A replica is an independently running copy of a component or service. Running several replicas lets requests execute in parallel instead of waiting behind one internally single-threaded component instance. section 6.3
In a wasmCloud deployment, replicas provide:
- Parallel request handling through NATS lattice load balancing. section 6.3
- Distribution across hosts, zones, machines, data centers, or geographic regions. section 6.4
- Improved fault tolerance and potentially lower latency through weighted regional placement. section 6.4
Scale and benchmark an HTTP component
Set the spreadscaler count to instances: 10, redeploy with `wash app delete rust-hello-world` and `wash app deploy ./wadm.yaml`, then run `oha -n 10 -c 5 http://127.0.0.1:8000`. The average response time falls from about 8 seconds with one replica to about 2 seconds with ten.
Figure 6.4 Updated architecture diagram for “Hello, World” wasmCloud app with 10 replicas of the http-component split across different hosts and regions (i.e., us-east-1 and us-west-1)
Trace how replicas of one HTTP component can be split across East and West Coast hosts and regions. section 6.4
