This workshop will provide developers with hands on experience in building Pivotal Cloud Cache(PCC) clients using Spring Data GemFire (SDG), Spring Data REST, Spring Cloud and Spring Boot. In this session we’ll be implementing a pizza store app for ordering pizza backed by PCC. Session includes presentations, demos and hands on labs.
Cloud Foundry services are managed through two primary types of operations:
- Create/Delete
These operations create or delete instances of a service. For a database this could mean creating/deleting a schema in an existing multitenant cluster or creating/deleting a dedicated database cluster.
- Bind/Unbind
These operations create or delete unique credential sets for an existing service instance that can then be injected into the environment of an application instance.
There are two ways to discover what services are available on Pivotal Cloudfoundry. The first is available on any instance of Cloud Foundry: the CLI. Just type:
$ cf marketplace
and you’ll get a list of services, their available plans, and descriptions. On Pivotal Cloudfoundry, the ``free'' tier of plans is normally the first one listed.
The second way is specific to Pivotal Cloudfoundry’s Apps Manager UI. If you haven’t already, login to it by visiting http://console.devcloudwest.inbcu.com
Click on the ''Marketplace'' link, and you’ll see the same service/plan/description listing in the browser:
Pizza store app will make service calls to this mysql instance
cf create-service p.mysql db-small workshop-db
Services can be created through Apps Manager Marketplace or by executing cf cli commands
cf marketplace -s p-cloudcache
cf create-service p-cloudcache extra-small workshop-pcc
cf create-service-key workshop-pcc devkey
cf service-key workshop-pcc devkey
This will return info about the cluster login.
"urls": {
"gfsh": "https://cloudcache-ca3ab6cf-53ea-4495-8f8f-57a201c32257.cf.*******.net/gemfire/v1",
"pulse": "https://cloudcache-ca3ab6cf-53ea-4495-8f8f-57a201c32257.cf.*******.net/pulse"
},
"users": [
{
"password": "*********",
"roles": [
"cluster_operator"
],
"username": "cluster_operator_*********"
}
Replace URL and password with the info retrieved above
connect --use-http=true --url=http://gemfire-xxxx-xxx-xx-xxxx.system.excelsiorcloud.com/gemfire/v1 --user=cluster_operator --password=*******
-
Change to the pizza-store-pcc-client application directory:
$ cd ../../pizza-store-pcc-client
Open the manifest.yml file. You will see the following:
applications: - name: pizza-store-pcc-client random-route: true path: target/pizza-store-pcc-client-0.0.1-SNAPSHOT.jar
This will app will be deployed with the name pizza-store-pcc-client using target/pizza-store-pcc-client-0.0.1-SNAPSHOT.jar and a random route url.
Let’s deploy without starting the app.
cf push --no-start
cf bind-service pizza-store-pcc-client workshop-db cf bind-service pizza-store-pcc-client workshop-pcc
Let’s load the db using /load endpoint. Make sure to use your route URL
New customers successfully saved into Database
Let’s fetch top 10 customers using /showdb endpoint. Make sure to use your route URL
First 10 customers are show here: Customer [id=0FsIMYG30, name=Brody England, [email protected], address=59644 New York, birthday=1973-07-16T03:34:20.016Z] Customer [id=2afB7G57Z, name=Ella Robbins, [email protected], address=72508 San Francisco, birthday=1942-05-23T20:49:41.209Z] Customer [id=2dO6dSUIs, name=Eva Fischer, [email protected], address=66604 New York, birthday=1949-05-29T22:01:01.136Z] Customer [id=2mDAi0qRr, name=Gianna Merritt, [email protected], address=25734 Washington, birthday=1969-01-20T14:08:36.941Z] ....
Lets look at the APIs available for our pizza store app using /pizzas endpoint.
Lets Order Some Pizza
-------------------------------
types: plain, fancy
GET /orderPizza?email={emailId}&type={pizzaType} - Order a pizza
GET /orders?email={emailId} - get specific value
Now we can order some pizza by invoking /orderPizza endpoint. Make sure to use your route URL and pick an email address from data returned by /showdb endpoint.
[email protected]&type=fancy" class="bare">https://pizza-store-pcc-client-boring-crane.cfapps.haas-69.pez.pivotal.io/orderPizza?email=[email protected]&type=fancy
Cache Miss Scenario
When you call that endpoint for the first time. Cache is missed and data is pulled from DB.
Result [Pizza{name='fancy', toppings=[arugula, chicken], sauce='pesto', Customer='Customer [id=05eKpgOFA, name=Lucy Norton, [email protected], address=48665 Washington, birthday=1965-02-10T06:20:27.828Z]'}]
Cache Miss for Customer [true]
Read from [MYSQL]
Elapsed Time [234 ms]
Data Returned From Cache
Invoke that endpoint again, and this time data will be pulled from Cache and it will return much faster.
[email protected]&type=fancy" class="bare">https://pizza-store-pcc-client-boring-crane.cfapps.haas-69.pez.pivotal.io/orderPizza?email=[email protected]&type=fancy
Result [Pizza{name='fancy', toppings=[arugula, chicken], sauce='pesto', Customer='Customer [id=05eKpgOFA, name=Lucy Norton, [email protected], address=48665 Washington, birthday=1965-02-10T06:20:27.828Z]'}]
Cache Miss for Customer [false]
Read from [PCC]
Elapsed Time [2 ms]
We can look at the order status using /orders endpoint.
But first, lets create an index in our cache, by going to gfsh cli.
create index --name=CustomerEmailIndex --region="/pizza_orders o" --expression="o.customerInfo.email"
If you look at the code under io.pivotal.data.repo.PizzaOrderRepo, we are using the index with our query.
Now lets fetch the order status
[email protected]" class="bare">https://pizza-store-pcc-client-boring-crane.cfapps.haas-69.pez.pivotal.io/orders?email=[email protected]
It will use the index and fetch data from Cache
Result [Pizza{name='plain', toppings=[cheese], sauce='red', Customer='Customer [id=2gOSU3V2O, name=Nathan Mendez, [email protected], address=91105 New York, birthday=1955-07-16T01:13:31.733Z]'}
]
Elapsed Time [24 ms]