oc-cluster destroy failed for reason "Permission denied".

[warmchang]

Run oc-cluster destroy failed for reason "Permission denied":

[root@appab-myproject ~]# oc-cluster list
# Using client for origin v1.5.1
Profiles:
- example
- workshop
- workshop2
[root@appab-myproject ~]# oc-cluster status
# Using client for origin v1.5.1
no cluster running
[root@appab-myproject ~]# oc-cluster destroy workshop2
# Using client for origin v1.5.1
Are you sure you want to destroy cluster with profile <workshop2> (y/n)? y
Removing profile workshop2
Removing /root/.oc/profiles/workshop2

/usr/bin/rm: cannot remove '/profiles/workshop2': Permission denied
Removing .kubeconfig profiles
error: cannot delete cluster workshop2, not in /root/.kube/config
error: cannot delete context workshop2, not in /root/.kube/config
Removing 0 images built with this cluster
[root@appab-myproject ~]#
[root@appab-myproject ~]# oc-cluster list
# Using client for origin v1.5.1
Profiles:
- example
- workshop
- workshop2

The docker version & OS version:

[root@appab-myproject ~]# docker version
Client:
 Version:         1.12.6
 API version:     1.24
 Package version: docker-1.12.6-28.git1398f24.el7.centos.x86_64
 Go version:      go1.7.4
 Git commit:      1398f24/1.12.6
 Built:           Fri May 26 17:28:18 2017
 OS/Arch:         linux/amd64

Server:
 Version:         1.12.6
 API version:     1.24
 Package version: docker-1.12.6-28.git1398f24.el7.centos.x86_64
 Go version:      go1.7.4
 Git commit:      1398f24/1.12.6
 Built:           Fri May 26 17:28:18 2017
 OS/Arch:         linux/amd64
[root@appab-myproject ~]# cat /etc/redhat-release
CentOS Linux release 7.3.1611 (Core)
[root@appab-myproject ~]#

Any advice to resolve this? thanks.

[warmchang]

After check the script, I found the inconsistency between "/root/.oc/profiles/" and "/root/.kube/config":

[root@appab-myproject ~]# ll /root/.oc/profiles/ total 0 drwxr-xr-x. 7 root root 124 Jun 30 12:40 example drwxr-xr-x. 7 root root 124 Jun 30 13:28 workshop drwxr-xr-x. 7 root root 124 Jun 30 13:50 workshop2

the /root/.kube/config has no info about workshop2.

This is caused by the forced shutdown before the "oc-cluster up workshop2" finished executing. Delete the workshop2 fouder and rerun oc-cluster up, then all thing is ok.

[warmchang]

Rerun the oc-cluster up in the above condition, can saw the following error:

[root@appab-myproject ~]# oc-cluster up workshop2
# Using client for origin v1.5.1
[INFO] Running a previously created cluster
oc cluster up --version v1.5.1 --image openshift/origin --public-hostname 127.0.0.1 --routing-suffix apps.127.0.0.1.nip.io --host-data-dir /root/.oc/profiles/workshop2/data --host-config-dir /root/.oc/profiles/workshop2/config --host-pv-dir /root/.oc/profiles/workshop2/pv --use-existing-config -e TZ=CST
-- Checking OpenShift client ... OK
-- Checking Docker client ... OK
-- Checking Docker version ... OK
-- Checking for existing OpenShift container ... OK
-- Checking for openshift/origin:v1.5.1 image ... OK
-- Checking Docker daemon configuration ... OK
-- Checking for available ports ...
   WARNING: Binding DNS on port 8053 instead of 53, which may not be resolvable from all clients.
-- Checking type of volume mount ...
   Using nsenter mounter for OpenShift volumes
-- Creating host directories ... OK
-- Finding server IP ...
   Using 192.168.10.130 as the server IP
-- Starting OpenShift container ...
   Starting OpenShift using container 'origin'
   Waiting for API server to start listening
   OpenShift server started
-- Removing temporary directory ... OK
-- Checking container networking ... OK
-- Server Information ...
   OpenShift server started.
   The server is accessible via web console at:
       https://127.0.0.1:8443

   To login as administrator:
       oc login -u system:admin

-- Permissions on profile dir fixed
error: no context exists with the name: "workshop2".
[root@appab-myproject ~]# oc-cluster list

I think can we fix the error above by script automatic? For example, adjust the order of the two commands. Run "${OC_BINARY} adm config use-context xxx" first, check the xxx context:

• context xxx is ok, then run the "oc cluster up ...":
• context xxx is wrong, no need to run the "oc cluster up ...", just print the error info & quit.

Or can do like this: If context xxx is wrong and the foulder "/root/.oc/profiles/xxx" is exist, we can just backup it (/root/.oc/profiles/xxx_backup) and renew the xxx, just like to make a new profile.

How about your opinion? Hope to hear the feedback from the contributors. Thanks.