add image volume source

reliability-v2/config/reliability-ivs.yaml

@@ -0,0 +1,223 @@
+reliability:
+  kubeconfig: <path_to_kubeconfig>
+  users:
+    - admin_file: <path_to_admin_file>
+    - user_file: <path_to_users_file>
+
+  appTemplates:
+    - template: cakephp-mysql-persistent
+    - template: nodejs-postgresql-persistent # N/A for FIPS
+    - template: django-psql-persistent # N/A for FIPS
+    - template: rails-pgsql-persistent
+    #- template: dancer-mysql-persistent
+    # For test without pvc
+    # - template: cakephp-mysql-example
+    # - template: dancer-mysql-example
+    # - template: django-psql-example
+    # - template: nodejs-postgresql-example
+
+  # The max number of projects is limited by the size of cluster.
+  # For 3 nodes m5.xlarge cluster, 25 to 30 projects are recomended.
+  # For 5 nodes m5.xlarge cluster, 60 projects are recomended.
+  # The max number of projects = groups x users x new_project
+  groups:
+    - name: admin
+      # For cluster created by Jenkins Flexy-install job, the admin user_name is kubeadmin.
+      # For rosa cluster created in Prow, the admin user_name is rosa-admin.
+      user_name: <admin_username> # admin username as kubeadmin or rosa-admin.
+      loops: forever # run group for loops times. integer > 0 or 'forever', default is 1.
+      trigger: 600 # wait trigger seconds between each loop
+      jitter: 60 # randomly start the users in this group in trigger seconds. Default is 0.
+      interval: 10 # wait interval seconds between tasks.
+      # If admin network policy function is planed in the test, uncomment the following line
+      pre_tasks:
+        - func apply_nonamespace -p "<path_to_content>/network-policy/00_banp_deny-triffic-reliability.yaml"
+        - func apply_nonamespace -p "<path_to_content>/network-policy/01_anp_allow-common-service.yaml"
+        - func apply_nonamespace -p "<path_to_content>/network-policy/02_anp_allow-traffic-egress-cidr-cluster-network-p20.yaml"
+        - func apply_nonamespace -p "<path_to_content>/network-policy/04_anp_allow-traffic-dev-test-p30.yaml"
+        - func apply_nonamespace -p "<path_to_content>/network-policy/03_anp_allow-traffic-dev-prod-p31.yaml"
+      tasks:
+        - func check_operators
+        - oc get project -l purpose=reliability
+        - func check_nodes
+        - kubectl get pods -A -o wide | egrep -v "Completed|Running"
+        # Run test case as scripts. KUBECONFIG of the current user is set as env variable by reliability-v2. 
+        #- . <path_to_script>/create-delete-pod-ensure-service.sh
+
+    - name: dev-test
+      user_name: testuser- # if user_start and user_end exist, this will be username prefix
+      # For cluster created by Jenkins Flexy-install job, the users start from testuser-0
+      # For cluster created in Prow and used the following step to create test users, the users start from testuser-1. 
+      # https://github.com/openshift/release/blob/master/ci-operator/step-registry/osd-ccs/conf/idp/htpasswd/multi-users/osd-ccs-conf-idp-htpasswd-multi-users-ref.yaml
+      user_start: 1 # user_start is inclusive, start with testuser-1 in the users file. 
+      user_end: 6 # user_end is exclusive, end with testuser-10 in the users file
+      loops: forever
+      trigger: 60
+      jitter: 600 # randomly start the users in this group in 10 minutes
+      interval: 10
+      tasks:
+        - func delete_all_projects # clear all projects
+        - func verify_project_deletion -n 2 # verfy project deletion in 2 namespaces
+        - func new_project -n 2 # new 2 projects
+        # If network policy is planed in the test, uncomment the following line
+        #- func apply -n 2 -p "<path_to_content>/network-policy/allow-same-namespace.yaml" # Apply network policy to 2 projects
+        - func check_all_projects # check all project under this user
+        - func apply -n 2 -p "<path_to_content>/nginx-deploy.yaml"
+        - func apply -n 2 -p "<path_to_content>/nginx-service.yaml"
+        - func apply -n 2 -p "<path_to_content>/route-http.yaml"
+        - func load_app -n 2 -p 10 # load apps in 2 namespaces with 10 clients for each
+        - func check_pods -n 2 # check pods in 2 namespaces 
+        - func delete_project -n 2 # delete project in 2 namespaces
+        - func verify_project_deletion -n 2 # verfy project deletion in 2 namespaces
+
+    - name: dev-test-build
+      user_name: testuser- # if user_start and user_end exist, this will be username prefix
+      # For cluster created by Jenkins Flexy-install job, the users start from testuser-0
+      # For cluster created in Prow and used the following step to create test users, the users start from testuser-1. 
+      # https://github.com/openshift/release/blob/master/ci-operator/step-registry/osd-ccs/conf/idp/htpasswd/multi-users/osd-ccs-conf-idp-htpasswd-multi-users-ref.yaml
+      user_start: 21 # user_start is inclusive, start with testuser-1 in the users file. 
+      user_end: 26 # user_end is exclusive, end with testuser-10 in the users file
+      loops: forever
+      trigger: 60
+      jitter: 600 # randomly start the users in this group in 10 minutes
+      interval: 10
+      tasks:
+        - func delete_all_projects # clear all projects
+        - func verify_project_deletion -n 2 # verfy project deletion in 2 namespaces
+        - func new_project -n 2 # new 2 projects
+        # If network policy is planed in the test, uncomment the following line
+        #- func apply -n 2 -p "<path_to_content>/network-policy/allow-same-namespace.yaml" # Apply network policy to 2 projects
+        - func check_all_projects # check all project under this user
+        - func new_app -n 2 # new app in 2 namespaces
+        - func load_app -n 2 -p 10 # load apps in 2 namespaces with 10 clients for each
+        - func build -n 1 # build app in 1 namespace
+        - func check_pods -n 2 # check pods in 2 namespaces 
+        - func delete_project -n 2 # delete project in 2 namespaces
+        - func verify_project_deletion -n 2 # verfy project deletion in 2 namespaces
+
+    - name: dev-prod
+      user_name: testuser-
+      user_start: 31
+      user_end: 36
+      loops: forever
+      trigger: 600
+      jitter: 3600
+      interval: 600
+      pre_tasks:
+        - func delete_all_projects
+        - func verify_project_deletion -n 8
+        - func new_project -n 8
+        # If network policy is planed in the test, uncomment the following line
+        #- func apply -n 2 -p "<path_to_content>/network-policy/allow-same-namespace.yaml" # Apply network policy to 2 projects
+        - func apply -n 8 -p "<path_to_content>/nginx-deploy.yaml"
+        - func apply -n 8 -p "<path_to_content>/nginx-service.yaml"
+        - func apply -n 8 -p "<path_to_content>/route-http.yaml"
+      tasks:
+        - func load_app -n 8 -p 50
+        - func scale_deployment -n 8 -p 2 # scale deployment in 2 namespaces to 2 replicas
+        - func scale_deployment -n 8 -p 1 # scale deployment in 2 namespaces to 1 replicas
+      post_tasks:
+        - func delete_project -n 8
+
+    - name: dev-cronjob
+      user_name: testuser- # if user_start and user_end exist, this will be username prefix
+      user_start: 41
+      user_end: 42
+      trigger: 600
+      jitter: 1200
+      loops: forever
+      pre_tasks:
+        - func delete_all_projects
+        - func verify_project_deletion -n 1
+        - func new_project -n 1
+        - <path_to_script>/cronjob.sh -n 10
+      tasks:
+        - <path_to_script>/cronjob.sh -c
+      post_tasks:
+        - <path_to_script>/cronjob.sh -d
+        - func delete_project -n 1
+        - func verify_project_deletion -n 1
+
+    - name: ivs-llama
+      user_name: <admin_username>
+      trigger: 3600
+      jitter: 600
+      interval: 10
+      loops: forever
+      pre_tasks:
+        - oc new-project ivs-llama
+        - oc label ns ivs-llama purpose=reliability
+        - oc create sa llama-sa -n ivs-llama
+        - oc adm policy add-scc-to-user anyuid -z llama-sa -n ivs-llama
+        - oc apply -f "<path_to_content>/ivs/ivs-ai-llama.yaml" -n ivs-llamacpp
+        - sleep 300
+        - oc get po -n ivs-llama
+      tasks:
+        - oc delete po llama-llama3-8b-instruct-bartowski-q5km-0-1 -n ivs-llama
+        - oc get po -n ivs-llama
+        - sleep 60
+        - oc get po -n ivs-llama
+
+    - name: ivs-nginx
+      user_name: <admin_username>
+      trigger: 600
+      jitter: 600
+      interval: 10
+      loops: forever
+      pre_tasks:
+        - oc new-project ivs-nginx
+        - oc label ns ivs-nginx purpose=reliability
+      tasks:
+        - oc apply -f "<path_to_content>/ivs/ivs-nginx.yaml" -n ivs-nginx
+        - sleep 60
+        - oc get po -n ivs-nginx
+        - oc delete ivs leaderworkerset-sample -n ivs-nginx
+
+  cerberusIntegration:
+    # start cerberus https://github.com/cloud-bulldozer/cerberus before starting reliabiity test.
+    cerberus_enable: False
+    # if cerberus_enable is false, the following 2 items are ignored.
+    cerberus_api: "http://0.0.0.0:8080"
+    # action to take when cerberus status is False, valid data: pause/halt/continue
+    cerberus_fail_action: pause
+
+  slackIntegration:
+    slack_enable: False
+    # the ID in the example is the id of slack channel #ocp-qe-reliability-monitoring.
+    slack_channel: C0266JJ4XM5
+    # slack_member is optional. If provided, the notification message will @ you. 
+    # you must be a member of the slack channel to receive the notification.
+    slack_member: <Your slack member id>
+
+  krakenIntegration:
+    kraken_enable: False
+    # supported Kraken scenarios: https://github.com/cloud-bulldozer/kraken-hub/blob/main/README.md#supported-chaos-scenarios
+    # pod-scenarios, container-scenarios, node-scenarios, zone-outages, time-scenarios, 
+    # node-cpu-hog, node-memory-hog, node-io-hog
+    # Please specify the parameters for each scenario. e.g. For pod-scenarios,
+    # refer to https://github.com/cloud-bulldozer/kraken-hub/blob/main/docs/pod-scenarios.md#supported-parameters
+    kraken_scenarios: 
+      - name: pod-scenarios_etcd
+        scenario: "pod-scenarios"
+        interval_unit: minutes # weeks,days,hours,minutes,seconds
+        interval_number: 8
+        # start_date: "2021-10-28 17:36:00" # Optional. format: 2021-10-20 10:00:00.
+        # end_date: "2021-10-28 17:50:00" # Optional.
+        # timezone: "Asia/Shanghai" # Optional. e.g. US/Eastern. Default is "UTC
+      - name: pod-scenarios_monitoring
+        scenario: "pod-scenarios"
+        interval_unit: minutes # weeks,days,hours,minutes,seconds
+        interval_number: 10
+        parameters:
+          NAMESPACE: openshift-monitoring
+          POD_LABEL: app.kubernetes.io/component=prometheus
+          EXPECTED_POD_COUNT: 2
+      - name: node-scenarios_workerstopstart
+        scenario: "node-scenarios"
+        interval_unit: minutes
+        interval_number: 12
+        parameters:
+          AWS_DEFAULT_REGION: us-east-2
+          AWS_ACCESS_KEY_ID: xxxx
+          AWS_SECRET_ACCESS_KEY: xxxx
+          CLOUD_TYPE: aws

reliability-v2/content/ivs/ivs-ai-llama.yaml

@@ -0,0 +1,27 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: ivs-llama
+spec:
+  replicas: 1
+  selector:
+    matchLabels:
+      app: ivs-llama
+  template:
+    metadata:
+      labels:
+        app: ivs-llama
+    spec:
+      containers:
+      - name: llamacpp
+        image: quay.io/ai/llamacpp:latest
+        args: ["--model", "/models/llama.bin"]
+        volumeMounts:
+        - name: volume
+          mountPath: /models
+          subPath: dir
+      volumes:
+      - name: volume
+        image:
+          reference: quay.io/crio/artifact:v1
+          pullPolicy: IfNotPresent

reliability-v2/content/ivs/ivs-nginx.yaml

@@ -0,0 +1,26 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: ivs-nginx
+spec:
+  replicas: 3
+  selector:
+    matchLabels:
+      app: ivs-nginx
+  template:
+    metadata:
+      labels:
+        app: ivs-nginx
+    spec:
+      containers:
+      - name: nginx
+        image: quay.io/bitnami/nginx:latest
+        volumeMounts:
+        - name: volume
+          mountPath: /usr/share/nginx/html/models
+          subPath: dir
+      volumes:
+      - name: volume
+        image:
+          reference: quay.io/crio/artifact:v1
+          pullPolicy: IfNotPresent