Terminology

• SmallRye - Out the box implementation of MicroProfiles, using JBoss stuff for getting up and running, basically "plugins", IBM competitor is openliberty

• MicroProfile - Eclipse MicroProfile, microservice framework composed of made up of (bare min)

  • JAX-RS - API for creating REST APIs
  • JSON-P - API for processing JSON
  • CDI 2.0 - Glue JAX-RS and JSON-P together dependency injection using java beans, share a class and manage its lifecycle
  • JSON-B - JSON binding (similar to jackson databind)
  • Open Tracing - standardise tracing/logging/profile metadata across a distributed service, provides wire protocol api spec and impl etc
  • Other stuff like Metrics (prometheus), watchdogs, retry, circuit breakers etc

• Loom - Virtual threads approach to non-blocking IO (as opposed to event-loop)

• Quarkus

  • ArC - CDI (injection framework), developed from Weld
  • Command mode - terminates exits after run as opposed to normal operation (wait for systemexit or pod to be killed)

• Jandex - Java annotations indexer (Jason Green dev)

• mod_cluster - undertow based Load Balancer

• Hotspot vs Native

  • Hotspot JVM (JIT) (Andrew Dinn in the office... well when theres not a global pandemic)
  • Native via GraalVM (AOT)

• Testing frameworks

  • JUnit - Main Java testing framework
  • Rest-assured - Test REST APIs
  • Awaitility - Java DSL test framework for async systems
  • FitNesse - A fork of Fit (Framework for Integrated Test) Acceptance testing framework

• Testing methodologies

  • Smoke testing - Simple tests for build verification eg. does it run without segfault/strack trace
  • Unit testing - Testing class/methods
  • Integration testing - Testing application framed by user scenarios
  • System testing - Holistic blackbox testing between integration components
  • Acceptance testing -
    • User acceptance testing - Factory acceptance testing (FAT), Site acceptance testing (SAT)
    • Operational acceptance testing - Things that allow a system to be maintained, checking of backups, checking of patching mechanisms etc
    • Contractural and regulation acceptance testing - Test against criteria as documented (Prod gatechecks for example)
  • Debugging (not actually testing but relevent experience, from gdb on ChibiOS to intelij)

General Notes

Quarkus built upon VERT-X for non-blocking IO, VERT-X mostly built upon netty, Vert-X provides a 80+ connectors for non-blocking IO like Kafka, mongodb, AMQP etc Quarkus two main deploy modes - native (graalvm) and jvm, some things dont do so well natively

//Mutiny async construct, onItem <- the result, onFailure <- Failed calls
Uni<String> uni
uni.onItem()
.onFailure()

The Job spec

• Implement test scenarios and develop enterpirse applications for testing
• Design and develop testing frameworks in Java for various environments, like cloud, multiple operating systems, large clusters, and containers
• Provide root cause analysis for complex bugs reported by our customers and communicate the impact of these bugs to customers, developers, and support specialists
• Participate in defining testing strategies
• Analyze and prioritize the quality risks posed by Red Hat products and determine the tests required to mitigate these risks
• Serve as an internal advocate for our customers to ensure that the product goes to customers without usability or user experience issues; properly document relevant information
• Document new bugs and work with project managers and developers on prioritization and a fix
• Mentor junior team members
• Constantly learn new things and maintain an overview of current technologies like Java, MicroProfile, and Reactive programming

Technical Job specs

• Practical experience with Java EE, MicroProfile, or Spring development
• Experience with testing in Junit or TestNG beyond unit tests
• Experience with Maven, Git, Jenkins, continuous integration, and pipelines
• Debugging and performance tuning skills
• Ability to install and maintain Red Hat Enterprise Linux (RHEL) or UNIX-like operating systems; basic administration skills of Jenkins nodes

Main technologies to learn about

• Microprofile + SmallRye
• JUnit
• TestNG - Many of its features on the roadmap for JUnit 5.x anyway
• Jenkins, CI, piplines Blue Ocean
• Maven - Already know most of this
  • Have fixed bugs in maven plugins
  • Used maven extensively in productisation
  • Good knowledge of prod proccess

JUnit5

Architecture

• Platform - Test engine itself
• Jupiter - JUnit5 API/Annotations
• Vintage - API for older JUnit4 tests
• Ext - 3rd Party extensions (custom API aka SPI)

"No news is good news" - We only care for failures

• @Test annotation to highlight test method we'd like to run as part of JUnit tests
• Common assert methods

assertEquals(expected, actual, "Description of failure") // Asserts both values are equal
assertArrayEquals(expectedArray, actualArray) //Verifies each item in the array are equal in the right position/order
assertIterableEquals(expectedArray, actualArray) // Verifies iterables, again equal and in the right position/order
fail() //Always fail... duh
assertAll(
    () -> assertEquals(expected, actual),
    //() -> ...
    )

Surefire plugin in maven to run JUnit tests (thing we always disabled in Prod 😉)

• Test lifecycle annotations

@BeforeAll //needs to be static method
@AfterAll //needs to be a static method
@BeforeEach
@AfterEach

//eg init a new pojo for each test
class Foo {
    SomePojo pojo;
    @BeforeEach
    void init() {
        pojo = new SomePojo();
    }
}

• Nested

class Foo {

    @Nested
    class Bar {
        @Test
        @DisplayName("Test Description")
        void dosomething()
        {
            assertEquals(expected, pojo.method(x,y), "Test description");
            assertEquals(Somethinexpected, pojo.method(x), "Test description 2"
        }
        void dosomethingelse()
        {
            assertEquals(expected, pojo.method(x,y), "Test description");
            assertEquals(Somethinexpected, pojo.method(x), "Test description 2"
        }
    }
}

• Lambda for lazy tests

assertEquals(expected, actual, () -> "big bad string: " + expected + "actual: " + actual) //Will only execute lambda if assert fires

• Repeated test @RepeatedTest(n) - Repeat a test multiple times (n number of times) has RepetitionInfo

• @Tag annotation allows running of many or few of multiple tags

• JUnit provides many providers (injected) eg. RepetitionInfo or testInfo

QE tests

Github

• Quarkus-startstop - Measures time and memory usage

  • Did something similar when working on opensplice DDS, combined a demo application with measurement output, used to track metrics not perfect as it didn't account for multiple tests/threads running, sort of solved with CPU pinning

• Quarkus jenkinsfiles - Jenkins pipelines, based on Groovy

  • Experience with groovy https://github.com/jboss-fuse/fuse-build-yaml/blob/7.x.redhat-7-x/src/main/groovy/formatYaml.groovy (all my groovy looks like java)
  • I've created Jenkins while in prod to do a number of tasks, namely
    • Interogating brew build root
    • Assembling various maven repository sets
    • Creating reports (build from source %)
    • Templates for email, html, Jira
  • Used a mixture of bash and python at the time but this is what prod tooling used at the time

• Openshift test suite - Integration tests for services on openshift

  • Deploys service via openshift YAML e.g. AMQ then tests quarkus application against it
  • Experience with deploying own openshift cluster and apps (Old laptop... still running now just cant access it!)
  • Experience with managing openshift cluster (Qucklabs)

• Beefy-scenarios -

The plan

• Understand technologies and terminologies
• Fire up Quarkus example (hello world)
• Design an LED conversion proxy for front porch
  • Listen to MQTT broker
  • Possibly use camel k to write mqtt to file or something
  • "Translate" and passthrough to porchLED
  • Write TestNG tests
  • Mockquito

Refs

• What is microprofile
• What is SmallRye
• CDI explained
• Mutiny lib/Vert-X
• Weld
• Quickstart hello world & Building native image - Used multistage build
• Awaitility