When updating a image template and "staging" selected, would be great to have new feature to select a host pool as a automated process

Avatar
Fabian Rodrigues

When updating a image template and "staging" selected, would be great to have new feature to select a host pool as a automated process

So when I image template is updated, and staging is enabled

 

Have a check box "Deploy staged image to host pool" & when this is selected enable a dropdown where customer can select a appropiate host pool.

Logic here, dedicated host pool purely for UAT\OAT of image updates.

Yes we do have the host pool "Re-Size\Re-image" option, but this is promoting a more automated image test cycle for customers

7

Comments (8 comments)

Sort by
1
Avatar
Lars Bertulies

Fabian kindly posted this on our behalf as result of a support ticket. Reasoning behind this:

We have a dedicated group of key users/power users that serve as pilots before a new image is rolled out into production. To minimize impact on production, we have a separate host pool to which staged images are deployed. To fully automate this process, Fabian's idea came to be.

Right now we have scheduled a "Resize/Re-image" task 3 hours after the scheduled image creation, but we have no way of knowing if the image creation was successful or not, or even if it completed in time...

1
Avatar
Raul Morales

That is an interesting idea. I will notate this for our developers and add it to our backlog. Thank you.

1
Avatar
Raul Morales

Lars, I would like to have a complete picture of the request. I understand that you already have a "Resize/Re-image" task scheduled after the scheduled image creation. That is excellent and does achieve an automated flow of creating a desktop image and then applying it to the host pool. 
You are also able to receive alerts of any task failures using the Alerts and Notifications feature (Task Conditions): https://nmmhelp.getnerdio.com/hc/en-us/articles/25498222093709-Alerts-and-Notifications

Suppose we add a "Deploy staged image to host pool" option. In that case, it will most likely still require initiating the "Resize/Re-image" process within the host pool afterward, as specific parameters need to be set for it to commence (such as how many session hosts to image, force users to log off, etc.)

1
Avatar
Lars Bertulies

The initial idea was a fully fledged solution to enable conditional task / script / runbook execution depending on results of previous executions (or other events). As you mentioned, the event engine is already there.

Practical examples:

  1. Run scheduled image creation task with staging enabled (e.g. the day after patch tuesday)
  2. If succesful: Run "Resize/Re-Image" Task on test pool; create scheduled deployment to production when staging period runs out
  3. If unsuccessful: Don't run "Resize/Re-Image" on test pool; don't create scheduled deployment to production; send mail to test users that no testing is needed and to wait for further instructions...
  1. Run scheduled image creation for personal single session hosts
  2. If successful: Run "Resize/Re-Image" Task on unassigned personal hosts; leave assigned hosts untouched (assumption being: If a device is in use, rebuilding could incur data loss and it will be patched through WUFB anyway. If it is not in use, destroy and rebuild on latest patch level to reduce setup time on assignment)

...the possibilites would be endless and could take automation to another level ☺️

Right now we are working with schedules which have no way of knowing if it would make any sense to run at all. If image creation fails, resize/re-image is of no use. If image creation for some reason takes more time than anticipated, resize/re-image could even run before the new image is ready!

Fabian's suggestion was to scale the idea down to make it easier to implement: Fully automate it for staging in an easy and accessible way.

1
Avatar
Raul Morales

Thank you, Lars, for the detailed insight. This triggers-and-run idea would be very interesting to implement. I can see some complexities involved, but we will certainly explore this as it would allow many avenues of automation.

1
Avatar
Denis Quinlan

This would be a real benifit to us as I've just been talking to Fabian about how best to utilise Nerdio to automate our patching, testing and production deployment of our image.  The asigning of the staged image is still a manual step within what could be a fully automated process :)

1
Avatar
Travis King

Having the option to select a host pool during staging would really streamline the image update process and reduce manual steps for us. A dedicated UAT/OAT host pool for testing image updates makes a lot of sense and is something i've thought about in the past funny enough. Looking forward to seeing if this gets implemented, thanks for bringing this up!

0
Avatar
Dan Goss

This would be a huge step forward in automating the release cycle. 

Please sign in to leave a comment.