SELinux boolean virt_sandbox_use_nfs does not exist

[gnunn1]

I'm getting the below nfs errors when trying to setup OpenShift in us-west-2. I did find an issue for this in openshift-ansible (https://github.com/openshift/openshift-ansible/issues/2294) with a workaround of removing this part of the script.

However I asked a colleague who also uses these scripts if he was getting this issue and he indicated he wasn't. So I'm wondering if this really is an issue or something I'm doing wrong?

TASK: [openshift_node | Configure Node Environment Variables] ***************** 
skipping: [ec2-35-162-253-219.us-west-2.compute.amazonaws.com]
skipping: [ec2-35-165-37-187.us-west-2.compute.amazonaws.com]
skipping: [ec2-35-164-139-103.us-west-2.compute.amazonaws.com]
skipping: [ec2-35-165-29-250.us-west-2.compute.amazonaws.com]
skipping: [ec2-35-164-165-246.us-west-2.compute.amazonaws.com]

TASK: [openshift_node | Install NFS storage plugin dependencies] ************** 
ok: [ec2-35-164-139-103.us-west-2.compute.amazonaws.com]
ok: [ec2-35-164-165-246.us-west-2.compute.amazonaws.com]
ok: [ec2-35-162-253-219.us-west-2.compute.amazonaws.com]
ok: [ec2-35-165-37-187.us-west-2.compute.amazonaws.com]
ok: [ec2-35-165-29-250.us-west-2.compute.amazonaws.com]

TASK: [openshift_node | Check for existence of virt_use_nfs seboolean] ******** 
ok: [ec2-35-164-139-103.us-west-2.compute.amazonaws.com]
ok: [ec2-35-164-165-246.us-west-2.compute.amazonaws.com]
ok: [ec2-35-162-253-219.us-west-2.compute.amazonaws.com]
ok: [ec2-35-165-37-187.us-west-2.compute.amazonaws.com]
ok: [ec2-35-165-29-250.us-west-2.compute.amazonaws.com]

TASK: [openshift_node | Set seboolean to allow nfs storage plugin access from containers] *** 
changed: [ec2-35-164-139-103.us-west-2.compute.amazonaws.com]
changed: [ec2-35-162-253-219.us-west-2.compute.amazonaws.com]
changed: [ec2-35-164-165-246.us-west-2.compute.amazonaws.com]
changed: [ec2-35-165-29-250.us-west-2.compute.amazonaws.com]
changed: [ec2-35-165-37-187.us-west-2.compute.amazonaws.com]

TASK: [openshift_node | Check for existence of virt_sandbox_use_nfs seboolean] *** 
ok: [ec2-35-164-165-246.us-west-2.compute.amazonaws.com]
ok: [ec2-35-162-253-219.us-west-2.compute.amazonaws.com]
ok: [ec2-35-164-139-103.us-west-2.compute.amazonaws.com]
ok: [ec2-35-165-37-187.us-west-2.compute.amazonaws.com]
ok: [ec2-35-165-29-250.us-west-2.compute.amazonaws.com]

TASK: [openshift_node | Set seboolean to allow nfs storage plugin access from containers(sandbox)] *** 
failed: [ec2-35-164-165-246.us-west-2.compute.amazonaws.com] => {"failed": true}
msg: SELinux boolean virt_sandbox_use_nfs does not exist.
failed: [ec2-35-162-253-219.us-west-2.compute.amazonaws.com] => {"failed": true}
msg: SELinux boolean virt_sandbox_use_nfs does not exist.
failed: [ec2-35-164-139-103.us-west-2.compute.amazonaws.com] => {"failed": true}
msg: SELinux boolean virt_sandbox_use_nfs does not exist.
failed: [ec2-35-165-37-187.us-west-2.compute.amazonaws.com] => {"failed": true}
msg: SELinux boolean virt_sandbox_use_nfs does not exist.
failed: [ec2-35-165-29-250.us-west-2.compute.amazonaws.com] => {"failed": true}
msg: SELinux boolean virt_sandbox_use_nfs does not exist.

FATAL: all hosts have already failed -- aborting

PLAY RECAP ******************************************************************** 
           to retry, use: --limit @/home/gnunn/openshift_setup.retry

ec2-35-162-253-219.us-west-2.compute.amazonaws.com : ok=310  changed=74   unreachable=0    failed=1   
ec2-35-164-139-103.us-west-2.compute.amazonaws.com : ok=128  changed=30   unreachable=0    failed=1   
ec2-35-164-165-246.us-west-2.compute.amazonaws.com : ok=310  changed=74   unreachable=0    failed=1   
ec2-35-165-29-250.us-west-2.compute.amazonaws.com : ok=128  changed=30   unreachable=0    failed=1   
ec2-35-165-37-187.us-west-2.compute.amazonaws.com : ok=475  changed=114  unreachable=0    failed=1   
localhost                  : ok=24   changed=0    unreachable=0    failed=0 

[detiber]

@gnunn1 what version of Ansible are you using? The latest 2.2 build should resolve this issue.

[gnunn1]

I'm using 1.9.6 because the repo says that it is what is required, when I tried 2.2 I was getting an error with CloudFormations template as referenced in issue #159.

[detiber]

@gnunn1 ah, that is indeed going to be an issue. I'm not sure when I'll get a chance to take a look at how to clean that up, until then you may want to try this tool instead: https://github.com/sborenst/ansible_aws_deployer

[thoraxe]

Definitely use the repo that @detiber mentioned. This repo is essentially deprecated and no one is really putting much effort into it.

I have put a lot of recent effort into ansible_aws_deployer.

Outside of that, I would recommend you check out the OpenShift AWS reference architecture if you need an HA environment:

https://access.redhat.com/articles/2623521

[gnunn1]

Does the ansible_aws_deployer support installing OpenShift without the bu_workshop artifacts, I.e just do a fairly generic OpenShift installation? I'm not overly familiar with Ansible at this point but looking at the repo and it's Readme it seems fairly tied to that?

[thoraxe]

All the plays are tagged, so you could do something like:

..... --skip-tags=workshop

And this would not install any of the workshop infrastructure bits. You can inspect the tags in the bu-workshop playbook to verify.

You could also modify the bu-workshop playbook and simply remove all of the workshop bits yourself.

Would you mind if I close this issue, as we're deprecating this repo? Thanks for your interest!

[gnunn1]

Thanks for the answer, closing issue.