feat: Traefik support. Fixes #516

[PhilippPlotnikov]

Fixes #516

Checklist:

• Either (a) I've created an enhancement proposal and discussed it with the community, (b) this is a bug fix, or (c) this is a chore.
• The title of the PR is (a) conventional, (b) states what changed, and (c) suffixes the related issues number. E.g. "fix(controller): Updates such and such. Fixes #1234".
• I've signed my commits with DCO
• I have written unit and/or e2e tests for my change. PRs without these are unlikely to be merged.
• My builds are green. Try syncing with master if they are not.
• My organization is added to USERS.md.

[codecov]

Codecov Report

Merging #1907 (ba6d753) into master (e42b4d4) will decrease coverage by 0.01%.
The diff coverage is 82.69%.

@@            Coverage Diff             @@
##           master    #1907      +/-   ##
==========================================
- Coverage   82.51%   82.50%   -0.02%     
==========================================
  Files         119      120       +1     
  Lines       16919    17054     +135     
==========================================
+ Hits        13961    14070     +109     
- Misses       2265     2283      +18     
- Partials      693      701       +8     

Impacted Files	Coverage Δ
utils/defaults/defaults.go	88.35% <ø> (ø)
rollout/trafficrouting/traefik/traefik.go	81.25% <81.25%> (ø)
rollout/trafficrouting.go	81.70% <100.00%> (+0.64%)	⬆️
metricproviders/prometheus/prometheus.go	96.42% <0.00%> (-3.58%)	⬇️
analysis/analysis.go	85.53% <0.00%> (-0.59%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update e42b4d4...ba6d753. Read the comment docs.

[leoluz]

Please update rollouts controller rbac permissions allowing it to update the traefik service resources.

[perenesenko]

have one comment for the VerifyWeight

[perenesenko]

Probably there is no need to implement this method in that way. You're accessing the same traefic service object (which you just modified in a SetWeight method)

Right now it's implemented for ALB only. The reason whay we have this implementation for the ALB is delays in traffic shifting.
For the ALB once you modified the Ingress object it take some time to really shift the traffic by creating/modifying target group. So here we need to make validation against the target group objects to see the traffic was really updated.

[sonarcloud]

Kudos, SonarCloud Quality Gate passed!   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
2 Code Smells

No Coverage information
2.6% Duplication

[leoluz]

LGTM

[perenesenko]

LGTM

[ayanhamza]

@PhilippPlotnikov Hello !
I am using Traefik ingress for Canary Deploy and used kind TraefikService for trafficRouting. Everything works fine when i start deploy canary release. Traffic for my services splits by weight for Canary pods by percentage. BUT when Canary version of my app is Degraded (in case of failure analysis) it's rollback to old version and requests for my Stable version not go. Because argo rollout after Degraded don't sends traffic via canary service. How to modify code to send traffic via canary service also after Degraded??
Cause in IngressRoute object of Traefik I declare in spec service "TraefikService". And TraefikService dont't work's properly when one of the services can't receive traffic to pods.

  strategy:
    canary:
      canaryService: colorapi-canary
      stableService: colorapi-stable
      trafficRouting:
        traefik:
          weightedTraefikServiceName: colorapi
      analysis:
        templates:
        - templateName: web
      steps:
      - setWeight: 10
      - pause:
          duration: 50s
      - setWeight: 50
      - pause:
          duration: 50s

apiVersion: traefik.containo.us/v1alpha1
kind: TraefikService
metadata:
  name: colorapi
  namespace: devops-services
spec:
  weighted:
    services:
      - name: colorapi-canary
        port: 5000
      - name: colorapi-stable
        port: 5000

apiVersion: traefik.containo.us/v1alpha1
kind: IngressRoute
metadata:
  annotations:
    kubernetes.io/ingress.class: traefik
  creationTimestamp: null
  name: colorapi
  namespace: devops-services
spec:
  entryPoints: []
  routes:
  - kind: Rule
    match: Host("colorapi") && PathPrefix("/")
    middlewares: []
    priority: 0
    services:
    - kind: TraefikService
      name: colorapi

[ayanhamza]

I found that after "Degraded" selector of kind:Service canary not return to old ReplicaSet Hash. And still remains in new ReplicaSet Hash. That's why canary-service doesn't have endpoints. And TraefikService not serve traffic to stable version,
Please help. How to modify code to change Hash of selector to old ReplicaSet if new canary version of app don't updated and Degraded ?

Name:              colorapi-canary
Namespace:         devops-services
Labels:            app=colorapi
                   app.kubernetes.io/instance=colorapi
Annotations:       argo-rollouts.argoproj.io/managed-by-rollouts: colorapi
Selector:          app=colorapi,rollouts-pod-template-hash=75f9cc4fb4
Type:              ClusterIP
IP Families:       <none>
IP:                192.168.18.218
IPs:               <none>
Port:              web  5000/TCP
TargetPort:        5000/TCP
Endpoints:         <none>
Session Affinity:  None
Events:            <none>

Name:              colorapi-stable
Namespace:         devops-services
Labels:            app=colorapi
                   app.kubernetes.io/instance=colorapi
Annotations:       argo-rollouts.argoproj.io/managed-by-rollouts: colorapi
Selector:          app=colorapi,rollouts-pod-template-hash=5c559c684b
Type:              ClusterIP
IP Families:       <none>
IP:                192.168.29.202
IPs:               <none>
Port:              web  5000/TCP
TargetPort:        5000/TCP
Endpoints:         192.168.189.142:5000,192.168.223.178:5000
Session Affinity:  None
Events:            <none>