
- 34 cards
- 3 topics
- Free
The outline, topic by topic
Each topic is weighted the way the exam weighs it.
- Concepts
- Scenarios
- Trade-offs and failures
A few cards from the deck
Load balancer: what is it?
Show the answer
A component that spreads incoming requests across several servers.
You scale by adding servers, and the service survives losing one of them.
Your only stateless web server is at 100 % CPU. What do you do?
- Add servers behind a load balancer
- Lengthen the timeout
- Add a database index
- Lengthen DNS caching
Show the answer
Add servers behind a load balancer
A stateless app multiplies easily: the load balancer shares the work.
Round robin: what is it?
Show the answer
Each new request goes to the next server in turn.
Simple and fair when servers and requests are alike.
The only load balancer fails. What happens?
- Nothing is reachable: it is a single point of failure
- Requests go straight to the servers
- Only new clients are affected
- Nothing, servers balance themselves
Show the answer
Nothing is reachable: it is a single point of failure
Run two (active and passive with a floating address) or use a redundant service.