update deployment strategy to rolling update

[vjayraghavan]

Description

Added deployment strategy to automatically rollout new deploynemt.

Fixes #309

Type of change

• Bug fix (non-breaking change which fixes an issue)

Checklist

• Unit tests and e2e tests updated
• Documentation updated

[Smiley73]

@vjayraghavan please update the description field

[vjayraghavan]

@vjayraghavan please update the description field

updated the description

[seanmalloy]

I'm re-running one of the Travis CI jobs to see if it passes. For reference all of the Travis CI jobs failed with the below error:

+'[' latest = v0.1.6 ']'
make[1]: Leaving directory '/home/travis/gopath/src/github.com/KohlsTechnology/eunomia'
Installing Eunomia Operator
2020/06/10 14:10:19 found symbolic link in path: /home/travis/gopath/src/github.com/KohlsTechnology/eunomia/deploy/helm/eunomia-operator/crds resolves to /home/travis/gopath/src/github.com/KohlsTechnology/eunomia/deploy/crds
namespace/test-eunomia-operator created
serviceaccount/eunomia-operator created
rolebinding.rbac.authorization.k8s.io/eunomia-operator created
customresourcedefinition.apiextensions.k8s.io/gitopsconfigs.eunomia.kohls.io created
clusterrole.rbac.authorization.k8s.io/eunomia-cluster-list created
clusterrole.rbac.authorization.k8s.io/gitopsconfig-admin created
clusterrole.rbac.authorization.k8s.io/gitopsconfig-viewer created
clusterrole.rbac.authorization.k8s.io/eunomia-operator created
clusterrolebinding.rbac.authorization.k8s.io/eunomia-operator created
role.rbac.authorization.k8s.io/eunomia-operator created
service/eunomia-operator created
deployment.apps/eunomia-operator created
deployment.apps/eunomia-operator condition met
error: unable to upgrade connection: container not found ("eunomia-operator")
Eunomia deployment failed
Makefile:62: recipe for target 'test-e2e' failed
make: *** [test-e2e] Error 1
The command "make test-dirty test" exited with 2.

[seanmalloy]

I'm re-running one of the Travis CI jobs to see if it passes

Well ... that didn't work. The Travis CI jobs keep failing with the same error.

[Smiley73]

What's this deployment.apps/eunomia-operator condition met. Never saw that before. Is this related to the PR?

[vinny-sabatini]

@Smiley73 / @seanmalloy that is a result of running the kubectl wait command where we make sure Eunomia is deployed.
Ref:

eunomia/scripts/e2e-test.sh

Line 244 in 30b563d

kubectl wait --for=condition=available --timeout=60s deployment/eunomia-operator -n $OPERATOR_NAMESPACE

The tests were failing as a result of this change and @vjayraghavan is working on updating the tests to properly ensure Eunomia is running with the new deployment strategy.

[codecov]

Codecov Report

Merging #349 into master will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master     #349   +/-   ##
=======================================
  Coverage   24.76%   24.76%           
=======================================
  Files           6        6           
  Lines         541      541           
=======================================
  Hits          134      134           
  Misses        378      378           
  Partials       29       29           

---

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 30b563d...6f72e6e. Read the comment docs.

[vinny-sabatini]

The rest looks good to me. Please remove the progressDeadlineSeconds and then I think this is ready to merge.