- Artifact account
- Configuring the Spin CLI for managing spinnaker pipelines efficiently
- Upgrading spinnaker version
- Spinnaker on minikube
- Spinnaker on GKE
- Enabling email notification in Spinnaker
- Configuring GCP's IAP authentication mechanism on Spinnaker UI
- Jenkins crumb issue
- Spinnaker Artifacts helps to pull the objects from the external resources, which can be further used in spinnaker pipeline.
- It could be a docker image, manifest file stored in VCS or an AMI
- All the configuration of Spinnaker is solely managed by its "halyard" component, which is the centralized hub for managing spinnaker.
- For the artifact account configuration, make sure you either generated the authentication token or auth through the username and password,specially for bitbucket you will be needing password.
- Usecase: The github or any other VCS can be used to store the k8s manifest file which can be used further.
hal config features edit --artifacts true
# Enabling the github artifact account
hal config artifact github enable
# Listing the artifact account
hal config artifact github account list
ARTIFACT_ACCOUNT_NAME=sample-github-artifact-account
# For this configuration, token-needs to be generated either on Github or github
echo "Secret_token" > ~/TOKEN
TOKEN_FILE=/home/spinnaker/TOKEN
hal config artifact github account add $ARTIFACT_ACCOUNT_NAME --token-file $TOKEN_FILE
hal config artifact bitbucket account add sample-bitbucket-account --username-password-file $FILE_PATH
# Enabling the artifact feature
hal config features edit --artifacts true
# Enabling bitbucket artifact account
hal config artifact bitbucket enable
# Listing the bitbucket artifact accounts
hal config artifact bitbucket account list
# Setting the artifact-account-name and username
#ARTIFACT_ACCOUNT_NAME=sample-bitbucket-account
USER_NAME=ishaq4466
# Configuring the bitbucket account and listing
hal config artifact bitbucket account add sample-bit-account --username $USER_NAME --password
hal config artifact bitbucket account list
# Once configured, need to hit the "hal deploy" to take changes on spinnaker end
hal deploy apply
# deleting the the artifact account
hal config artifact bitbucket account delete $ARTIFACT_ACCOUNT_NAME
cloud_driver, echo, igor service will be restarted
If artifact account configure correctly you will the below image result on the spinnaker console
Api call for my reference:
curl --request GET --user ishaq4466:<password> https://api.bitbucket.org/2.0/repositories/ishaq4466/kubernetes/src/master/deployment/sample_deployment.yaml
*it happens sometime the artifact account doesn't appear, try to upgrade the spinnaker version and re-try to configure the artifact account
- CLI configuration
- Before installing "spin", we need to know the spinnaker's-gate svc endpoint, maker sure we have before-hand
wget https://storage.googleapis.com/spinnaker-artifacts/spin/$(curl -s https://storage.googleapis.com/spinnaker-artifacts/spin/latest)/linux/amd64/spin
chmod +x ./spin
mv ./spin /usr/local/bin/spin
mkdir ~/.spin/
# configuring the gate endpoint so that spin can talk with spinnaker through gate
cat<<EOF> ~/.spin/config
gate:
endpoint: http://<spinnaker-gate-endpoint>:8084
EOF
# Listing all the pipelines in Spinnaker within application
spin pipeline list --application <app-name>
# Exporting the JSON Spinnaker pipeline from an application
spin pipeline get --name <pipline-name> --application <app-name>
# Deleting a pipeline
spin pipeline delete <pipline-name> --application <app-name>
# Saving a spinnaker template as a pipeline
spin pipeline save --file template1.txt, configure and deploy
- Some useful-bash-hacks
# Spliting single json template to many json template with "//" as a delimeter
csplit --digits=2 --quiet --prefix=template ./allPipelines.txt "////+1" "{*}"
# Looping and saving the pipeline
for x in `ls template*`
do
echo $x
echo spin pipeline save --file $x
done
# Deleting the pipelines in one-loop
for x in $(cat pipeline_names.txt)
do
echo spin pipeline delete --name $x --application <app-name>
done
- Exec into the halyard pod
kubectl exec -it <spin-halyard-podname> bash
- Check the spinnaker version, select the version to deployed, configure and deploy
hal version list #checking the current version
version=1.17.5
hal config version edit --version $version
hal deploy apply
hal version list
Step 1. For running spinnaker minikube vm must atleast have 8Gi,4 cores of compute power for deploying Spinnaker smootly, without any trouble.
minikube start --vm-driver=virtualbox --kubernetes-version=1.16.0 --memory=8192 --cpus=4 --disk-size=50g
Step 2. Deploy the minikube spinnaker stack i.e spin-minikube.yaml once the minikube cluster-up running
kubectl create -f spin-minikube.yaml
Deploying spinnaker on minikube is patience keeping process, with the above least configuration
Step1. Create the "spin-gkestack.yaml" on GKE
Step2. Expose the spin-deck and spin-gate(if needed) as NodePort or LoadBalancer svc through "spin-svc.yaml"
Note: All the spinnaker micro-services are deployed to the spinnaker namespace though it could be changed
hal config provider kubernetes account edit ACCOUNT --add-omit-namespace default
Save the below configuration file in halyard pod followed by hal deploy apply
cat>>EOF<<-/home/spinnaker/.hal/default/profiles/settings-local.js
window.spinnakerSettings = window.spinnakerSettings || {};
window.spinnakerSettings.notifications = window.spinnakerSettings.notifications || {};
window.spinnakerSettings.notifications.email = window.spinnakerSettings.notifications.email || {};
window.spinnakerSettings.notifications.email.enabled = true;
EOF
cat>>EOF<<-/home/spinnaker/.hal/default/profiles/echo-local.yml
mail:
enabled: true
host: smtp.gmail.com
from: [email protected]
properties:
mail:
smtp:
auth: true
starttls:
enable: true
spring:
mail:
host: smtp.gmail.com
port: 587
username: [email protected]
password: Octa@123
properties:
mail:
smtp:
auth: true
starttls:
enable: true
EOF
