Loan Broker example using microservices in Kubernetes.
The Loan Broker is a very common example of demonstrating how you "compose" services.
Taken from the Enterprise Integration Pattern examples:
When shopping for a loan, a customer usually calls several banks to find the deal with the best possible interest rate.
Each bank asks the customer for his or her social security number, the amount of the loan and the desired term , i.e.
the number of months until the loan has to be paid off. Each bank then investigates the customer's credit background,
usually by contacting a credit agency. Based on the requested terms and the customer's credit history, the bank responds
with an interest rate quote to the consumer (or declines thankfully). Once the customer has received quotes from all banks,
he or she can then select the best offer, i.e. the lowest interest rate.
This example demonstrates the use case above using Kubernetes, Spring Boot, Hystrix, Ribbon and OpenZipkin.
The quickest way to get started is to deploy everything in a single step:
cd examples/loanbroker
mvn -Pf8-local-deploy
Or you can build and deploy each part of the example individually.
The credit buraeu is responsible for evaluating the creditworthiness of the client. It provides a rest endpoint which accepts the client's SSN as a parameter and return a score (0-1000).
To build to the project:
cd examples/loanbroker/credit-buraeu
mvn clean package fabric8:build fabric8:deploy
You can access the servlet at: http://loanbroker-credit-bureau.vagrant.f8/hello (or wherever the loanbroker-credit-buraeu service is bound).
Each bank communicates with the credit bureau and receives the creditworthiness score of the client. It uses an algorithm that accepts as input the base rate of the bank, the score and the duration of the loan and calculates the interest.
The bank communicate with the credit bureau via ribbon.
To create a bank
cd examples/loanbroker/bank
mvn clean package fabric8:build fabric8:deploy -Dbank.id=<bank's id> -Dbank.name=<bank's name> -Dbank.rate=<base rate>
You can create as many banks as you like by repeating the last command with different arguments. For example:
mvn clean package fabric8:build fabric8:deploy -Dbank.id=bank1 -Dbank.name="Bank One" -Dbank.rate=15
mvn clean package fabric8:deploy -Dbank.id=bank2 -Dbank.name="Bank Two" -Dbank.rate=20
mvn clean package fabric8:deploy -Dbank.id=bank3 -Dbank.name="Bank Three" -Dbank.rate=25
The broker as the name implies accepts a loan request from the client and broadcasts the request to all "known" banks. Then it collects the results and presents them to the user.
Discovery is done via Kubernetes services (it discovers all services with the label {project: loanbroker-bank}.
Internally the broker is using Spring's Rest Template to communicate with the Banks and each request is wrapped inside Hystrix command.
To build and create the broker:
cd examples/loanbroker/broker
mvn clean package fabric8:build fabric8:deploy
Once the broker has been created you can send http requests to to the broker url, for example:
curl http://loanbroker-broker.vagrant.f8/quote?ssn=12345678&duration=12&amount=10000
To generate the requests automatically a small generator app has been added to the loanbroker example. Its a simple camel/springboot application that generates random loan requests to the rboker.
To build and create the generator:
cd examples/loanbroker/broker
mvn clean package fabric8:build fabric8:deploy
It's not a mandatory step but it does help a lot if you bring up the Turbine Server and Hystrix Dashboard and have a glimpse at the dashboard.
The loanbroker example is using Kubernetes integration with Spring Cloud which among others supports discovery and integration with OpenZipkin.
This means that if zipkin is discovered, then all loan requests will be traced and the traces will be visilbe in the ZipKin console:
And here is the dependency view taken from the Zipkin Condole
