JAX-RS Sample application for Wildfly enhanced with kensu monitoring capabilities

An example of a Rest web application, using JAX-RS, Hibernate, MySQL, and Clickhouse.

The data lineages and data quality are tracked automatically between HTTP endpoints and Databases using Kensu collectors.

By default, the Jaeger backend tracer + Zipkin reporter is used to keep the spans published to a observability platform, this tracer is combined with the Kensu tracer to collect what is important for Kensu to build lineage and stats.

One-click run all in Docker

First run ./one-click-start-all-prebuilt.sh (to start this app with MySQL & Zipkin inside a docker), and, after waiting a few minutes for it to start up, visit some of these endpoints:

This will trigger some metadata on data movement and data quality getting collected, and sent to Zipkin and/or Kensu servers.

Check results: Zipkin & Kensu app

If you have access to Kensu App, on its UI you may first check - data catalog, lineages, and projects.

To understand the internals (debug the raw data sent), visit Zipkin at http://127.0.0.1:9411/zipkin/?serviceName=sample-service and click on Run query - you'll see list of traces/spans that were sent, including info on: http request (url/params), SQL query & stats, and http response (json schema), like this:

That's it! All the rest info below is only for advanced techy users wanting to explore the internals.

Build it yourself and run

Starting Jax-Rs server for sample app (incl Wildflow & MySQL in docker):

( cd ./docker-dir && run-docker.sh )
./build-dependencies.sh
./build-and-run-app.sh

Then, follow the same steps as in the instructions above.

Technical details

Databases

MySQL

P.S. Data is preloaded into the Docker. You need the below info only for custom setup.

Tutorial

The MySQL import script from the sample MySQL database is located in src/test/resources/data-mysql.sql.

Kensu demodb

This instance is available on Kensu cloud and is used for the public Demos (used by Tableau).

Clickhouse

Tutorial database

You can load it as explained here.

Ported MySQL database

The MySQL import script from the sample MySQL database is ported in test/resources/data-clickhouse.sql and can loaded as a regular script in Clickhouse (clickhouse-client < src/test/resources/data-clickhouse.sql).

Application

Sources

The application source is in the package io.kensu.example.jboss and consists of services:

VisitService: returns data from Clickhouse tutorial database
OrderDetailsService and ProductLineService: querying the sample MySQL database

Configuration

Databases

The data sources access for JPA/Hibernate are defined in src/main/resources/META-INF/persistence.xml.

Note: the Demo server has its password configured as an interpolated MAVEN property whic should be set by a profile.

JBOSS env

src/main/webapp/WEB-INF/beans.xml: enables CDI discover (Weld) for the KensuTracerFactory see ([Tracer Bean](#Tracer Bean)).
src/main/webapp/WEB-INF/jboss-deployment-structure.xml: to avoid the default tracing capabilities of JBOSS (microprofile / smallrye) to take the precedence over Kensu one (use KensuTracerFactory bean).

Collector

Configuration

Tracer Bean

src/main/java/io/kensu/collector/config/KensuTracerFactory.java: Configures the Jaeger and Kensu tracers, and registers it as a Bean.
src/main/resources/app.properties: Maven interpolated file defining the application name and version.
src/main/resources/kensu-tracer.properties: configures the Kensu collector (using MAVEN interpolation to avoid committing URLs and Tokens) and high level information like application (process) name, code base, user, and such.
src/main/resources/META-INF/services/io.opentracing.contrib.tracerresolver.TracerFactory: Java service loader for the tracerresolver dependency making sure that `io.opentracing.contrib.tracerresolver.TracerResolver.resolve picks the one we must use

Jaeger & more

Jaeger is configured using the Java properties or variable env by default, like explained here.

Other configurations or initializers happen in the following classes as well:

src/main/java/io/kensu/collector/config/DamProcessEnvironment.java: utility to set Kensu model high level metadata about the application and its context (process name, run, user, code, ...).
src/main/java/io/kensu/collector/config/TracingContextListener.java: makes sure GlobalTracer is set.

Dependencies

Open Tracing

io.opentracing:opentracing-api: open tracing api definitions;
optional: io.opentracing.contrib:span-reporter: to define custom, combined reporters (to jaeger and Kensu);
io.opentracing.contrib:opentracing-tracerresolver: provides the mechanism to pick the right tracer;
io.opentracing.contrib:opentracing-jaxrs2: injects JAX-RS request with span and tags;
io.jaegertracing:jaeger-client: Jaeger backend implementation (could be made optional...).

Utilities

com.github.wnameless.json:json-flattener: from deep JSON to flattened one, used by src/main/java/io/kensu/collector/utils/json/KensuJsonSchemaInferrer.java.

Kensu

io.kensio.third.contrib:opentracing-jdbc: intercepts JDBC connection, statement, and results (added by Kensu) with dedicated spans (this is a fork in kensuio-oss);

Capabilities

Interceptors

src/main/java/io/kensu/collector/interceptors/DiscoverableSpanFinishingFilter.java: installs for all request the JAX-RS Filter from opentracing JAX-RS support, this injects the Span in the request header and catches the response HTTP code as span tag
src/main/java/io/kensu/collector/interceptors/KensuTracingInterceptorFeature.java: this decorates the span from all requests with additional tags like:
- http.request.url.path.pattern: sanitized URL pattern (without values);
- http.request.url.path.parameters: path parameters;
- http.request.url.query.parameters: query parameters.
src/main/java/io/kensu/collector/interceptors/ResponseInterceptor.java: intercepts the response to read the response JSON and injects these tags using src/main/java/io/kensu/collector/utils/json/KensuJsonSchemaInferrer.java:
- response.schema the schema of the JSON;
- response.stats the stats (e.g. min, max, count, na.count) of the JSON.

From (Opentracing/Kensu) JDBC

Intercepts JDBC Connection, PreparedStatement to inject dedicated spans with database, driver, SQLs and such information.

Kensu version add injection of ResultSet only supported from PreparedStatement ATM to compute stats while iterating over the ResultSet (e.g. count, min, na.count, ...);

What's being reported to Observability system (Kensu servers)

The span structure looks typically as follow:

> GET:/v1/order-details/product-line/{productLine}
  - http.request.url.path.pattern=/v1/order-details/product-line/{productLine}?maxResults={maxResults}
  - http.request.url.query.parameters=[maxResults]

    > Query 
      - db.instance=classicmodels
      - db.statement=select orderdetai0_.orderNumber as orderNum1_0_, orderdetai0_.orderLineNumber as orderLin2_0_...    

        > QueryResultStats
          - db.column.5.md={nullable=false, type=INT, name=quantityOrdered, schemaName=, tableName=orderdetails}
          - db.column.5.stats={mean=36.875, count=16.0, sum=590.0}
          - db.count=17

    > Query 
        > QueryResultStats
        
    > Query 
        > QueryResultStats

    > serialize
      - media.type=application/json
      - response.schema=[class FieldDef { name: [].quantityOrdered fieldType: number nullable: false }, class FieldDef...
      - response.stats={[].product.productName={count=17.0}, [].product.productVendor={count=17.0}, ...

Deployment

Optional: Install Jaeger (there's Docker)

The Docker is perfectly OK: https://www.jaegertracing.io/docs/1.21/getting-started/#all-in-one.

Optional: Install Wildfly (there's Docker)

Follow instructions on https://www.wildfly.org/.

Start it in standalone mode: ${wildfly.home}/bin/standalone.sh.

Maven

Install maven following instructions on https://maven.apache.org/.

As this is using private Kensu libraries, the $HOME/.m2/settings.xml of maven should be updated to point to the Kensu mirrors in Nexus, see below.

Also to allow automatic local deployments, you enable activate the copy-war profile and create yours with your path to wildfly.

Moreover, the access to Kensu API Url and Tokens are configured in app.properties, therefore one can configure a profile for each environment to be used (demo101, demo102, qa3, ...) by setting the variable appropriately. The the maven build command needs to be adapted to use the profile like mvn -P kensu-demo-102

Settings.xml

YOU NEED TO UPDATE IT WITH YOUR wilfly.home and PUT_TOKEN_HERE

<?xml version="1.0" encoding="UTF-8"?>
<settings xmlns="http://maven.apache.org/SETTINGS/1.1.0"
          xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
          xsi:schemaLocation="http://maven.apache.org/SETTINGS/1.1.0 http://maven.apache.org/xsd/settings-1.1.0.xsd">

    <servers>
        <server>
            <id>nexus</id>
            <username>$NEXUS_USER$</username>
            <password>$NEXUS_PASSWORD$</password>
        </server>
    </servers>

    <mirrors>
        <mirror>
            <id>nexus</id>
            <name>nexus</name>
            <url>https://public.usnek.com/n/repository/maven-public/</url>
            <mirrorOf>*</mirrorOf>
        </mirror>
    </mirrors>
  
    <profiles>
        <profile>
            <id>kensu-demo-102</id>
            <activeByDefault/>
            <properties>
                <kensu.api.url>https://api-demo102.usnek.com</kensu.api.url>
                <kensu.api.token>${PUT_TOKEN_HERE}</kensu.api.token>
            </properties>
        </profile>
        <profile>
            <id>demodb</id>
            <activeByDefault/>
            <properties>
                <kensu.demodb.password>${PUT_PASSWORD_HERE}</kensu.demodb.password>
                <kensu.demodb.uri>some-domain.com:3306/demodb</kensu.demodb.uri>
            </properties>
        </profile>
        <profile>
            <id>my-copy-war</id>
            <activeByDefault/>
            <properties>
                <path.to.jboss.deployments.dir>${wilfly.home}/standalone/deployments</path.to.jboss.deployments.dir>
            </properties>
        </profile>
    </profiles>
  
    <activeProfiles>
      <activeProfile>copy-war</activeProfile>
    </activeProfiles>

</settings>

Dependencies

Make sure you have the collectors' dependencies installed on your maven machine, or they are published in Nexus.

Build

To build, simply run: mvn clean package to create the package (war file).

Note in order to inject the properties from your profiles in settings.xml don't forget to add the -P to your command!

This will compile, run the unit tests, and create a war file that can be deployed into an JEE app server.

The war file:

will be created somewhere like target/sample-service-1.3.1.war (sensitive to artifactId and version changes in pom.xml...).
can be published to the wildfly standalone deployments folder:
- manually: cp target/sample-service-1.3.1.war ${wilxfly.home}/standalone/deployments
- automatically using a maven profile see Settings.xml

SSL (on kensu servers)

Install certificate to your java keystore:

java.home=/Users/andy/Library/Java/JavaVirtualMachines/openjdk-15.0.1/Contents/Home/
${java.home}bin/keytool -keystore ${java.home}lib/security/cacerts -noprompt -storepass changeit -importcert \
 -trustcacerts -alias kensu_import_vault_ca -file  ~/Downloads/kensuio_vault_ca.crt

Try it

Some examples:

Check on Jaeger: http://localhost:16686/search

Check on Kensu UI: data catalog, lineages, and projects.

Name		Name	Last commit message	Last commit date
Latest commit History 41 Commits
.github		.github
.settings		.settings
docker-dir		docker-dir
docs/screenshots		docs/screenshots
src		src
.classpath		.classpath
.gitignore		.gitignore
.project		.project
LICENSE		LICENSE
README.md		README.md
build-and-run-app.sh		build-and-run-app.sh
build-dependencies.sh		build-dependencies.sh
one-click-start-all-prebuilt.sh		one-click-start-all-prebuilt.sh
pom.xml		pom.xml
sample-service.iml		sample-service.iml

License

kensuio-oss/jaxrs-sample

Folders and files

Latest commit

History

Repository files navigation

JAX-RS Sample application for Wildfly enhanced with kensu monitoring capabilities

One-click run all in Docker

Check results: Zipkin & Kensu app

Build it yourself and run

Technical details

Databases

MySQL

Tutorial

Kensu demodb

Clickhouse

Tutorial database

Ported MySQL database

Application

Sources

Configuration

Databases

JBOSS env

Collector

Configuration

Tracer Bean

Jaeger & more

Dependencies

Open Tracing

Utilities

Kensu

Capabilities

Interceptors

From (Opentracing/Kensu) JDBC

What's being reported to Observability system (Kensu servers)

Deployment

Optional: Install Jaeger (there's Docker)

Optional: Install Wildfly (there's Docker)

Maven

Settings.xml

Dependencies

Build

SSL (on kensu servers)

Try it

About

Resources

License

Stars

Watchers

Forks

Releases

Packages 0

Contributors 5

Languages

Packages