Failed Install and odd STDOUT Messages about installcentos HOT 16 CLOSED

On line 142 of the install-openshift.sh script the envsubst command substitutes the wrong values into inventory.ini

The following workaround worked for me:

1. Download install-openshift.sh
2. Edit install-openshift.sh and comment these lines:
   echo "* Your domain is $DOMAIN "
   echo "* Your IP is $IP "
   echo "* Your username is $USERNAME "
   echo "* Your password is $PASSWORD "
   echo "* OpenShift version: $VERSION "
   #echo "******"
3. run install-openshift.sh from the local file system.

P.S It's not a fix, just a crude workaround.

from installcentos.

not sure why it's happening, but have changed the way the switch between interactive/non-interactive is done and seems to work now. Managed to deploy cluster, but Logging failed ... will keep investigating tomorrow.

from installcentos.

Thanks for sharing notes! I'm doing further poking around, and it seems to possibly have something to do with how piping to | /bin/bash works (or rather doesn't). I'm still unsure why...

from installcentos.

Does the rest of the script work for you, @koksj? The underlying ansible-playbook is hitting this error and the script concludes as follows:

TASK [openshift_master_certificates : Create the master server certificate] ******************************************************************************************************************************

TASK [openshift_master_certificates : Generate the loopback master client config] ************************************************************************************************************************
ERROR! |failed expects hostvars is a dict
htpasswd: cannot modify file /etc/origin/master/htpasswd; use '-c' to create it
install-openshift.sh: line 159: oc: command not found
Failed to restart origin-master-api.service: Unit not found.
* Your console is https://console.97.116.1.1.nip.io:8443
* Your username is kentos
* Your password is fuzzypickles
* Login using:
$ oc login -u kentos -p fuzzypickles https://console.97.116.1.1.nip.io:8443/
install-openshift.sh: line 170: oc: command not found

I wonder if this project would have better luck locking down the install to specific git commits on it's dependent software projects.

from installcentos.

@IRobL I get the exact same error. Spent all day trying to get past it. No such luck yet.

from installcentos.

On the second error, I'm following up here and making some progress... albeit towards other errors.

The following commands use the patch mentioned in that thread, but sadly I am halted on another, possibly unrelated error :(

sudo su
export PASSWORD='changeme123'
export USERNAME='john'
export DOMAIN='97.116.33.228.nip.io'

curl https://raw.githubusercontent.com/irobl/installcentos/master/install-openshift.sh > install-openshift.sh

chmod 0775 install-openshift.sh
sed -i '/^echo "\**"$/d' install-openshift.sh # just in-case

/bin/bash install-openshift.sh -a

Notice it's relying on these two patched up forks in the hopes to produce stability:

• https://github.com/IRobL/installcentos
• https://github.com/IRobL/openshift-ansible

from installcentos.

@IRobL Was seeing this exact error. However it has only just crept in over the last week, as this script worked perfectly last week for a fresh Centos 7 install. I think that's because Ansible has upgraded itself from 2.4 to 2.6 somewhere, either in the main package repos of Centos or as a dependency somewhere else. If you haven't already you should maybe submit a pull request to the openshift ansible repo as this patch seems to have fixed it for me.

from installcentos.

OK, with explicitly stating the variable in the inventory file, the Logging installation finished fine. Running

curl https://raw.githubusercontent.com/gshipley/installcentos/master/install-openshift.sh | INTERACTIVE=false /bin/bash

on a VM with 16G memory gives me a complete and functional deployment.

from installcentos.

Just tried to deploy Openshift on my VM but I get this error now:

FAILED - RETRYING: wait for api server to be ready (60 retries left). FAILED - RETRYING: wait for api server to be ready (59 retries left). FAILED - RETRYING: wait for api server to be ready (58 retries left). FAILED - RETRYING: wait for api server to be ready (57 retries left). FAILED - RETRYING: wait for api server to be ready (56 retries left). FAILED - RETRYING: wait for api server to be ready (55 retries left). FAILED - RETRYING: wait for api server to be ready (54 retries left). FAILED - RETRYING: wait for api server to be ready (1 retries left). fatal: [10.100.0.80]: FAILED! => {"attempts": 60, "changed": false, "content": "", "msg": "Status code was -1 and not [200]: Request failed: <urlopen error [Errno 111] Connection refused>", "redirected": false, "status": -1, "url": "https://apiserver.kube-service-catalog.svc/healthz"} to retry, use: --limit @/root/openshift-ansible/playbooks/deploy_cluster.retry

PLAY RECAP ********************************************************************* 10.100.0.80 : ok=926 changed=333 unreachable=0 failed=1 localhost : ok=13 changed=0 unreachable=0 failed=0

INSTALLER STATUS *************************************************************** Initialization : Complete (0:00:10) Health Check : Complete (0:00:01) etcd Install : Complete (0:01:15) Master Install : Complete (0:04:53) Master Additional Install : Complete (0:00:28) Node Install : Complete (0:02:16) Hosted Install : Complete (0:01:13) Web Console Install : Complete (0:01:26) Metrics Install : Complete (0:01:36) Logging Install : Complete (0:02:24) Service Catalog Install : In Progress (0:12:12) This phase can be restarted by running: playbooks/openshift-service-catalog/config.yml

Failure summary: Hosts: 10.100.0.80 Play: Service Catalog Task: wait for api server to be ready Message: Status code was -1 and not [200]: Request failed: <urlopen error [Errno 111] Connection refused> Adding password for user marko cluster role "cluster-admin" added: "marko"

Your console is https://console.internal.mkgraz.ovh:8443 Your username is marko Your password is xxxxxxxxx

from installcentos.

@marscydonia that can be many things, what does syslog say?

from installcentos.

@marekjelen I have tested the script and it worked for me.
P.S
A big thank you, to you and the other contributors to make this script possible.

from installcentos.

Hi all,

any solution for this? After changing the install-openshift.sh script (remove the echo "**" lines) I ran into the oc: command not found error.

The two errors during the playbook were:

PLAY [Ensure that all non-node hosts are accessible] ***************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************

TASK [Gathering Facts] *********************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************
fatal: [xx.xx.xx.xx]: UNREACHABLE! => {"changed": false, "msg": "Failed to connect to the host via ssh: Permission denied (publickey,gssapi-keyex,gssapi-with-mic,password).\r\n", "unreachable": true}

from installcentos.

@devops-42

Try downloading the script to your local file system. Don't make any changes to the script. Also download vol.yaml to your local file system.

Then follow the automaton instructions as detailed in
https://github.com/gshipley/installcentos Readme. Excepect run the script from your local file system.

Just for the record running it from curl worked for me after @marekjelen made his fixes.

from installcentos.

@koksj Thanks, but new errors occur :( Have proceeded as suggested:

curl https://raw.githubusercontent.com/gshipley/installcentos/master/install-openshift.sh > install-openshift.sh
curl https://raw.githubusercontent.com/gshipley/installcentos/master/vol.yaml > vol.yaml

export DOMAIN=172......nip.io 
export USERNAME=osadmin
export PASSWORD=********
export VERSION="v3.9.0"
chmod +x install-openshift.sh 
INTERACTIVE=false ./install-openshift.sh 

leads to:

TASK [Evaluate oo_etcd_to_migrate] *****************************************************************************************************************************************************************************************************************************************
ok: [localhost] => (item=172........)
 [WARNING]: Could not match supplied host pattern, ignoring: oo_lb_to_config

 [WARNING]: Could not match supplied host pattern, ignoring: oo_nfs_to_config


PLAY [Ensure that all non-node hosts are accessible] ***********************************************************************************************************************************************************************************************************************

TASK [Gathering Facts] *****************************************************************************************************************************************************************************************************************************************************
fatal: [172........]: UNREACHABLE! => {"changed": false, "msg": "Failed to connect to the host via ssh: Permission denied (publickey,gssapi-keyex,gssapi-with-mic,password).\r\n", "unreachable": true}

NO MORE HOSTS LEFT *********************************************************************************************************************************************************************************************************************************************************
	to retry, use: --limit @/home/*****/openshift-ansible/playbooks/prerequisites.retry

PLAY RECAP *****************************************************************************************************************************************************************************************************************************************************************
172........               : ok=1    changed=0    unreachable=1    failed=0   
localhost                  : ok=11   changed=0    unreachable=0    failed=0   


INSTALLER STATUS ***********************************************************************************************************************************************************************************************************************************************************
Initialization             : In Progress (0:00:01)

with final error:

htpasswd: cannot modify file /etc/origin/master/htpasswd; use '-c' to create it
./install-openshift.sh: line 155: oc: command not found
Failed to restart origin-master-api.service: Unit not found.
******

from installcentos.

@devops-42

Hi

I ran the flowing on a Centos 7 Virtual Box VM with success as root:
8 gb ram
2 x vcpu

yum -y install vim wget (my personal preferences)
wget https://raw.githubusercontent.com/gshipley/installcentos/master/install-openshift.sh
wget https://raw.githubusercontent.com/gshipley/installcentos/master/vol.yaml

export DOMAIN= mydomain.io
export USERNAME=xxxx
export PASSWORD=xxx
export DISK="/dev/sdb" ( I added a disk to the vm)
export INTERACTIVE=true (just for kicks)

chmod +x install-openshift.sh
./install-openshift.sh

from installcentos.

@koksj

Problem found :) The user account could not login via public key, so the playbook stopped.

Thanks for your help!

from installcentos.