Skip to content

Latest commit

 

History

History
743 lines (639 loc) · 21.4 KB

how-it-works.md

File metadata and controls

743 lines (639 loc) · 21.4 KB

How it works

Flagger takes a Kubernetes deployment and optionally a horizontal pod autoscaler (HPA) and creates a series of objects (Kubernetes deployments, ClusterIP services and Istio or App Mesh virtual services) to drive the canary analysis and promotion.

Flagger Canary Process

Canary Custom Resource

For a deployment named podinfo, a canary promotion can be defined using Flagger's custom resource:

apiVersion: flagger.app/v1alpha3
kind: Canary
metadata:
  name: podinfo
  namespace: test
spec:
  # deployment reference
  targetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: podinfo
  # the maximum time in seconds for the canary deployment
  # to make progress before it is rollback (default 600s)
  progressDeadlineSeconds: 60
  # HPA reference (optional)
  autoscalerRef:
    apiVersion: autoscaling/v2beta1
    kind: HorizontalPodAutoscaler
    name: podinfo
  service:
    # container port
    port: 9898
    # service port name (optional, will default to "http")
    portName: http-podinfo
    # Istio gateways (optional)
    gateways:
    - public-gateway.istio-system.svc.cluster.local
    # Istio virtual service host names (optional)
    hosts:
    - podinfo.example.com
  # promote the canary without analysing it (default false)
  skipAnalysis: false
  # define the canary analysis timing and KPIs
  canaryAnalysis:
    # schedule interval (default 60s)
    interval: 1m
    # max number of failed metric checks before rollback
    threshold: 10
    # max traffic percentage routed to canary
    # percentage (0-100)
    maxWeight: 50
    # canary increment step
    # percentage (0-100)
    stepWeight: 5
    # Prometheus checks
    metrics:
    - name: request-success-rate
      # minimum req success rate (non 5xx responses)
      # percentage (0-100)
      threshold: 99
      interval: 1m
    - name: request-duration
      # maximum req duration P99
      # milliseconds
      threshold: 500
      interval: 30s
    # external checks (optional)
    webhooks:
      - name: integration-tests
        url: http://podinfo.test:9898/echo
        timeout: 1m
        # key-value pairs (optional)
        metadata:
          test: "all"
          token: "16688eb5e9f289f1991c"

Note that the target deployment must have a single label selector in the format app: <DEPLOYMENT-NAME>:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: podinfo
spec:
  selector:
    matchLabels:
      app: podinfo
  template:
    metadata:
      labels:
        app: podinfo

Besides app Flagger supports name and app.kubernetes.io/name selectors. If you use a different convention you can specify your label with the -selector-labels flag.

The target deployment should expose a TCP port that will be used by Flagger to create the ClusterIP Service and the Istio Virtual Service. The container port from the target deployment should match the service.port value.

Istio routing

Flagger creates an Istio Virtual Service based on the Canary service spec. The service configuration lets you expose an app inside or outside the mesh. You can also define HTTP match conditions, URI rewrite rules, CORS policies, timeout and retries.

The following spec exposes the frontend workload inside the mesh on frontend.test.svc.cluster.local:9898 and outside the mesh on frontend.example.com. You'll have to specify an Istio ingress gateway for external hosts.

apiVersion: flagger.app/v1alpha3
kind: Canary
metadata:
  name: frontend
  namespace: test
spec:
  service:
    # container port
    port: 9898
    # service port name (optional, will default to "http")
    portName: http-frontend
    # Istio gateways (optional)
    gateways:
    - public-gateway.istio-system.svc.cluster.local
    - mesh
    # Istio virtual service host names (optional)
    hosts:
    - frontend.example.com
    # HTTP match conditions (optional)
    match:
      - uri:
          prefix: /
    # HTTP rewrite (optional)
    rewrite:
      uri: /
    # Envoy timeout and retry policy (optional)
    headers:
      request:
        add:
          x-envoy-upstream-rq-timeout-ms: "15000"
          x-envoy-max-retries: "10"
          x-envoy-retry-on: "gateway-error,connect-failure,refused-stream"
    # cross-origin resource sharing policy (optional)
    corsPolicy:
      allowOrigin:
        - example.com
      allowMethods:
        - GET
      allowCredentials: false
      allowHeaders:
        - x-some-header
      maxAge: 24h

For the above spec Flagger will generate the following virtual service:

apiVersion: networking.istio.io/v1alpha3
kind: VirtualService
metadata:
  name: frontend
  namespace: test
  ownerReferences:
    - apiVersion: flagger.app/v1alpha3
      blockOwnerDeletion: true
      controller: true
      kind: Canary
      name: podinfo
      uid: 3a4a40dd-3875-11e9-8e1d-42010a9c0fd1
spec:
  gateways:
    - public-gateway.istio-system.svc.cluster.local
    - mesh
  hosts:
    - frontend.example.com
    - frontend
  http:
  - appendHeaders:
      x-envoy-max-retries: "10"
      x-envoy-retry-on: gateway-error,connect-failure,refused-stream
      x-envoy-upstream-rq-timeout-ms: "15000"
    corsPolicy:
      allowHeaders:
      - x-some-header
      allowMethods:
      - GET
      allowOrigin:
      - example.com
      maxAge: 24h
    match:
    - uri:
        prefix: /
    rewrite:
      uri: /
    route:
    - destination:
        host: podinfo-primary
        port:
          number: 9898
      weight: 100
    - destination:
        host: podinfo-canary
        port:
          number: 9898
      weight: 0

Flagger keeps in sync the virtual service with the canary service spec. Any direct modification to the virtual service spec will be overwritten.

To expose a workload inside the mesh on http://backend.test.svc.cluster.local:9898, the service spec can contain only the container port:

apiVersion: flagger.app/v1alpha3
kind: Canary
metadata:
  name: backend
  namespace: test
spec:
  service:
    port: 9898

Based on the above spec, Flagger will create several ClusterIP services like:

apiVersion: v1
kind: Service
metadata:
  name: backend-primary
  ownerReferences:
  - apiVersion: flagger.app/v1alpha3
    blockOwnerDeletion: true
    controller: true
    kind: Canary
    name: backend
    uid: 2ca1a9c7-2ef6-11e9-bd01-42010a9c0145
spec:
  type: ClusterIP
  ports:
  - name: http
    port: 9898
    protocol: TCP
    targetPort: 9898
  selector:
    app: backend-primary

Flagger works for user facing apps exposed outside the cluster via an ingress gateway and for backend HTTP APIs that are accessible only from inside the mesh.

Canary Stages

Flagger Canary Stages

A canary deployment is triggered by changes in any of the following objects:

  • Deployment PodSpec (container image, command, ports, env, resources, etc)
  • ConfigMaps mounted as volumes or mapped to environment variables
  • Secrets mounted as volumes or mapped to environment variables

Gated canary promotion stages:

  • scan for canary deployments
  • check Istio virtual service routes are mapped to primary and canary ClusterIP services
  • check primary and canary deployments status
    • halt advancement if a rolling update is underway
    • halt advancement if pods are unhealthy
  • call pre-rollout webhooks are check results
    • halt advancement if any hook returned a non HTTP 2xx result
    • increment the failed checks counter
  • increase canary traffic weight percentage from 0% to 5% (step weight)
  • call rollout webhooks and check results
  • check canary HTTP request success rate and latency
    • halt advancement if any metric is under the specified threshold
    • increment the failed checks counter
  • check if the number of failed checks reached the threshold
    • route all traffic to primary
    • scale to zero the canary deployment and mark it as failed
    • call post-rollout webhooks
    • post the analysis result to Slack
    • wait for the canary deployment to be updated and start over
  • increase canary traffic weight by 5% (step weight) till it reaches 50% (max weight)
    • halt advancement if any webhook call fails
    • halt advancement while canary request success rate is under the threshold
    • halt advancement while canary request duration P99 is over the threshold
    • halt advancement if the primary or canary deployment becomes unhealthy
    • halt advancement while canary deployment is being scaled up/down by HPA
  • promote canary to primary
    • copy ConfigMaps and Secrets from canary to primary
    • copy canary deployment spec template over primary
  • wait for primary rolling update to finish
    • halt advancement if pods are unhealthy
  • route all traffic to primary
  • scale to zero the canary deployment
  • mark rollout as finished
  • call post-rollout webhooks
  • post the analysis result to Slack
  • wait for the canary deployment to be updated and start over

Canary Analysis

The canary analysis runs periodically until it reaches the maximum traffic weight or the failed checks threshold.

Spec:

  canaryAnalysis:
    # schedule interval (default 60s)
    interval: 1m
    # max number of failed metric checks before rollback
    threshold: 10
    # max traffic percentage routed to canary
    # percentage (0-100)
    maxWeight: 50
    # canary increment step
    # percentage (0-100)
    stepWeight: 2
  # deploy straight to production without
  # the metrics and webhook checks
  skipAnalysis: false

The above analysis, if it succeeds, will run for 25 minutes while validating the HTTP metrics and webhooks every minute. You can determine the minimum time that it takes to validate and promote a canary deployment using this formula:

interval * (maxWeight / stepWeight)

And the time it takes for a canary to be rollback when the metrics or webhook checks are failing:

interval * threshold

In emergency cases, you may want to skip the analysis phase and ship changes directly to production. At any time you can set the spec.skipAnalysis: true. When skip analysis is enabled, Flagger checks if the canary deployment is healthy and promotes it without analysing it. If an analysis is underway, Flagger cancels it and runs the promotion.

A/B Testing

Besides weighted routing, Flagger can be configured to route traffic to the canary based on HTTP match conditions. In an A/B testing scenario, you'll be using HTTP headers or cookies to target a certain segment of your users. This is particularly useful for frontend applications that require session affinity.

You can enable A/B testing by specifying the HTTP match conditions and the number of iterations:

  canaryAnalysis:
    # schedule interval (default 60s)
    interval: 1m
    # total number of iterations
    iterations: 10
    # max number of failed iterations before rollback
    threshold: 2
    # canary match condition
    match:
      - headers:
          user-agent:
            regex: "^(?!.*Chrome).*Safari.*"
      - headers:
          cookie:
            regex: "^(.*?;)?(user=test)(;.*)?$"

If Flagger finds a HTTP match condition, it will ignore the maxWeight and stepWeight settings.

The above configuration will run an analysis for ten minutes targeting the Safari users and those that have a test cookie. You can determine the minimum time that it takes to validate and promote a canary deployment using this formula:

interval * iterations

And the time it takes for a canary to be rollback when the metrics or webhook checks are failing:

interval * threshold

Make sure that the analysis threshold is lower than the number of iterations.

HTTP Metrics

The canary analysis is using the following Prometheus queries:

HTTP requests success rate percentage

Spec:

  canaryAnalysis:
    metrics:
    - name: request-success-rate
      # minimum req success rate (non 5xx responses)
      # percentage (0-100)
      threshold: 99
      interval: 1m

Istio query:

sum(
    rate(
        istio_requests_total{
          reporter="destination",
          destination_workload_namespace=~"$namespace",
          destination_workload=~"$workload",
          response_code!~"5.*"
        }[$interval]
    )
) 
/ 
sum(
    rate(
        istio_requests_total{
          reporter="destination",
          destination_workload_namespace=~"$namespace",
          destination_workload=~"$workload"
        }[$interval]
    )
)

App Mesh query:

sum(
    rate(
        envoy_cluster_upstream_rq{
          kubernetes_namespace="$namespace",
          kubernetes_pod_name=~"$workload",
          response_code!~"5.*"
        }[$interval]
    )
) 
/ 
sum(
    rate(
        envoy_cluster_upstream_rq{
          kubernetes_namespace="$namespace",
          kubernetes_pod_name=~"$workload"
        }[$interval]
    )
)

HTTP requests milliseconds duration P99

Spec:

  canaryAnalysis:
    metrics:
    - name: request-duration
      # maximum req duration P99
      # milliseconds
      threshold: 500
      interval: 1m

Istio query:

histogram_quantile(0.99, 
  sum(
    irate(
      istio_request_duration_seconds_bucket{
        reporter="destination",
        destination_workload=~"$workload",
        destination_workload_namespace=~"$namespace"
      }[$interval]
    )
  ) by (le)
)

App Mesh query:

histogram_quantile(0.99, 
  sum(
    irate(
      envoy_cluster_upstream_rq_time_bucket{
        kubernetes_pod_name=~"$workload",
        kubernetes_namespace=~"$namespace"
      }[$interval]
    )
  ) by (le)
)

Note that the metric interval should be lower or equal to the control loop interval.

Custom Metrics

The canary analysis can be extended with custom Prometheus queries.

  canaryAnalysis:
    threshold: 1
    maxWeight: 50
    stepWeight: 5
    metrics:
    - name: "404s percentage"
      threshold: 5
      query: |
        100 - sum(
            rate(
                istio_requests_total{
                  reporter="destination",
                  destination_workload_namespace="test",
                  destination_workload="podinfo",
                  response_code!="404"
                }[1m]
            )
        )
        /
        sum(
            rate(
                istio_requests_total{
                  reporter="destination",
                  destination_workload_namespace="test",
                  destination_workload="podinfo"
                }[1m]
            )
        ) * 100

The above configuration validates the canary by checking if the HTTP 404 req/sec percentage is below 5 percent of the total traffic. If the 404s rate reaches the 5% threshold, then the canary fails.

  canaryAnalysis:
    threshold: 1
    maxWeight: 50
    stepWeight: 5
    metrics:
    - name: "rpc error rate"
      threshold: 5
      query: |
        100 - (sum
            rate(
                grpc_server_handled_total{
                  grpc_service="my.TestService",
                  grpc_code!="OK"
                }[1m]
            )
        )
        /
        sum(
            rate(
                grpc_server_started_total{
                  grpc_service="my.TestService"
                }[1m]
            )
        ) * 100

The above configuration validates the canary by checking if the percentage of non-OK GRPC req/sec is below 5 percent of the total requests. If the non-OK rate reaches the 5% threshold, then the canary fails.

When specifying a query, Flagger will run the promql query and convert the result to float64. Then it compares the query result value with the metric threshold value.

Webhooks

The canary analysis can be extended with webhooks. Flagger will call each webhook URL and determine from the response status code (HTTP 2xx) if the canary is failing or not.

There are three types of hooks:

  • Pre-rollout hooks are executed before routing traffic to canary. The canary advancement is paused if a pre-rollout hook fails and if the number of failures reach the threshold the canary will be rollback.
  • Rollout hooks are executed during the analysis on each iteration before the metric checks. If a rollout hook call fails the canary advancement is paused and eventfully rolled back.
  • Post-rollout hooks are executed after the canary has been promoted or rolled back. If a post rollout hook fails the error is logged.

Spec:

  canaryAnalysis:
    webhooks:
      - name: "smoke test"
        type: pre-rollout
        url: http://migration-check.db/query
        timeout: 30s
        metadata:
          key1: "val1"
          key2: "val2"
      - name: "load test"
        type: rollout
        url: http://flagger-loadtester.test/
        timeout: 15s
        metadata:
          cmd: "hey -z 1m -q 5 -c 2 http://podinfo-canary.test:9898/"
      - name: "notify"
        type: post-rollout
        url: http://telegram.bot:8080/
        timeout: 5s
        metadata:
          some: "message"

Note that the sum of all rollout webhooks timeouts should be lower than the analysis interval.

Webhook payload (HTTP POST):

{
    "name": "podinfo",
    "namespace": "test",
    "phase": "Progressing", 
    "metadata": {
        "test":  "all",
        "token":  "16688eb5e9f289f1991c"
    }
}

Response status codes:

  • 200-202 - advance canary by increasing the traffic weight
  • timeout or non-2xx - halt advancement and increment failed checks

On a non-2xx response Flagger will include the response body (if any) in the failed checks log and Kubernetes events.

Load Testing

For workloads that are not receiving constant traffic Flagger can be configured with a webhook, that when called, will start a load test for the target workload. If the target workload doesn't receive any traffic during the canary analysis, Flagger metric checks will fail with "no values found for metric istio_requests_total".

Flagger comes with a load testing service based on rakyll/hey that generates traffic during analysis when configured as a webhook.

Flagger Load Testing Webhook

First you need to deploy the load test runner in a namespace with Istio sidecar injection enabled:

export REPO=https://raw.githubusercontent.com/weaveworks/flagger/master

kubectl -n test apply -f ${REPO}/artifacts/loadtester/deployment.yaml
kubectl -n test apply -f ${REPO}/artifacts/loadtester/service.yaml

Or by using Helm:

helm repo add flagger https://flagger.app

helm upgrade -i flagger-loadtester flagger/loadtester \
--namespace=test \
--set cmd.timeout=1h

When deployed the load tester API will be available at http://flagger-loadtester.test/.

Now you can add webhooks to the canary analysis spec:

webhooks:
  - name: load-test-get
    url: http://flagger-loadtester.test/
    timeout: 5s
    metadata:
      type: cmd
      cmd: "hey -z 1m -q 10 -c 2 http://podinfo.test:9898/"
  - name: load-test-post
    url: http://flagger-loadtester.test/
    timeout: 5s
    metadata:
      type: cmd
      cmd: "hey -z 1m -q 10 -c 2 -m POST -d '{test: 2}' http://podinfo.test:9898/echo"

When the canary analysis starts, Flagger will call the webhooks and the load tester will run the hey commands in the background, if they are not already running. This will ensure that during the analysis, the podinfo.test virtual service will receive a steady stream of GET and POST requests.

If your workload is exposed outside the mesh with the Istio Gateway and TLS you can point hey to the public URL and use HTTP2.

webhooks:
  - name: load-test-get
    url: http://flagger-loadtester.test/
    timeout: 5s
    metadata:
      type: cmd
      cmd: "hey -z 1m -q 10 -c 2 -h2 https://podinfo.example.com/"

The load tester can run arbitrary commands as long as the binary is present in the container image. For example if you you want to replace hey with another CLI, you can create your own Docker image:

FROM quay.io/stefanprodan/flagger-loadtester:<VER>

RUN curl -Lo /usr/local/bin/my-cli https://github.com/user/repo/releases/download/ver/my-cli \
    && chmod +x /usr/local/bin/my-cli

Load Testing Delegation

The load tester can also forward testing tasks to external tools, by now nGrinder is supported.

To use this feature, add a load test task of type 'ngrinder' to the canary analysis spec:

webhooks:
  - name: load-test-post
    url: http://flagger-loadtester.test/
    timeout: 5s
    metadata:
      # type of this load test task, cmd or ngrinder
      type: ngrinder
      # base url of your nGrinder controller server
      server: http://ngrinder-server:port
      # id of the test to clone from, the test must have been defined.
      clone: 100
      # user name and base64 encoded password to authenticate against the nGrinder server
      username: admin
      passwd: YWRtaW4=
      # the interval between between nGrinder test status polling, default to 1s
      pollInterval: 5s

When the canary analysis starts, the load tester will initiate a clone_and_start request to the nGrinder server and start a new performance test. the load tester will periodically poll the nGrinder server for the status of the test, and prevent duplicate requests from being sent in subsequent analysis loops.