Commit 85137d22 authored by syasonik's avatar syasonik

Add QA test for prometheus alerts

The configuration of alerts depends on a lot of moving parts and has
been prone to issue. This test sets up a cluster, creates a project,
connects the cluster to the project, sets up Auto Devops, deploys the
application, then performs basic CRUD operations on alerts from the
metrics dashboard. Alerts are configurated by updating Prometheus, so
a future test could build upon the assertions by ensuring that the alert
can indeed be fired.
parent 3b9944fc
......@@ -311,7 +311,10 @@ export default {
<gl-tooltip :target="() => $refs.graphTitle" :disabled="!showTitleTooltip">
{{ graphData.title }}
</gl-tooltip>
<div class="prometheus-graph-widgets js-graph-widgets flex-fill">
<div
class="prometheus-graph-widgets js-graph-widgets flex-fill"
data-qa-selector="prometheus_graph_widgets"
>
<slot></slot>
</div>
</div>
......
......@@ -353,7 +353,7 @@ export default {
</script>
<template>
<div class="prometheus-graphs">
<div class="prometheus-graphs" data-qa-selector="prometheus_graphs">
<div
v-if="showHeader"
ref="prometheusGraphsHeader"
......
......@@ -138,6 +138,7 @@ export default {
v-gl-tooltip
class="ml-auto mx-3"
toggle-class="btn btn-transparent border-0"
data-qa-selector="prometheus_widgets_dropdown"
:right="true"
:no-caret="true"
:title="__('More actions')"
......@@ -161,7 +162,11 @@ export default {
>
{{ __('Generate link to chart') }}
</gl-dropdown-item>
<gl-dropdown-item v-if="alertWidgetAvailable" v-gl-modal="`alert-modal-${index}`">
<gl-dropdown-item
v-if="alertWidgetAvailable"
v-gl-modal="`alert-modal-${index}`"
data-qa-selector="alert_widget_menu_item"
>
{{ __('Alerts') }}
</gl-dropdown-item>
</gl-dropdown>
......
......@@ -331,6 +331,7 @@ export default {
:loading="isRetrying"
:disabled="isRetrying"
container-class="js-pipelines-retry-button btn btn-default btn-retry"
data-qa-selector="pipeline_retry_button"
@click="handleRetryClick"
>
<icon name="repeat" />
......
......@@ -240,7 +240,7 @@
- if project_nav_tab? :environments
= nav_link(controller: :environments, action: [:metrics, :metrics_redirect]) do
= link_to metrics_project_environments_path(@project), title: _('Metrics'), class: 'shortcuts-metrics' do
= link_to metrics_project_environments_path(@project), title: _('Metrics'), class: 'shortcuts-metrics', data: { qa_selector: 'operations_metrics_link' } do
%span
= _('Metrics')
......
......@@ -239,11 +239,12 @@ export default {
<gl-dropdown
id="alert-query-dropdown"
:text="queryDropdownLabel"
toggle-class="dropdown-menu-toggle"
toggle-class="dropdown-menu-toggle qa-alert-query-dropdown"
>
<gl-dropdown-item
v-for="query in relevantQueries"
:key="query.metricId"
data-qa-selector="alert_query_option"
@click="selectQuery(query.metricId)"
>
{{ query.label }}
......@@ -282,6 +283,7 @@ export default {
v-model.number="threshold"
:disabled="formDisabled"
type="number"
data-qa-selector="alert_threshold_field"
/>
</gl-form-group>
</div>
......
......@@ -289,6 +289,8 @@ module QA
autoload :AddExisting, 'qa/page/project/operations/kubernetes/add_existing'
autoload :Show, 'qa/page/project/operations/kubernetes/show'
end
autoload :Metrics, 'qa/page/project/operations/metrics'
end
module Wiki
......
......@@ -128,6 +128,8 @@ module QA
module Kubernetes
autoload :Show, 'qa/ee/page/project/operations/kubernetes/show'
end
autoload :Metrics, 'qa/ee/page/project/operations/metrics'
end
module Packages
......
# frozen_string_literal: true
module QA
module EE
module Page
module Project
module Operations
module Metrics
EXPECTED_LABEL = 'Total (GB)'
def self.prepended(page)
page.module_eval do
view 'ee/app/assets/javascripts/monitoring/components/alert_widget_form.vue' do
element :alert_query_dropdown
element :alert_query_option
element :alert_threshold_field
end
end
end
def wait_for_alert(operator = '>', threshold = 0)
wait_until(reload: false) { has_alert?(operator, threshold) }
end
def has_alert?(operator = '>', threshold = 0)
within_element :prometheus_graphs do
has_text?([EXPECTED_LABEL, operator, threshold].join(' '))
end
end
def write_first_alert(operator = '>', threshold = 0)
open_first_alert_modal
click_on operator
fill_element :alert_threshold_field, threshold
within('.modal-content') { click_button(class: 'btn-success') }
end
def delete_first_alert
open_first_alert_modal
within('.modal-content') { click_button(class: 'btn-danger') }
wait_for_requests
end
def open_first_alert_modal
all_elements(:prometheus_widgets_dropdown, minimum: 1).first.click
click_element :alert_widget_menu_item
click_element :alert_query_dropdown unless has_element?(:alert_query_option, wait: 3)
all_elements(:alert_query_option, minimum: 1).first.click
end
end
end
end
end
end
end
# This is stripped down version of the .gitlab-ci.yml found
# here: https://gitlab.com/joshlambert/autodevops-deploy.
#
# It performs only the deploy stage.
image: alpine:latest
variables:
# AUTO_DEVOPS_DOMAIN is the application deployment domain and should be set as a variable at the group or project level.
AUTO_DEVOPS_DOMAIN: my-fake-domain.com
POSTGRES_USER: user
POSTGRES_PASSWORD: testing-password
POSTGRES_ENABLED: 'false'
POSTGRES_DB: $CI_ENVIRONMENT_SLUG
KUBERNETES_VERSION: 1.11.6
HELM_VERSION: 2.12.2
DOCKER_DRIVER: overlay2
stages:
- production
# This job continuously deploys to production on every push to `master`.
production:
stage: production
script:
- check_kube_domain
- install_dependencies
- download_chart
- ensure_namespace
- initialize_tiller
- create_secret
- deploy
- persist_environment_url
environment:
name: production
url: http://$CI_PROJECT_PATH_SLUG.$AUTO_DEVOPS_DOMAIN
artifacts:
paths: [environment_url.txt]
only:
refs:
- master
kubernetes: active
# ---------------------------------------------------------------------------
.auto_devops: &auto_devops |
# Auto DevOps variables and functions
[[ "$TRACE" ]] && set -x
auto_database_url=postgres://${POSTGRES_USER}:${POSTGRES_PASSWORD}@${CI_ENVIRONMENT_SLUG}-postgres:5432/${POSTGRES_DB}
export DATABASE_URL=${DATABASE_URL-$auto_database_url}
export CI_APPLICATION_REPOSITORY=$CI_REGISTRY_IMAGE/$CI_COMMIT_REF_SLUG
export CI_APPLICATION_TAG=$CI_COMMIT_SHA
export CI_CONTAINER_NAME=ci_job_build_${CI_JOB_ID}
export TILLER_NAMESPACE=$KUBE_NAMESPACE
# Extract "MAJOR.MINOR" from CI_SERVER_VERSION and generate "MAJOR-MINOR-stable" for Security Products
export SP_VERSION=$(echo "$CI_SERVER_VERSION" | sed 's/^\([0-9]*\)\.\([0-9]*\).*/\1-\2-stable/')
function get_replicas() {
track="${1:-stable}"
percentage="${2:-100}"
env_track=$( echo $track | tr -s '[:lower:]' '[:upper:]' )
env_slug=$( echo ${CI_ENVIRONMENT_SLUG//-/_} | tr -s '[:lower:]' '[:upper:]' )
if [[ "$track" == "stable" ]] || [[ "$track" == "rollout" ]]; then
# for stable track get number of replicas from `PRODUCTION_REPLICAS`
eval new_replicas=\$${env_slug}_REPLICAS
if [[ -z "$new_replicas" ]]; then
new_replicas=$REPLICAS
fi
else
# for all tracks get number of replicas from `CANARY_PRODUCTION_REPLICAS`
eval new_replicas=\$${env_track}_${env_slug}_REPLICAS
if [[ -z "$new_replicas" ]]; then
eval new_replicas=\${env_track}_REPLICAS
fi
fi
replicas="${new_replicas:-1}"
replicas="$(($replicas * $percentage / 100))"
# always return at least one replicas
if [[ $replicas -gt 0 ]]; then
echo "$replicas"
else
echo 1
fi
}
# Extracts variables prefixed with K8S_SECRET_
# and creates a Kubernetes secret.
#
# e.g. If we have the following environment variables:
# K8S_SECRET_A=value1
# K8S_SECRET_B=multi\ word\ value
#
# Then we will create a secret with the following key-value pairs:
# data:
# A: dmFsdWUxCg==
# B: bXVsdGkgd29yZCB2YWx1ZQo=
function create_application_secret() {
track="${1-stable}"
export APPLICATION_SECRET_NAME=$(application_secret_name "$track")
env | sed -n "s/^K8S_SECRET_\(.*\)$/\1/p" > k8s_prefixed_variables
kubectl create secret \
-n "$KUBE_NAMESPACE" generic "$APPLICATION_SECRET_NAME" \
--from-env-file k8s_prefixed_variables -o yaml --dry-run |
kubectl replace -n "$KUBE_NAMESPACE" --force -f -
export APPLICATION_SECRET_CHECKSUM=$(cat k8s_prefixed_variables | sha256sum | cut -d ' ' -f 1)
rm k8s_prefixed_variables
}
function deploy_name() {
name="$CI_ENVIRONMENT_SLUG"
track="${1-stable}"
if [[ "$track" != "stable" ]]; then
name="$name-$track"
fi
echo $name
}
function application_secret_name() {
track="${1-stable}"
name=$(deploy_name "$track")
echo "${name}-secret"
}
function deploy() {
track="${1-stable}"
percentage="${2:-100}"
name=$(deploy_name "$track")
replicas="1"
service_enabled="true"
postgres_enabled="$POSTGRES_ENABLED"
# if track is different than stable,
# re-use all attached resources
if [[ "$track" != "stable" ]]; then
service_enabled="false"
postgres_enabled="false"
fi
replicas=$(get_replicas "$track" "$percentage")
if [[ "$CI_PROJECT_VISIBILITY" != "public" ]]; then
secret_name='gitlab-registry'
else
secret_name=''
fi
create_application_secret "$track"
env_slug=$(echo ${CI_ENVIRONMENT_SLUG//-/_} | tr -s '[:lower:]' '[:upper:]')
eval env_ADDITIONAL_HOSTS=\$${env_slug}_ADDITIONAL_HOSTS
if [ -n "$env_ADDITIONAL_HOSTS" ]; then
additional_hosts="{$env_ADDITIONAL_HOSTS}"
elif [ -n "$ADDITIONAL_HOSTS" ]; then
additional_hosts="{$ADDITIONAL_HOSTS}"
fi
if [[ -n "$DB_INITIALIZE" && -z "$(helm ls -q "^$name$")" ]]; then
echo "Deploying first release with database initialization..."
helm upgrade --install \
--wait \
--set service.enabled="$service_enabled" \
--set releaseOverride="$CI_ENVIRONMENT_SLUG" \
--set image.repository="registry.gitlab.com/joshlambert/ruby-gke/master" \
--set image.tag="63492726c2264a0277141d6a6573c3d22ecd7de3" \
--set image.pullPolicy=IfNotPresent \
--set image.secrets[0].name="$secret_name" \
--set application.track="$track" \
--set application.database_url="$DATABASE_URL" \
--set application.secretName="$APPLICATION_SECRET_NAME" \
--set application.secretChecksum="$APPLICATION_SECRET_CHECKSUM" \
--set service.url="$CI_ENVIRONMENT_URL" \
--set service.additionalHosts="$additional_hosts" \
--set replicaCount="$replicas" \
--set postgresql.enabled="$postgres_enabled" \
--set postgresql.nameOverride="postgres" \
--set postgresql.postgresUser="$POSTGRES_USER" \
--set postgresql.postgresPassword="$POSTGRES_PASSWORD" \
--set postgresql.postgresDatabase="$POSTGRES_DB" \
--set application.initializeCommand="$DB_INITIALIZE" \
--set gitlab.app="$CI_PROJECT_PATH_SLUG" \
--set gitlab.env="$CI_ENVIRONMENT_SLUG" \
--namespace="$KUBE_NAMESPACE" \
"$name" \
chart/
echo "Deploying second release..."
helm upgrade --reuse-values \
--wait \
--set application.initializeCommand="" \
--set application.migrateCommand="$DB_MIGRATE" \
--namespace="$KUBE_NAMESPACE" \
"$name" \
chart/
else
echo "Deploying new release..."
helm upgrade --install \
--wait \
--set service.enabled="$service_enabled" \
--set releaseOverride="$CI_ENVIRONMENT_SLUG" \
--set image.repository="registry.gitlab.com/joshlambert/ruby-gke/master" \
--set image.tag="63492726c2264a0277141d6a6573c3d22ecd7de3" \
--set image.pullPolicy=IfNotPresent \
--set image.secrets[0].name="$secret_name" \
--set application.track="$track" \
--set application.database_url="$DATABASE_URL" \
--set application.secretName="$APPLICATION_SECRET_NAME" \
--set application.secretChecksum="$APPLICATION_SECRET_CHECKSUM" \
--set service.url="$CI_ENVIRONMENT_URL" \
--set service.additionalHosts="$additional_hosts" \
--set replicaCount="$replicas" \
--set postgresql.enabled="$postgres_enabled" \
--set postgresql.nameOverride="postgres" \
--set postgresql.postgresUser="$POSTGRES_USER" \
--set postgresql.postgresPassword="$POSTGRES_PASSWORD" \
--set postgresql.postgresDatabase="$POSTGRES_DB" \
--set application.migrateCommand="$DB_MIGRATE" \
--set gitlab.app="$CI_PROJECT_PATH_SLUG" \
--set gitlab.env="$CI_ENVIRONMENT_SLUG" \
--namespace="$KUBE_NAMESPACE" \
"$name" \
chart/
fi
kubectl rollout status -n "$KUBE_NAMESPACE" -w "deployment/$name"
}
function install_dependencies() {
apk add -U openssl curl tar gzip bash ca-certificates git
wget -q -O /etc/apk/keys/sgerrand.rsa.pub https://alpine-pkgs.sgerrand.com/sgerrand.rsa.pub
wget https://github.com/sgerrand/alpine-pkg-glibc/releases/download/2.23-r3/glibc-2.23-r3.apk
apk add glibc-2.23-r3.apk
rm glibc-2.23-r3.apk
curl "https://kubernetes-helm.storage.googleapis.com/helm-v${HELM_VERSION}-linux-amd64.tar.gz" | tar zx
mv linux-amd64/helm /usr/bin/
mv linux-amd64/tiller /usr/bin/
helm version --client
tiller -version
curl -L -o /usr/bin/kubectl "https://storage.googleapis.com/kubernetes-release/release/v${KUBERNETES_VERSION}/bin/linux/amd64/kubectl"
chmod +x /usr/bin/kubectl
kubectl version --client
}
function download_chart() {
if [[ ! -d chart ]]; then
auto_chart=${AUTO_DEVOPS_CHART:-gitlab/auto-deploy-app}
auto_chart_name=$(basename $auto_chart)
auto_chart_name=${auto_chart_name%.tgz}
else
auto_chart="chart"
auto_chart_name="chart"
fi
helm init --client-only
helm repo add gitlab https://charts.gitlab.io
if [[ ! -d "$auto_chart" ]]; then
helm fetch ${auto_chart} --untar
fi
if [ "$auto_chart_name" != "chart" ]; then
mv ${auto_chart_name} chart
fi
helm dependency update chart/
helm dependency build chart/
}
function ensure_namespace() {
kubectl describe namespace "$KUBE_NAMESPACE" || kubectl create namespace "$KUBE_NAMESPACE"
}
function check_kube_domain() {
if [ -z ${AUTO_DEVOPS_DOMAIN+x} ]; then
echo "In order to deploy or use Review Apps, AUTO_DEVOPS_DOMAIN variable must be set"
echo "You can do it in Auto DevOps project settings or defining a secret variable at group or project level"
echo "You can also manually add it in .gitlab-ci.yml"
false
else
true
fi
}
function initialize_tiller() {
echo "Checking Tiller..."
export HELM_HOST="localhost:44134"
tiller -listen ${HELM_HOST} -alsologtostderr > /dev/null 2>&1 &
echo "Tiller is listening on ${HELM_HOST}"
if ! helm version --debug; then
echo "Failed to init Tiller."
return 1
fi
echo ""
}
function create_secret() {
echo "Create secret..."
if [[ "$CI_PROJECT_VISIBILITY" == "public" ]]; then
return
fi
kubectl create secret -n "$KUBE_NAMESPACE" \
docker-registry gitlab-registry \
--docker-server="$CI_REGISTRY" \
--docker-username="$CI_REGISTRY_USER" \
--docker-password="$CI_REGISTRY_PASSWORD" \
--docker-email="$GITLAB_USER_EMAIL" \
-o yaml --dry-run | kubectl replace -n "$KUBE_NAMESPACE" --force -f -
}
function persist_environment_url() {
echo $CI_ENVIRONMENT_URL > environment_url.txt
}
before_script:
- *auto_devops
# frozen_string_literal: true
module QA
module Page
module Project
module Operations
class Metrics < Page::Base
EXPECTED_TITLE = 'Memory Usage (Total)'
LOADING_MESSAGE = 'Waiting for performance data'
view 'app/assets/javascripts/monitoring/components/dashboard.vue' do
element :prometheus_graphs
end
view 'app/assets/javascripts/monitoring/components/charts/time_series.vue' do
element :prometheus_graph_widgets
end
view 'app/assets/javascripts/monitoring/components/panel_type.vue' do
element :prometheus_widgets_dropdown
element :alert_widget_menu_item
end
def wait_for_metrics
wait_for_data
return if has_metrics?
wait_until(max_duration: 180) do
wait_for_data
has_metrics?
end
end
def wait_for_data
wait_until(reload: false) { !has_text?(LOADING_MESSAGE) } if has_text?(LOADING_MESSAGE)
end
def has_metrics?
within_element :prometheus_graphs do
has_text?(EXPECTED_TITLE)
end
end
end
end
end
end
end
QA::Page::Project::Operations::Metrics.prepend_if_ee('QA::EE::Page::Project::Operations::Metrics')
......@@ -9,6 +9,7 @@ module QA::Page
view 'app/assets/javascripts/pipelines/components/pipelines_table_row.vue' do
element :pipeline_commit_status
element :pipeline_retry_button
end
def click_on_latest_pipeline
......@@ -18,11 +19,26 @@ module QA::Page
end
def wait_for_latest_pipeline_success
wait_for_latest_pipeline_status { has_text?('passed') }
end
def wait_for_latest_pipeline_completion
wait_for_latest_pipeline_status { has_text?('passed') || has_text?('failed') }
end
def wait_for_latest_pipeline_status
wait_until(reload: false, max_duration: 300) do
within_element_by_index(:pipeline_commit_status, 0) do
has_text?('passed')
within_element_by_index(:pipeline_commit_status, 0) { yield }
end
end
def wait_for_latest_pipeline_success_or_retry
wait_for_latest_pipeline_completion
if has_text?('failed')
click_element :pipeline_retry_button
wait_for_latest_pipeline_success
end
end
end
end
......
......@@ -12,6 +12,7 @@ module QA
view 'app/views/layouts/nav/sidebar/_project.html.haml' do
element :link_operations
element :operations_environments_link
element :operations_metrics_link
end
end
end
......@@ -24,6 +25,14 @@ module QA
end
end
def go_to_operations_metrics
hover_operations do
within_submenu do
click_element(:operations_metrics_link)
end
end
end
def go_to_operations_kubernetes
hover_operations do
within_submenu do
......
# frozen_string_literal: true
module QA
context 'Monitor' do
describe 'Alerts', :orchestrated, :kubernetes do
before do
@cluster = Service::KubernetesCluster.new.create!
end
after do
@cluster&.remove!
end
it 'allows configuration of alerts' do
Flow::Login.sign_in
project = create_project
create_kubernetes_cluster(project, @cluster)
push_repository(project)
wait_for_deployment
Page::Project::Operations::Metrics.perform do |metrics|
verify_metrics(metrics)
verify_add_alert(metrics)
verify_edit_alert(metrics)
verify_persist_alert(metrics)
verify_delete_alert(metrics)
end
end
private
def create_project
Resource::Project.fabricate_via_api! do |p|
p.name = 'alerts'
p.description = 'Project with alerting configured'
end
end
def create_kubernetes_cluster(project, cluster)
Resource::KubernetesCluster.fabricate_via_browser_ui! do |c|
c.project = project
c.cluster = cluster
c.install_helm_tiller = true
c.install_prometheus = true
c.install_runner = true
end
end
def push_repository(project)
Resource::Repository::ProjectPush.fabricate! do |push|
push.project = project
push.directory = Pathname
.new(__dir__)
.join('../../../../../../fixtures/monitored_auto_devops')
push.commit_message = 'Create Auto DevOps compatible gitlab-ci.yml'
end
end
def wait_for_deployment
Page::Project::Menu.perform(&:click_ci_cd_pipelines)
Page::Project::Pipeline::Index.perform(&:wait_for_latest_pipeline_success_or_retry)
Page::Project::Menu.perform(&:go_to_operations_metrics)
end
def verify_metrics(metrics)
metrics.wait_for_metrics
expect(metrics).to have_metrics
expect(metrics).not_to have_alert
end
def verify_add_alert(metrics)
metrics.write_first_alert('>', 0)
expect(metrics).to have_alert
end
def verify_edit_alert(metrics)
metrics.write_first_alert('<', 0)
expect(metrics).to have_alert('<')
end
def verify_persist_alert(metrics)
metrics.refresh
metrics.wait_for_metrics
metrics.wait_for_alert('<')
expect(metrics).to have_alert('<')
end
def verify_delete_alert(metrics)
metrics.delete_first_alert
expect(metrics).not_to have_alert('<')
end
end
end
end
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment