Closed rktidwell closed 6 years ago
Thank you for pointing this out. We'll take a look.
@rktidwell I have been unable to reproduce in my lab. I have 3 VSDs and 2 VSCs deployed on CentOS 7.3 KVM. My Ansible is also on CentOS 7.3. I am testing Nuage version 5.1.1-U1. How does this compare to your environment? Will you provide your ansible.log file for triage?
I've gone back and tried replaying this, I'm wondering if that message is coming from my ansible host due to "connection: local" being used somewhere.
@bacastelli Just to follow up, I had earlier assumed that I was getting an error because of some string escaping problem when sending the command to the VSD. However, I think what is happening is that command is being invoked on my ansible host which of course does not have any of the ejabberd tools installed on it.
@rktidwell I have been unable to reproduce. Will you provide the ansible.log that shows the failure so that we can take the investigation deeper? If you need to rerun, please make sure to rerun with -vvvv
option to maximize information in the log. Thank you.
More than a month with no update. Going to close. Please re-open if additional triage materials are provided.
When running the vsp_postupgrade_health playbook, the vsd-health role attempts to run "/opt/ejabberd/bin/ejabberdctl list_p1db". This step fails because it appears to invoke the entire string as a single executable rather than a command with 1 parameter. This causes the health check to fail because with an error "command /opt/ejabberd/bin/ejabberdctl list_p1db cannot be found". If you log in to the VSD and issue the command it runs successfully. It appears there is just some escaping of the command string needed to iron this out.