* Running operatorhub tests on kubernetes cluster
* Fixing script path
* Align spec field in service binding subscription yaml
* Adding debugging steps for Finding ImagePullBackOff Reason
* Adding mongodb operator and modifying test accordingly to resolve https://github.com/openshift/odo/issues/4651#issuecomment-851269315
* Fixing command typo for kubernetes
* Fixing regex issue
* Fixing mongodb subscription yaml format
* Injecting docker account secrets into kubernetes cluster for etcd services
* remove setup-kube-operators file
* Separate setup for kubernetes operator
* Adding operators common for both ocp and minikube
* sync function for ocp and minikube
* Adding some tests refactor
* refactoring operator hub tests for redis operator
* Finding reason for the pending state of redis pod and fix regex format
* Fixing regex format
* Using redis instance CR definition from example file
* Modifying test to use RedisCluster operator
* Fixing redis-secret failure for specific namespace in the test
* removing not used function for copying example
* Refactoring operator test
* redis pod name are changing so frequently so updating it once again
* Refactored operator link test
* Skipping postgresql test for kubernetes
* Fixing invalid devfile with link
* Fixing file not found error
* Address review comments
* Run all tests on ocp and kubernetes
* Shifting to Redis CRD instead of RedisCluster and review comments
* Fixing failure related to no name
* Address cosmetic changes as per review
* olm installation on minikube
* initial work towards param based odo service create
* change
* add type data to cr
* add to devfile
* update olm
* remove logs
* towards tests
* init tests
* more test
* fix tests
* fix unit again
* set default service name
* mod fix
* addressed some comments
* add postgres operator in setup and in test
* fix
* fix some tests
* fix helpers
* helper flag
* merge conflicts
* add tests and install operator
* certain cases fix
* fix intg tests
* update scripts
* fix intg test
* changed function name
* changes
* fix comments
* docs update
* link fix
* small fix
* comment
* unit tests
* doc fix
* renamed as per comments
* comment
* Add nodejs latest image on ocp 4.7 as supported
* updating the odo init version to 1.18
* Updated odo init image version
* updated init image version
* Moving to static htpasswd file as it is for ephermal test cluster anyway
- password still remains password@123
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Moving kubeadmin to its own script
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Updating script name
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Adding check for non-existent KUBECONFIG
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Removing reset_kubeconfig
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Fixing reset_kubeconfig as it should not be called unless user wants to do it
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Addressing comment by girish
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Spliting configure-installer-test-cluster into a library.
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Fixing password for developer and libdir
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Some fixes
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Fixing periodic tests developer pass
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Changing way script gets it current dir
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Renameing function and doing a set -e
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Fixup for usage where caller script is called with `.`
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Updating remaining scripts to use same auth lib
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Moving htpass up one level
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Adding some echos
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Setting +e for login
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Removing exit on error for parts of the script, where it should not matter
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Adding ubi8/nodejs-12 images as supported
* Updated image stream yaml
* Revert image stream yaml changes
* verify ubi8 nodejs image with e2e test
* Added image namespace in the script file
* changing dc from source bin to deployment
* Revert the dc change
* update odo-init-container to v1.1.6
Co-authored-by: Tomas Kral <tkral@redhat.com>
* Copy kubeconfig to temporary file and grant the permissions to read write execute
* Limiting the file permission to read write only
* Updating copy kubeconfig to temp dir on the basis of review comment
* Fixing login kubeconfig issue
* Add dependency on operator-lifecycle-manager (release-4.1)
* Add Operator Hub client to kclient package
* Add function to list installed operators
To be more specific, it lists the `ClusterServiceVersions` in the
cluster. But a `ClusterServiceVersion` is available only when
Operator(s) is/are installed by the cluster admin.
* List the CSVs in the cluster using odo commands
* Add dependency on github.com/golang/groupcache
* Removes futile call to get services from catalog
We're already getting services from the catalog in `Complete` function
and storing it in `o.services`.
* List the Operators only if experimental mode is enabled
* Update CI cluster setup script to setup operators
* Add OperatorHub tests to test-cmd-service
* Integration tests for listing operators
* Remove specific CSV version to let it be automatically selected
* Add test suite for OperatorHub tests to run
* Separate PHONY for OperatorHub tests and run it on 4.x cluster
* Use a CRD's kind instead of DisplayName to make service creation simpler in future
* Set environment variable for experimental mode
* Modify cluster setup to install operators on 4.1
* Execute different steps to setup OperatorHub if running on OCP 4.1
* Fix operator setup script for OCP 4.x
* Changes as per the PR review
https://github.com/openshift/odo/pull/2593#pullrequestreview-363449893
* Added klog dependency through `glide update -v`
* Changes as per the PR review
https://github.com/openshift/odo/pull/2593#pullrequestreview-372480988https://github.com/openshift/odo/pull/2593#pullrequestreview-372484636
* Making pre-test setup script check api server stability
The script now checks if api server is stable after success login
and if not does reasonable retry of login
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Stabalizing OC_STABLE_LOGIN by initializing and adding `set -x`
* fix java_e2e tests
increase waitForPodTimeOut
increate timeout for 'oc import-image'
configure-installer-test-cluster: show version
rename openjdk image stream to java, and don't import it on OpenShift CI
* fix source_e2e test
* e2e: waitForCmdOut show command out on failure
* e2e: use oc to delete projects
* fix e2e_test.go
* specify branch in git tests
* fix component e2e (project)
* e2e: add debug to cmdRunner
* e2e add project fflag
* Removing unnessaary make bin from script.
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Modifying e2e tests to not check project name
* Increasing time out
* Removing timeout as its added to test
* Fixing script to be able to skip logging in as admin, if needed
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Adding make bin
* Updating script to try success login multiple times
* Adding reference for CI environment variable
* Removing hard sleep in favor of checking successful login
* Removing unused env
* Adding scripts to handle initialization of cluster for openshift CI.
This also fixes couple of usernames on the side
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
* Adding http-tools to build root so we can use htpasswd
* Adding login as developer and creation of first project
* Moving script to ./scripts and renaming makefile command
* Renaming scripts to incude tests in configure cluster script
* Making script more configurable
* Adding namespace to secret commands.
* Fixing missing username
* Simplifying script by removing unnessasary variables
* Adding TODO and slightly reduction wait time for auth config
* Making assets dir overwritable
* Making configure script executable.
* Fixing message for not existing kubeconfig
* Redirecting out and err of commands where we are not interested in msg
* Updating to 18 secs default sleep which can be overridden
* Adding workaround for missing wildfly in OpenShift 4.0
* Adding basic ci docs for prow
Signed-off-by: Mohammed Zeeshan Ahmed <mohammed.zee1000@gmail.com>
