Skip to content
This repository has been archived by the owner on Jul 23, 2020. It is now read-only.

FAQ

Jump to bottom
George Gastaldi edited this page Feb 5, 2018 · 66 revisions

Account / Tenant / Username Questions

I try to login and I see my username is not approved.

Click on the Register Now link in the landing page, and sign-in using your @redhat.com account.

I am an approved user, but I'm unable to get past the landing page.

Try hitting https://openshift.io/openshiftio and then logging in.

The username in Keycloak is different than what I see on OpenShift.io

Re-login to OpenShift.io to be able to see your updated username. This was fixed as part of https://github.com/fabric8-services/fabric8-wit/pull/1367

My username on OpenShift.io is different than the name of the project created in my OpenShift Online account

The reason is unknown as of now, however the debugging session was successful in resolving this for chmouel This will cause issues with creating your QuickStart projects and pipelines. ( need to fill in details )

How do I change my password?

Visit Red Hat Developers and change your password there.

How do I update my tenant?

Your own jenkins, che, etc is running in your OpenShift Online account in the https://console.starter-us-east-2.openshift.com cluster. The Deployment Configs have to be updated every few days. To do so,

  • Login to OpenShift.io , if you face issues while trying to login, please see ( add link )
  • Visit your profile by clicking on the top-right corner.
  • Click "Update profile"
  • Scroll down to the bottom of the page and click "Update tenant". This doesn't provide any feedback as of now, but you could check for a /api/user 200 OK response in the browser console.

How do I confirm that my tenant is updated?

To confirm you need to be checking which build version of Jenkins/Che you are running, and compare it with the intended version.

To find out which version of the Jenkins 2 build you are running, use the oc CLI , and try running oc export dc jenkins -n oc {ID}-jenkins | grep "version:"

If the above doesn't work, use the UI:

  • Login to OpenShift Online https://console.starter-us-east-2.openshift.com
  • Select project {ID}-jenkins
  • Navigate to ``Applications->Deployments` -> `jenkins`.
  • Open the latest 'active' deployment from the list.
  • On the details tab, there should a version=3.x.x

My tenant update does not work

https://github.com/fabric8io/fabric8-init-tenant/issues/74

Applications / Resources / OpenShift Online Questions

Where are my applications running?

Currently, the tenant URL is https://console.starter-us-east-2.openshift.com

What resources are used in my OpenShift Online account?

For your OSO username of ID the following resources are used

  • namespace ${ID}: contains your BuildConfigs and Builds of any pipelines along with a jenkins Service. If running integration tests these will run pods in this namespace. You should not have a jenkins here though. By default on OSO if you create a BuildConfig with a pipeline then OSO will lazily create a Jenkins here. Though we use a jenkins service here to prevent that. If one gets created you will probably wanna delete your jenkins DeploymentConfig, ReplicationControllers and Pods
  • namespace: ${ID}-stage: the stage environments for your apps you stage via pipelines
  • namespace: ${ID}-run: the main running environment for your apps you promote from staging via pipelines
  • namespace: ${ID}-che: contains your Che workspaces
  • namespace: ${ID}-jenkins: contains your Jenkins master pod along with the content-repository for storing released artifacts. Also contains any S2I Build objects used inside a pipeline and any build pipeline pods.

I already have a project I had created in my OpenShift Online account.

Assuming your old project is pke-test1 and your OpenShift.io username is pkettman, this might cause your pipelines to not show up since there is no pkettman namespace.

Clean-up your existing project by running the following:

Visit https://console.starter-us-east-2.openshift.com/oauth/token/request to get your oc auth token needed for the CLI login.

oc login https://console.starter-us-east-2.openshift.com  --token=TOKEN
oc delete project pke-test1
oc new-project pkettman
oc delete all --all -n pkettman-test
oc delete all --all -n pkettman-stage
oc delete all --all -n pkettman-run

Projects / Forge Wizard Questions

I am seeing an error when I create a new project - what should I do?

The actions that you should take will depend on the nature of the error. The first step to take is to enable the Chrome console log to provide you with detailed information on the any error that you encounter. Some possible project creation errors and their workarounds are:

  • Error code 500 (service error) - This general error is usually accompanied by a second, more specific error. If the second error includes an error code of 400 (bad request), you may need to have Red Hat Engineering try detaching and re-attaching your github account to OSIO and then try again to create the a project.
  • Error code 503 (service unavailable) - The most likely cause of this error is that the Jenkins service in you rtenant space is not running. As is described earlier in this FAQ, the workaround is to restart the Jenkins pod, and then try again to create the a project.
  • Error code 401 (permissions error) - This is an intermittent error related to the creation of the project's code repository in github. If you see this error, just try again to create a new project.
  • Error in the Jenkins log that indicates that the project's jar file cannot be found - The problem is that the project is running out of space on the the maven .m2 repository's Persistent Volume (PV). The steps to take are: ** oc project [your username]-jenkins ** oc get pods (look for the name of the "content repository" pod, for example: "content-repository-5-3cw6j") ** oc exec -ti content-repository-5-3cw6j bash ** rm -rf /var/www/html/content/repositories
  • Error in the Jenkins log that indicates - java.lang.OutOfMemoryError: Metaspace - As is described earlier in this FAQ, the workaround is to restart the Jenkins pod.
  • The new project's Che codespace is not created due to a 504 (timeout) error - This is an intermittent error. If you see this error, just try again to create a new project.

Pipelines / Builds / Jenkins Questions

I add a new QuickStart project but the build doesn't start on the Pipelines page.

This happens because of a bug in the jenkins-sync plugin which restarts all old builds. Since, we have a restriction of only allowing 1 build at a time, the newly created project's build pipeline falls low into the build queue. To fix this, abort all the queued up builds except the one which was created 'just now'.

  • On the Pipelines UI, click on the Jenkins URL ( this part needs better instructions )
  • In the Jenkins UI, locate the queue visualisation on the left-hand side of the page.
  • Click on the red abort button for each build.

Jenkins is not responding, says "Application is not available"

"When I goto https://jenkins-username-jenkins.8a09.starter-us-east-2.openshiftapps.com I just get Application Not Available"

This happens when your Jenkins service is struggling to load because of resource issues.

Workaround :

My pipeline seems to have updated staging but my latest commit isn't there

It's likely that the staging server hasn't actually been updated. OpenShift.io (IO) updates the staging application version number when the new deployment configuration is sent to OpenShift Online (OSO). If your resource quota in OSO has been exceeded however, you won't be able to update the existing staging pod. In this case click on the build number link in the Pipelines tab in IO to open in OSO. Once in the OSO console, look at the <your-username>-stage project and watch for running staging servers from older spaces.

My Pipeline build fails with out-of-memory exceptions

[Pipeline] End of Pipeline
java.lang.OutOfMemoryError: Metaspace
Finished: FAILURE

The solution is same as the one above, if you are using the OpenShift CLI, you could use these command alternatively to restart the Jenkins pod

oc scale dc/jenkins --replicas=0 -n burrsutter-jenkins
oc scale dc/jenkins --replicas=1 -n burrsutter-jenkins

https://github.com/openshiftio/openshift.io/issues/206

How do I enable or disable integration tests in my pipelines?

You can edit your Jenkinsfile to comment out the mavenIntegrationTest section on a per app basis.

If you are doing demos you may want to disable integration tests to speed up demos ;) To do that you can globally configure pipelines (once the new Tenant is released) to disable integration tests via updating your fabric8-pipelines ConfigMap.

Type:

oc edit cm fabric8-pipelines

Then add the following to the data: section of the ConfigMap to disable the CD and/or CI integration tests

data:
  disable-itests-cd: 'true'
  disable-itests-ci: 'true'

NOTE that there is an issue with RHOAR boosters so we've temporarily disabled all integration tests anyway!

My Pipelines are not being properly filtered by my Space?

When you switch between spaces the Pipelines tab should filter by those apps created in the space.

If you created an App a while back (say before June 2017) then the created Pipeline (which is a BuildConfig inside OpenShift) will not have the space label associated so that per space filtering doesn't work.

To work around this from the openshift command line type:

oc edit bc foo

then ensure that the metadata.labels has a value of space: cheese like this:

metadata:
  name: foo
  labels:
     space: cheese
...

that will then ensure that the foo pipeline is filtered to only show on the cheese space!

How do I clean-up all apps and pipelines in my OpenShift Online

In OpenShift Online, go to the Help question mark (?) in the upper right.

  1. Select "Command Line Tools"
  2. If you have not done so already download the oc binary for your platform via the link on that page and put it on your $PATH environment variable
  3. Use the copy icon to copy the first command to get your token
  4. Paste that into a terminal
  5. Download the gofabric8 binary for your platform and add the gofabric8 binary to your $PATH environment variable
  6. Type the following command:
gofabric8 tenant delete

This will remove all pipelines, jenkins jobs and deployed apps. It will also clean out the various Persistent Volumes in your account for the jenkins jobs, the content repository (where released and versioned artifacts go like pom.xml files, jars, kubernetes manifests etc) and the local maven repository used as a persistent cache to speed up builds and reduce noise in build logs.

You will then need to Update Tenant to update your tenant services.

You can check the status of your tenant via:

gofabric8 tenant check

If you wish to just clean up different parts of your tenant you can try the more fine grained clean commands:

gofabric8 clean jenkins
gofabric8 clean content-repo
gofabric8 clean maven-local-repo
gofabric8 clean apps

You can use the help to get more information:

gofabric8 help clean

Which version should my Jenkins be?

To find out the intended version, visit http://central.maven.org/maven2/io/fabric8/online/packages/fabric8-online-jenkins/1.0.167/fabric8-online-jenkins-1.0.167-openshift.yml and look for the section:

provider: fabric8
      project: jenkins-openshift
      version: 3.0.37

Re-construct the above URL by replacing 1.0.167 with the value in the file "TEAM_VERSION" in https://github.com/fabric8io/fabric8-init-tenant/commit/COMMIT_HASH where COMMIT_HASH is the hash in https://github.com/openshiftio/saas/blob/master/dsaas-services/f8-tenant.yaml#L2

Whenever my build completes, it starts another one

This could have happened if lots of builds were triggered and never got provisioned due to quotas and there's lots of Build resources left over.

Delete all the build resources by doing a oc delete build --all

Eclipse Che Questions

My Che workspace isn't starting when I hit the "Create Workspace" button

Ensure that your tenants are up-to-date (see the account section at the head of this page). If it still doesn't work contact us.

I'm not getting auto-complete for my Java project

Right-click on the project in the Project Explorer and select "Configuration" from the Project menu item. Ensure that you are configured as a Maven project, not Blank.

How do I clean-up all Che workspaces in my OpenShift Online

From a running workspace click on the gray arrow on the top left corner of the workspace (or go to URL https://che-${ID}-che.8a09.starter-us-east-2.openshiftapps.com/dashboard). The Che dashboard will open. Select the "Workspaces" page in the left menu, then you can select all the workspaces you don't need and delete them.

Which version should my Che be?

To find out the intended version, open a shell and execute the following command:

curl -sSL http://central.maven.org/maven2/io/fabric8/online/apps/che/maven-metadata.xml | grep latest | sed -e 's,.*<latest>\([^<]*\)</latest>.*,\1,g'

How do I retrieve Che logs?

There are at least 3 interesting Che logs you can look at if you are looking for details of Che bahviour:

  • Che server (wsmaster) logs
  • Che workspace (wsagent) logs
  • Recommendation (bayesian) language server logs

Here are some commands that should help:

ID=<your osio id>
TOKEN=<your oso token>
oc login https://api.starter-us-east-2.openshift.com --token=${TOKEN}
oc project ${ID}-che

# get che server logs
MASTER_POD_ID=$(oc get pods -o=custom-columns=NAME:.metadata.name --no-headers | grep che-[0-9])
oc logs ${MASTER_POD_ID}

# get workspace logs
WS_POD_ID=$(oc get pods -o=custom-columns=NAME:.metadata.name --no-headers | grep che-ws)
oc rsh ${WS_POD_ID} \
     sh -c "cat /home/user/che/ws-agent/logs/catalina-0.log"

# get bayesian LS logs
oc rsh ${WS_POD_ID} \
     sh -c "cat /home/user/che/ls-bayesian/bayesian.log"
Clone this wiki locally