I am taking the Ansible training (D0407). Towards the end of the on-line class (section 12) I ran into unrecoverable errors. I opened a ticket and the support system said to "get a fresh set of VMs"... I have no idea what that means. There is a DELETE button on the online lab, but I can't find any definition of what that means (so, I'm a little to reluctant to hit the delete button now that I'm at the end of the lab, but not quite done).
Also, the review lab has you step back through previous labs, but (since I've already completed those labs) the files are already present on the workstation or remote hosts.
So... does anyone know how to get a "fesh set of VMs" or otherwise reset / reload the lab servers so you can start from scratch?
Listed in the 'Outcomes' section for the first exercise in the Comprehensive Review lab is:
"If you did not reset all of your virtual machines at the end of the last lab, do so now. This may take a few minutes."
At one point we had a section entitled Before you begin. I'll have to ask the content authors why we're not using that in this case, as this sentence is in the middle of several other things talking about what should be the case at the end of the lab. To reset all the systems, you could individually select the 'reset' action on each of them. A Delete will remove the entire lab environment, but you can Provision Lab back a new copy to get all fresh systems. Because there are 6 systems in this lab, I'd do the Delete/Provision for the sake of expediency. I think the support guys were suggesting you do a Delete/Provision to, as they put it, get a fresh set of VMs. I'll bring this up in my call with them next week; while we work with the platform every day, that is not always the case for users. Where possible, I'd like the support guidance to align with the actual process, and button labels, one should use to perform the task they think should be completed.
@DanK: A system reset is actually the following cloud API calls...
If running: Power off -> detach disk image -> attach new disk image from library -> start
If stopped: detach disk image -> attach new disk image from library
Really, we should go ahead and start the machines if they were initially stopped, I don't recall the reason we didn't initially. Maybe because we wanted to leave them in the same state as we found them (running labs charge time, where as stopped labs do not).
Everything except the machine start should happen pretty quickly, like within 30 seconds. Because it generally happens pretty quickly, the machine might be STARTED then transition to STARTING state because it stopped and swapped disk images in between polling intervals for the lab UI.
But realize that my idea of completing a 'reset' is doing all the cloud stuff to make a system boot from it's initial state disk again. The actual time it takes to boot that machine, or have services run on that machine (like all the OpenStack stuff) is beyond my ability to control. My bits of reset should take roughly 30-90 seconds. If you're reproduceably getting 15 minute timeframes from hitting the reset action to having the machine in STARTING state, you should have support take a look at that. If, instead, the majority of that 15 minutes is the machine in STARTED state and you're waiting for it to complete booting or start services ... if only there was someone on the content curriculum team you could address that with ... ;-)