This repo contains code used in the Harness CD Community Edition which is licensed under the PolyForm Shield License 1.0.0. This repo also contains code belonging to Harness CD Enterprise Plan which is licensed under the PolyForm Free Trial License 1.0.0. You may obtain a copy of these licenses in the licenses directory at the root of this repository.

https://harness.atlassian.net/wiki/spaces/BT/pages/21101969665/How+to+setup+local+system+to+build+and+run+backend+code+harness-core+using+Bazel.

1. Install Homebrew:

/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

~~2. DEPRECATED STEPS DO NOT USE Download and Install Java 8

Note: Brew will download and install the latest version of OpenJDK/JRE, its recommended to install OpenJDK/JRE_1.8.0_242 to be in sync with the version everyone is using in the team.

To setup the recommended version, download the OpenJDK 1.8-242 (jdk8u242-b08) JRE .pkg from AdoptOpenJDK and install it. Make sure to update JAVA_HOME and PATH accordingly (see step 5).~~

2. Install JDK11 locally - 11.0.15 Temurin build

Download JDK 11 for MAC curl -OL https://github.com/adoptium/temurin11-binaries/releases/download/jdk-11.0.15%2B10/OpenJDK11U-jdk_x64_mac_hotspot_11.0.15_10.tar.gz

Untar it tar -xzf OpenJDK11U-jdk_x64_mac_hotspot_11.0.15_10.tar.gz

Update java home (Also update in .bashrc/.zshrc) export JAVA_HOME=path_to_folder/jdk-11.0.15+10/Contents/Home

Update path (Also update in .bashrc/.zshrc) export PATH=$JAVA_HOME/bin:$PATH

Check java version, should be 11 java --version

3. Install bazel:

brew install bazelisk

4. Install npm (used for front-end)

brew install npm

5. Update /etc/hosts to reflect your hostname

255.255.255.255	broadcasthost
127.0.0.1  <your hostname>
::1        <your hostname>

6. Download and install buf Complete this step only if you are actively working with the protocol buffer files.

brew tap bufbuild/buf
brew install buf

To check if your protobuf files are according to the coding standards execute in the root of the repo

buf lint

7. Install Docker

Official steps to install docker on mac: docker.com.

8. If you are on MacOS make sure you have Xcode installed

1. Setup your harness email to the git config. You can do it globally or just for the harness-core repo:

   git config --global user.email "[email protected]"

   or just for harness-core with

   git config user.email "[email protected]"

2. Setup your name to the git config. We are using First and Last name. Please make sure you use the same spelling as you did for your github account.

   git config --global user.name "FirstName LastName"

   or just for harness-core with

   git config user.name "FirstName LastName"

3. Install git hooks. harness-core comes with a set of convenient productivity booster set of hooks. For security reasons they cannot be enabled automatically. To do so execute the following command from the root of already cloned locally repo:

   toolset/git-hooks/install.sh

   NOTE: if you clone the repo to another location you will have to do this again. On the other side you will be getting fixes and updates with no extra effort.

https://harness.atlassian.net/wiki/spaces/BT/pages/1910047082/Bazel+best+practices+for+faster+and+efficient+builds.

1. Clone form git repository: https://github.com/harness/harness-core.git

   (Optional) Follow https://help.github.com/articles/adding-a-new-ssh-key-to-your-github-account/ to setup your SSH keys. You can then use SSH to interact with git

   NOTE: If you have regular bazel installed, please uninstall bazel and install bazelisk. It allows us to use the git repo to synchronize everyone's installation of bazel.

2. Setup the build purpose

   You need to set environment variable BUILD_PURPOSE with one of the following values:

   • DEVELOPMENT - set this when you building for development purposes
   • PR_CHECK - set this when you building for executing check for the pr
   • FEATURE - set this when you building for internal feature testing
   • RELEASE - set this when you building for release

   if you seting up the project to build locally, you should simply add to your .bash_profile file:

   export BUILD_PURPOSE=DEVELOPMENT
   

3. Go to harness-core directory and run

   scripts/bazel/generate_credentials.sh
   bazel build //... or bazel build :all
   

4. If Global Search is not required:

   Install and start MongoDB Docker Image (v4.2):

   docker run -p 127.0.0.1:27017:27017 -v ~/_mongodb_data:/data/db --name mongoContainer -d --rm mongo:4.2
   

   Verify the container is running using docker ps

   Install & use RoboMongo client to test MongoDB connection.

5. If Global search has to be enabled (OPTIONAL):

   Install and start Elasticsearch Docker Image for Search(v7.3):

   docker run -p 9200:9200 -p 9300:9300 -v ~/_elasticsearch_data:/usr/share/elasticsearch/data -e "discovery.type=single-node" docker.elastic.co/elasticsearch/elasticsearch:7.3.0
   

   In harness-core/360-cg-manager/config.yml set searchEnabled to true.

   Run mongo in replica set:

   docker-compose -f <Directory to harness-core>/harness-core/docker-files/mongo-replicaset/docker-compose.yml up -d
   

   Add this to /etc/hosts:

   127.0.0.1       mongo1
   127.0.0.1       mongo2
   127.0.0.1       mongo3
   

   Run brew tap mongodb/brew Run brew install [email protected]

   Run mongo --port 30001

   Run these in the mongo console:

   rs.initiate()
   rs.add('mongo2:30002')
   rs.add('mongo3:30003')
   

   In 360-cg-manager/config.yml set mongo.uri to mongodb:https://mongo1:30001,mongo2:30002,mongo3:30003/harness. Do the same in config-datagen.yml and verification-config.yml.

6. If TimeScaleDB has to be enabled (Optional for now)

   a. Start TimeScaleDB using the following docker command: docker run -d --name harness-timescaledb -v ~/timescaledb/data:/var/lib/postgresql/data -p 5432:5432 --rm -e POSTGRES_USER=admin -e POSTGRES_DB=harness -e POSTGRES_PASSWORD=password timescale/timescaledb:latest-pg14

   b. Set the TimeScaleDB config in the config.yml

timescaledb:
  timescaledbUrl: jdbc:postgresql:https://localhost:5432/harness
  timescaledbUsername: admin
  timescaledbPassword: password

7. Install Redis - Follow the instructions from here

cd to harness-core directory

1. Start server by running following commands :

   • bazel build //360-cg-manager:module_deploy.jar
   • curl https://storage.googleapis.com/harness-prod-public/public/shared/tools/alpn/release/8.1.13.v20181017/alpn-boot-8.1.13.v20181017.jar --output alpn-boot-8.1.13.v20181017.jar [NOTE: alpn.jar will get download as part of generate_credentials.sh script as well.]
   • java -Xms1024m -Xmx4096m -XX:+HeapDumpOnOutOfMemoryError -XX:+PrintGCDetails -XX:+PrintGCDateStamps -Xloggc:mygclogfilename.gc -XX:+UseParallelGC -XX:MaxGCPauseMillis=500 -Xbootclasspath/p:alpn-boot-8.1.13.v20181017.jar -Dfile.encoding=UTF-8 -jar ~/.bazel-dirs/bin/360-cg-manager/module_deploy.jar server 360-cg-manager/config.yml > harness-core.log &

2. Generate sample data required to run the services locally by running the following step only once. DataGenUtil: Open a new terminal and run following command (Make sure you setup HARNESS_GENERATION_PASSPHRASE environment variable in your Bash profile):

   • java -Xmx1024m -jar ~/.bazel-dirs/bin/160-model-gen-tool/module_deploy.jar server 160-model-gen-tool/config-datagen.yml

   or, preferably, with this command from bash console:

   • bazel run 160-model-gen-tool:module --jvmopt="-Xbootclasspath/p:alpn-boot-8.1.13.v20181017.jar" server <harness-core project location>/160-model-gen-tool/config-datagen.yml

   After command finishes, you might confirm in the mongodb account table that accountKey value is properly set.

3. Start Delegate

   • java -Xmx1536m -XX:+HeapDumpOnOutOfMemoryError -XX:+PrintGCDetails -XX:+PrintGCDateStamps -Xloggc:mygclogfilename.gc -XX:+UseParallelGC -XX:MaxGCPauseMillis=500 -jar ~/.bazel-dirs/bin/260-delegate/module_deploy.jar 260-delegate/config-delegate.yml &

4. Start Verification service (Optional)

   • java -Xms1024m -Xmx4096m -XX:+HeapDumpOnOutOfMemoryError -XX:+PrintGCDetails -XX:+PrintGCDateStamps -Xloggc:mygclogfilename.gc -XX:+UseParallelGC -XX:MaxGCPauseMillis=500 -Xbootclasspath/p:alpn-boot-8.1.13.v20181017.jar -Dfile.encoding=UTF-8 -jar ~/.bazel-dirs/bin/270-verification/module_deploy.jar server 79-verification/verification-config.yml > verification.log &

5. Start Command Library Service (Optional)

   • java -Xms1024m -Xmx4096m -XX:+HeapDumpOnOutOfMemoryError -XX:+PrintGCDetails -XX:+PrintGCDateStamps -Xloggc:mygclogfilename.gc -XX:+UseParallelGC -XX:MaxGCPauseMillis=500 -Xbootclasspath/p:alpn-boot-8.1.13.v20181017.jar -Dfile.encoding=UTF-8 -jar ~/.bazel-dirs/bin/210-command-library-server/module_deploy.jar 210-command-library-server/command-library-server-config.yml > command_library_service.log &

6. Start UI (Optional)

• Create a shell script which pulls harness ui docker image and starts ui app. Name it e.g. startui.sh and replace and with the harness docker user credentials, found in Vault.

run_ui ()
{
    echo '<DOCKERHUBPASS>' | sudo docker login -u <DOCKERHUBUSER> --password-stdin;
    if [ ! -z "$1" ]; then
        tag=":$1";
    fi;
    sudo docker pull harness/ui$tag;
    sudo docker run -it -p 8000:8080 --rm -e API_URL=https://localhost:9090 harness/ui$tag
}  

alias runui='run_ui'

• Add following line to ~/.bashrc: source <path-to-startui-script>/startui.sh
• Execute .bashrc: source ~/.bashrc
• Start ui by simply entering: runui
• UI url will be found in logs.

More info on this can be found here.

1. Install clang-format (11.0.0) If on MacOS you can download this via brew brew install clang-format@11

else

Download the clang 11.0.0 tar from this page Untar the downloaded file and add it to your PATH in ~/.bashrc or ~/.zshrc

echo "export PATH="$PATH:$HOME/<path-to-above-directory>/bin" >> ~/.zshrc

to format .graphql files: you can follow these steps:

• npm install --global [email protected]

• prettier --write --print-width=120 <filename> - formats given graphql file

helper shell scripts:

• git clang-format - makes sure all staged in git files are reformatted

• find . -iname *.java | xargs clang-format -i - formats all java files from the current directory down

1. Install IntelliJ

   NOTE: Download IntelliJ as per your macOS architecture i.e. x86_64 (intel) or arm64 (Apple silicon). You can check your architecture by command “uname -m“.

   It's recommended to install the latest version of IntelliJ that's supported by the Bazel Plugin. To find the latest supported version, please refer to the bazel github project. As of June, 2022 recommended version of IntelliJ is 2021.3.3 After identifying the desired version of IntelliJ, you can download it from jetbrains.com.

2. Install bazel project plugin from the IntelliJ marketplace

3. Import harness-core as a Bazel project a. Open File > Import Bazel Project.... b. Enter /path/to/repo/harness-core for Workspace, click Next. c. Select Import project view file and enter project/bazelproject as the Project view. NOTE: If you do not see this file project/bazelproject in the list then select option Create from scratch and paste the content of project/bazelproject in the text area.

4. Install ClangFormatIJ Plugin: https://plugins.jetbrains.com/plugin/8396-clangformatij (use Ctrl/Cmd-Alt-K to format current statement or the selection)

   WARNING: For unclear reason in some environments the plugin causes IntelliJ to hang. If you are unlucky to be one of those cases there is alternative. Please use the external 3rd-party tool integration as described here: https://www.jetbrains.com/help/idea/configuring-third-party-tools.html. Configure the tool to look like shown on the image:

   .

   Then follow these instructions https://www.jetbrains.com/help/idea/configuring-keyboard-shortcuts.html to assign whatever key combination you would like it to be triggered on.

5. Install Lombok Plugin: https://projectlombok.org/setup/intellij

   Not required for IntelliJ 2020.3 and later versions - the Lombok Plugin comes pre-bundled with IntelliJ

6. Install SonarLint plugin:

   • This plugin is really helpful to analyze your code for issues as you code.

   • Go to Preferences -> Plugins -> type SonarLint -> Install plugin. (Will need to restart Intellij)

   • Go to Preferences -> Tools -> SonarLint.

     Check "Automatically trigger analysis".

     Add a SonarQube connection to https://sonar.harness.io (requires VPN connection). As part of the connection-setup you'll be directed to sonar.harness.io to create a custom security token.

   • Go to Preferences -> Tools -> SonarLint -> Project Settings. Check "Bind project to sonarqube", and select the connection, and set project as portal_bazel. This is so that we use the same rules locally instead of the default rules. .

   • Go to Preferences -> Editor -> Colorscheme -> Sonarlint. For Blocker, Critical & Major, untick "Inherit values from" checkbox and configure a different highlighting style. These violations are treated as release blockers and this configuration is to highlight them differently from regular warnings. .

   • Just right click on file in intellij and "Analyze with SonarLint" or enable autoscan.

7. Install the Checkstyle-Idea Plugin.

   1. Run Maven build of the tools directory

      mvn -f tools/ clean install -DskipTests
      

   2. Setup Checkstyle plugin. In Preferences -> Tools -> Checkstyle add tools/config/target/config-0.0.1-SNAPSHOT-jar-with-dependencies.jar and tools/checkstyle/target/checkstyle-0.0.1-SNAPSHOT.jar jars in the repo to the 3rd party checks classpath. Add configuration file harness_checks.xml (Choose the option to resolve the file from the 3rd party checks classpath - it's within the config jar) and choose it as the default active. Set scan scope to java sources including tests. In case Intellij complains about missing Harness rule files add following jar to Third-Party Checks tools/checkstyle/target/checkstyle-0.0.1-SNAPSHOT.jar. Additionally, check version of Checkstyle plugin to be 8.20 Preferences > Tools > Checkstyle > Checkstyle Version:

   • .
   • .

8. Change settings to mark injected fields as assigned. (Preferences -> Editor -> Inspections -> Java -> Declaration Redundancy -> Unused Declarations -> Entry Points -> Annotations -> Mark field as implicitly written if annotated by) Click add, then search for "Inject". Add both google and javax annotations.

   • .

9. Increase Build Process Heap Size (Preferences > Build, Execution, Development > Compiler, search for "Build Process Heap Size" and set it to 2048 or higher if you still see an out of memory exception in future)

10. If facing build issues make sure you have enabled "Always update snapshots" in IntelliJ (Preferences > Build, Execution, Deployment > Build Tools > Maven)

Run configurations for the different applications are already checked into the repo. Choose the appropriate run configuration from the menu. While running an app from pre checked in configs, Add JAVA_HOME as an environment variable in Intellij.

If you are using zsh (which is default on MacOS Catalina and later), basic git integration comes out of the box.

If you are using bash, add the following to your ~/.bash_profile to display the current git branch in the command prompt:

parse_git_branch() {
  git branch 2 > / dev / null | sed - e '/^[^*]/d' - e 's/* \(.*\)/ (\1)/'
}
export PS1="\[\033[34m\]\w\[\033[36m\]\$(parse_git_branch)\[\033[31m\] $\[\033[0m\] "

Alternatively, use Fish shell: brew install fish then set iterms command to /usr/local/bin/fish

1. Make sure your mongodb is running first.

2. Run API Server (WingsApplication): [Run > Run... > WingsApplication]

   • If you get ALPN processor missing at start of WingsApp execute following curl command

     curl https://storage.googleapis.com/harness-prod-public/public/shared/tools/alpn/release/8.1.13.v20181017/alpn-boot-8.1.13.v20181017.jar --output alpn-boot-8.1.13.v20181017.jar

3. Run DataGenApp: [Run > Run... > DataGenApp]. Add HARNESS_GENERATION_PASSPHRASE environment variable to DataGenApp config in intellij.

4. Run DelegateApplication: [Run > Run... > DelegateApplication]

The admin username and password are in BaseIntegrationTest.java.

1. To build UI Go to wingsui and follow READ me instructions.

2. To apply database migrations run following command in dbmigrations folder:

   mvn clean compile exec:java

• If you get an error about missing build.properties when you start the server, do a mvn clean install.

• If you get a SupportedEllipticCurvesExtension NoClassDefFoundError, Its likely that jsse.jar in /Library/Java/JavaVirtualMachines//Contents/Home/jre/lib folder does not have this class definition. Copy this file from a Team member.

  • If you have jsse.jar but still getting that error, then make sure the default JDK for your maven module is set correctly in IntelliJ. Right Click Module in left sidebar > Open Module Settings > Platform Settings > SDKs)

• If you go to https://localhost:8000/#/login and don't see content, go to https://localhost:8181/#/login to enable the certificate then try again.

• If still face not able to login then got to https://localhost:9090/api/version and enable certificate and try again.

• If you get ALPN processor missing at start of WingsApp execute following curl command

  curl https://storage.googleapis.com/harness-prod-public/public/shared/tools/alpn/release/8.1.13.v20181017/alpn-boot-8.1.13.v20181017.jar --output alpn-boot-8.1.13.v20181017.jar

• Refer to the readme under python/splunk_intelligence

1. Install Go 1.17.8 here
2. Add this to .bash_profile: export PATH=$PATH:~/go/bin
3. Install dependent tools by running:

harness-core/tools/go/go_setup.sh

4. On mac: brew install bazelisk
   • Other platforms: Follow the instrictions here

Jetbrains has GoLand editor but its not free. If we continue using intelliJ for Go, then we need a plugin called Go, but it’s not supported on Community Edition(free) of intelliJ.

• So recommendation is to use VsCode(free) which is better than intelliJ for Go development.
• Once you install VsCode, open and install the plugin Microsoft Go
• Note: If your autocomplete is not working, disable gopls

Alternately, gopls has a langserver implementation which hooks up cleanly to IntelliJ as LSP client.

This page contains the most common commands and recommendations, for more details go to bazel homepage. Bazel has extensive documentation available online.

You can start with the user guide and best practices

Documentation about the Go ruleset is available on github as part of bazelbuild/rules_go, most importantly go through the set of core rules for Go here.

You should use the bazel build command to build a project:

bazel build //path/to/project/...

For more information about build target pattern syntax run:

bazel help target-syntax

Examples:

bazel build //commons/go/...

Note that // stands for the repo root, if you run command from the repo root then you may omit it and instead run:

bazel build commons/go/...

bazel build //...

bazel build

bazel build ...

bazel build foo:all

For additional information, run:

bazel help target-syntax

You can cross-compile any go_binary target on demand to a specific platform by running:

bazel build --platforms=@io_bazel_rules_go//go/toolchain:linux_amd64 //cmd

For details see the Go rules for Bazel.

Tests can be executed using the bazel test command:

bazel test //path/to/project/... # will run all tests under //path/to/project
bazel test //path/to/project:target # will run the test named "target" under //path/to/project

Note that test targets tagged as manual are skipped in ... if not specified explicitly.

Examples:

bazel test //commons/go/lib/...:all --test_summary=detailed  # will run all tests under lib

bazel test //...

bazel test

bazel test //commons/go/lib/logs:go_default_test

bazel test ...

bazel test foo:

bazel run //commons/go/lib/<module>:"internal_tests" -- -test.bench=<keyword identifying the resolver> -test.benchmem
bazel run //commons/go/lib/logs:go_default_test -- -test.bench=harness -test.benchmem # an example

bazel coverage <target pattern>
bazel coverage //commons/go/lib/logs:go_default_test # an example

Bazel allows running executable targets using:

bazel run //path/to/target:target
bazel run //commons/go/lib/logs:go_default_test # an example

1. brew install --cask google-cloud-sdk
2. Add gcloud to your PATH
   1. Either, manually follow the onscreen instructions from brew
   2. Or, run the SDKs installer /usr/local/Caskroom/google-cloud-sdk/latest/google-cloud-sdk/install.sh
3. Run gcloud init to configure your installation
4. Please select platform project in GCP during the configuration.
5. Once all configurations done then you should be able to pull images from gcr registry.

We have added flexibilities of building docker images with bazel.
Docker rule reference: https://github.com/bazelbuild/rules_docker.
To build docker images through bazel locally(i.e. access private images, push etc) we need to configure gcloud auth for docker. You can run these commands to configure it locally:

gcloud components install docker-credential-gcr
gcloud auth login
gcloud auth configure-docker

BUILD.bazel files contain build rules. If you've added/removed packages or modified dependencies in the source code, or added new rules manually, then you should run gazelle to update and format your BUILD.bazel files. This tool will add any missing rules, update dependencies and format all BUILD.bazel files that you've touched. Run in the root of the harness-core repository:

gazelle

The above comand creates or updates BUILD.bazel files where needed. If the above command fails, it is likely due to using an incorrect version of gazelle. Currently we are using version 0.21.

The rough overview is here, if you want more complete instructions go to the bazelbuild github page https://github.com/bazelbuild/bazel-gazelle.

git clone [email protected]:bazelbuild/bazel-gazelle.git
cd bazel-gazelle
git reset origin/release-0.21 --hard
cd cmd/gazelle
bazelisk build gazelle
$(bazelisk info bazel-bin)/cmd/gazelle/gazelle_/gazelle
# it expands out to something like below, giving the 0.21 binary
/home/tp/.cache/bazel/_bazel_tp/46ccc68b31f8c833946cfcd24410eb45/execroot/bazel_gazelle/bazel-out/k8-fastbuild/bin/cmd/gazelle/gazelle_/gazelle

This can now be used in harness-core to fix dependencies.

We need to update the dependencies in harness-core/WORKSPACE. Run the following for your new/updated go.mod

update_bazel_repo.sh commons/go/lib/go.mod # an example

This updates the harness-core/WORKSPACE file with new dependencies. Check-in harness-core/WORKSPACE file and any updated go.mod and go.sum files.

• We are using bazel for building and testing go source code. bazel needs BUILD.bazel files and doesn't need go.mod or go.sum files go.mod and go.sum files are needed for native go tools

• bazel uses $PROJECT_ROOT/WORKSPACE to store go repositories which has information found in both go.mod and go.sum bazel uses $PROJECT_ROOT/WORKSPACE for dependency management and BUILD.bazel for actual execution of targets

• To use local modules in an application

• for eg., to use module lib in an application ci-addon, we need to perform these actions:

cd ci/addon #cd to the application folder where main.go is located

go mod init <module-name>   # this generates go.mod and go.sum

• update go.mod by adding the following line to point to local repository (replace module and relative path to match yours):

replace github.com/harness/harness-core/commons/go/lib => ../../../commons/go/lib

• (the above replace is needed only if you are outside the module you want to import)

go get  # this updates go.mod

gazelle  # generates, updates BUILD.bazel

• To update go_repository() at harness-core/WORKSPACE, run this script as:

harness-core/tools/go/update_bazel_repo.sh go.mod

• Using GOPATH

Some editors like vscode requires the code to be structured in a GOPATH directory tree. In this case, you may use the :gopath target to generate a directory tree with symlinks to the original source files.

bazelisk build :gopath

You can then set the GOPATH in your editor to bazel-bin/gopath. If bazel-bin folder is not present in the root folder, use $(bazel info bazel-bin)/gopath.

NOTE: Below changes are only recommended in local environment and changes shall not be pushed.

AWS SDK library internal logging is done using SLF4J. SLF4J serves as a simple facade or abstraction for various logging frameworks (e.g. java.util.logging, logback, log4j)

We are already using logback framework in our application, so it is simple to enable logging as it is already supported in SLF4J.

• Delegate - To enable AWS SDK logging in delegate, update root logger level to TRACE in logback.xml file in 260-delegate module resources folder and restart delegate.
• Manager - To enable AWS SDK logging in manager, update root logger level to TRACE in logback.xml file in 360-cg-manager module resources folder and restart manager.

1. Create a new JIRA ticket(Don't use the feature branch).
2. Checkout the branch to hotfix on and make the changes.
3. Bump up the build number in the build.properties.
4. Create a pull request on this branch and master as well, make sure you merge them on master within the next 3 4 hours.