Need help to setup Eclipse-che 7 on Openshift 3 #15481
Labels
area/install
Issues related to installation, including offline/air gap and initial setup
kind/question
Questions that haven't been identified as being feature requests or bugs.
lifecycle/stale
Denotes an issue or PR has remained open with no activity and has become stale.
Comments
svrajasekar
added
the
kind/question
che-bot
added
the
status/need-triage
tolusha
added
area/install
|
hi @svrajasekar Have you check the installation guide? Also it will be very helpful if you provide any logs. |
|
Pls fill in form to have more clear picture of what is going on. |
Eclipse che server installation on CentOS 7 timesout.
Che server in multiuser mode must be up and running
Client Version: version.Info{Major:"1", Minor:"11+", GitVersion:"v1.11.0+d4cacc0", GitCommit:"d4cacc0", GitTreeState:"clean", BuildDate:"2018-10-10T16:38:01Z", GoVersion:"go1.10.3", Compiler:"gc", Platform:"linux/amd64"}
oc v3.11.0+0cbc58b Server https://127.0.0.1:8443
chectl server:start --platform=openshift --domain=127.0.0.1.nip.io --multiuser
events |
|
Keycloack didn't start properly. Could you show its logs https://www.eclipse.org/che/docs/che-7/viewing-external-service-logs/#viewing-keycloak-server-logs_viewing-keycloak-logs |
keycloack logs |
|
it seems fine. |
|
Since che server log is pretty huge. I will be sending it in parts. |
Che Server Logs |
Che Server Logs |
Che Server Logs |
|
a wrong route @ibuziuk |
|
Please help me out. I need to desperately set this up. I have no clue as to what to do next. I tried with Minikube, Minishift, Kubernetes and Openshift. Everywhere I faced timeout issue. I am using CentOS 7 VM with 16 gig ram and 4 vcpus. |
|
@svrajasekar If you are running in minishift/minikube, |
|
I am using Openshift 3.11 only and not minishift. Please help me to install Eclipse-che successfully on my host. I even assigned the --public-hostname to the public ip. But no use. Today I tried to install using che-operator. Except che-server pod, all other pods work properly. The che pod times out. Also if you looked at the log that I had sent, you could see a lot exceptions relate to jgroups and ch.qos.logback. It throws illegalstate exception. I don't know to proceed further. I need to set this up asap to my company. |
|
@svrajasekar You can see in the Che server logs The Che server is trying to communicate with keycloak by pinging localhost, which will obviously not work since keycloak is a separate Pod on the cluster. E.g. on my cluster: In your cluster you should see something like Step "Preparing OpenShift 3 for installing Che" in the docs covers retrieving your |
|
@svrajasekar I'm having the exact same issue. @amisevsk The domain is set - 127.0.0.1.nip.io I noticed that when I hit the endpoint from the error logs: http://keycloak-che.127.0.0.1.nip.io/auth/realms/che/.well-known/openid-configuration I get the same error. However, when I update the URL to point to port 8080 on the endpoint, it does seem to retrieve json data. I updated the the keycloak service to also forward port 80 to 8080 which enables the original url to return the json. However, the sever still fails. |
|
Yes I have been banging my head on this issue. Today I brought up Openshift 3 like this It came up successfully I then created a project named "che" I executed the below command to install Eclipse/che chectl server:start --platform=openshift --installer=operator --domain=osmaster.maantt.com --multiuser Still no luck. It times out. On top of it, I had a look at the logs, it still shows http://keycloak-che.127.0.0.1.nip.io I don't know why? You need to help me out. I am desperate and I have to set it up fast. I have already spent a week on this. |
|
Hi Atleast this could save my face in the organization. Please help me out. |
|
@svrajasekar could you try > oc cluster up --public-hostname=osmaster.maantt.com --routing-suffix=osmaster.maantt.com.nip.io |
|
Tried all the combos. Nothing seem to work. Keycloak does not seem to start properly though it's status shows as "running". Because of the above issue, che server fails to start. Please let me know any old stable version that can be used straight from Docker instead of Kubernetes or Openshift All the old versions starting 6.1.0 till 6.19.0 fails because these point to keycloak.openshift which is not in the repository. I will have to explore alternative to eclipse/che I guess. I have already waster 10 days on this. |
|
I have the following entries in /etc/hosts file 192.168.70.101 che-che.chemaster.maantt.com I am able to ping using the hostname. But chectl reports "UnknownHostException" for keycloak-che.chemaster.maantt.com May I know why? |
|
I tried to use https://keycloak-che.<PUBLIC-IP>/auth in the CHE_KEYCLOAK_AUTH__SERVER__URL. |
|
Are we talking about minishift? |
|
No, it is Openshift 3.11 |
|
I tried with https://keycloak-che.{PUBLIC-IP}.nip.io/auth as well. Throws UnknownHostException |
|
It only connects to Cluster IP |
|
Atlast I could successfully install Eclipse Che using external keycloak server. But when I launch eclipse che in the browser, I don't find any items in the stacks page Why? |
|
I have been asking for help for more than a week. But no one seems to suggest a positive step to resolve my concerns. Everybody is saying "try this" and "try that". |
|
@svrajasekar Yes, sorry about that. Nobody really understands the installation you have and what is wrong. If you send me an email - we could try to organize a quick meeting if it works for you |
|
Atlast I could successfully install Eclipse Che using external keycloak server. But when I launch eclipse che in the browser, I don't find any items in the stacks page Can you help me out? |
|
I have no idea. Maybe HAR could show more |
|
Can't start a new workspace. Complains that it is unable to install plugins due to timeout. I even hosted the local che-plugin-registry built using --offline flag. Still no use. I need your help on this. One issue after another pops up. Don't know when it is going to function properly. My VM is hosted in AWS. It has 32 gig ram with 8 vCpus. I installed Openshift 3.11 successfully and could start it properly. Regards |
|
Hi,
Review for installing oc and chectl:
https://www.eclipse.org/che/docs/che-7/installing-che-on-openshift-3-using-the-operator/
See more bug in your case:
https://docs.openshift.com/container-platform/3.11/release_notes/ocp_3_11_release_notes.html#ocp-3-11-43
…On Sun, Dec 29, 2019, 19:58 Rajasekar S V ***@***.***> wrote:
Can't start a new workspace. Complains that it is unable to install
plugins due to timeout. I even hosted the local che-plugin-registry built
using --offline flag. Still no use. I need your help on this. One issue
after another pops up. Don't know when it is going to function properly.
My VM is hosted in AWS. It has 32 gig ram with 8 vCpus.
I installed Openshift 3.11 successfully and could start it properly.
Regards
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#15481>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AEYAML4YU657GIEG62S4BULQ3CNHBANCNFSM4J23M2EQ>
.
|
|
I don't know how many times to iterate. I followed the instructions stated in the Che Getting Started Guid step by step. Still I faced and still facing issues. Nobody has helped me so far. Everyone says "Try this" and "Try that". But no clear instruction. There is a bug in chectl tool. When it starts up Eclipse Che, every other pod starts successfully. But not che pod. The reason: Che pod tries to communicate to Keycloak pod using the published route. It will not work. Because no 2 pods within a cluster can communicate with each other over public url. To address this, I had to install Keycloak server separately outside the cluster in my host and point this url to CHE_KEYLOAK_AUTH__SERVER__URL of the che pod. This solved the timeout problem, Also we can's install devfile and plugin registries inside the cluster as it will hit the same problem stated above. Either host it outside the cluster or use the public url. This solved the 2nd problem The last problem I am facing now is that the workspace is started because of timeout issues. Which means Eclipse Che is trying to download the plugin images. But somewhere it times out. I am yet get the help for this. Regards |
|
Sorry for the type: The last problem I am facing now is that the workspace is NOT started because of timeout issues. Which means Eclipse Che is trying to download the plugin images. But somewhere it times out. I am yet get the help for this. |
|
Hi,
Could you wonder permission for get image!?
…On Sun, Dec 29, 2019, 23:16 Rajasekar S V ***@***.***> wrote:
Sorry for the type:
The last problem I am facing now is that the workspace is NOT started
because of timeout issues. Which means Eclipse Che is trying to download
the plugin images. But somewhere it times out. I am yet get the help for
this.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#15481>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AEYAML2Q6GKASWM55HTV5CLQ3DENDANCNFSM4J23M2EQ>
.
|
|
After I launched eclipse che and created a workspace using one of the available stacks, it timed out stating "Plugin installation timed out" Next time, when the workspace was being started, I looked at the che-init-plugin-broker docker logs: docker logs f4811284c186 If you notice the ws:// endpoint, it strips the routing suffix and tries to open the socket. It is where it times out and fails. Any solution to this problem? |
|
Hi,
some doc to refer for troubleshooting.
https://technology.amis.nl/2018/10/28/quick-start-with-eclipse-che-browser-based-ide-running-on-docker/
Dont forget update lastest repo.
…On Wed, Jan 1, 2020, 02:45 Rajasekar S V ***@***.***> wrote:
After I launched eclipse che and created a workspace using one of the
available stacks, it times out stating "Plugin installation timed out"
Next time, when the workspace was being started, I looked at the docker
logs:
docker logs f4811284c186
2019/12/31 19:37:42 Broker configuration
2019/12/31 19:37:42 Push endpoint: ws://che-che./api/websocket
2019/12/31 19:37:42 Auth enabled: true
2019/12/31 19:37:42 Runtime ID:
2019/12/31 19:37:42 Workspace: workspacetaac25syghuc0u6y
2019/12/31 19:37:42 Environment: default
2019/12/31 19:37:42 OwnerId: 3d5836ba-ba9a-4a53-8732-c6e5db1b6e25
2019/12/31 19:37:42 Couldn't connect to endpoint
'ws://che-che./api/websocket', due to error 'dial tcp: lookup che-che. on
172.30.0.2:53: server misbehaving'
If you notice the ws:// endpoint, it strips the routing suffix and tries
to open the socket. It is where it times out and fails.
Any solution to this problem?
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#15481>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AEYAMLY27CNTGJ5YIXO6IQDQ3OOPBANCNFSM4J23M2EQ>
.
|
|
@IveJ Thanks for the troubleshooting doc. But the steps shown in the doc talks about starting eclipse che in a docker container. The old way. Now that Eclipse Che latest version cannot be started without either Kubernetes or Openshift. Second, the doc only refers to single user che. I didn't face any problems with Che 6.19.0 starting in single user mode. The question I asked is why Eclipse Che strips the routing suffix from the exposed route url when trying to open socket connection? Especially the push endpoint. Please help me. I am desperate. |
|
Hi,
As you can see on doc che-plugin-docker, this broker runs as an init
container on the workspace pod. Its job is to take in a list of plugin
identifiers (either references to a plugin in the registry or a link to a
plugin meta.yaml) and ensure that the correct .vsix and .theia extenions
are downloaded into the /plugins directory, for each plugin requested for
the workspace.
And more understand when you look at in dockerfile backward in registry.
So think of timeout occur for unsufficent extension during download.
https://github.com/eclipse/che-plugin-broker
https://www.eclipse.org/che/docs/che-7/che-workspaces-architecture/
…On Wed, Jan 1, 2020, 10:03 Rajasekar S V ***@***.***> wrote:
@IveJ <https://github.com/IveJ> Thanks for the troubleshooting doc. But
the steps shown in the doc talks about starting eclipse che in a docker
container. The old way. Now that Eclipse Che latest version cannot be
started without either Kubernetes or Openshift.
Second, the doc only refers to single user che. I didn't face any problems
with Che 6.19.0 starting in single user mode.
The question I asked is why Eclipse Che strips the routing suffix from the
exposed route url when trying to open socket connection?
Especially the push endpoint.
Please help me. I am desperate.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#15481>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AEYAML2RQG4ACX7ALEN2TFTQ3QBWVANCNFSM4J23M2EQ>
.
|
|
@IveJ Please note that I have not contributed anything to Eclipse Che development. I have been trying hard to setup for my company. It is going to be 3 weeks and still I have not got any help. What you are doing is, sending me the link to documentation. What I want is a solution. Please let me know if you can provide a solution or not? I can send you whatever the information you want. I have achieved 90% out of my own initiative. Atleast on this last issue, I hope you can help me out. |
|
Hi Raja,
Follow quick start guide in doc for your enviroment:
https://www.eclipse.org/che/docs/che-7/che-quick-starts/
Notice to your project deployment, with ssl option, https or wss. Url as
path your project.
…On Wed, Jan 1, 2020, 16:16 Rajasekar S V ***@***.***> wrote:
@IveJ <https://github.com/IveJ> Please note that I have not contributed
anything to Eclipse Che development. I have been trying hard to setup for
my company. It is going to be 3 weeks and still I have not got any help.
What you are doing is, sending me the link to documentation. What I want
is a solution.
Please let me know if you can provide a solution or not?
I can send you whatever the information you want.
I have achieved 90% out of my own initiative. Atleast on this last issue,
I hope you can help me out.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#15481>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AEYAML2QXXB4W5VV7ER37NTQ3RNNJANCNFSM4J23M2EQ>
.
|
|
@svrajasekar
Sorry that you're still stuck at the same place too. This would have been a good tool for my development team but there just doesn't seem to be a fix right now for implementation on Openshift 3. The odd thing for me is that it installs fine on a Minishift deployment in a similar environment.
Are you able to deploy it using Minishift as well?
|
|
@svrajasekar I finally got it to run on the Openshift 3.11 cluster. I'm not sure if this helps or why it worked but it possibly could work for you as well. I've tried several different combinations of the the First I launched the cluster with This just kept forwarded me to Afterwards I took the cluster down with: Finally, I relaunched the cluster with: This is similar to what you tried previously. However, when I tried to reach the console at From there I was able to setup my users, roles, and the che project and then install the che7 as in the documentation. Again, sounds trivial, but it worked for me. Also note that chectl has updated since you opened this issue. I just ran: Good luck! |
|
@tag-ii Could you create a workspace?. If not, please try that. Also as which user did you login to web console?. developer?. Is it through the console that you created users, roles and project che? As developer user, did it allow you to create users and roles?. Before typing chectl ...., did you login as developer or system:admin? Pls don't think I am asking too many questions. I have been banging my head on this for more than 3 weeks. My company is steam-rolling me everyday asking for status. Please respond, |
|
Ok, so as far as I could get yesterday, the che server pod crashed in the
process of starting a workspace.
I initially logged into the web console as user just to verify it was
working. I then logged into oc as system:admin and created another user and
assigned it the cluster admin role. That user is the one I then used to
install che with chectl.
No problem with the questions. We're all trying to get it going.
I'm just getting back to it now myself. I'll update here if I'm able to
discover what caused the che server to crash or if I'm able to get a
workspace going.
… |
|
Issues go stale after Mark the issue as fresh with If this issue is safe to close now please do so. Moderators: Add |
che-bot
added
the
lifecycle/stale
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I have been trying hard to setup Eclipse-che on Openshift 3 over the last few days. But could not successfully do it.
I tried in Minikube, Minishift, Kubernetes and Openshift. It failed in all.
Che-server pod crashes in multiuser mode. I tried with various versions of Openshift and Eclipse che. But no use.
I am able to start Eclipse-che in single user mode. But in multiuser mode, che-server pod crashes.
I need help desperately.
The text was updated successfully, but these errors were encountered: