Self Service VM Provisioning with Morpheus, Part 1: Solving Some Existential Crises

Here in Part 1, I hope to answer the question, “why Morpheus?” with some dark existentialism along the way. . . . And I finally get to go on a rant I have been dying to unleash for quite some time.

In Part 2, I will get more technical and show you some automated VM provisioning with Morpheus. I haven’t written it yet, so I have no idea if it will live up to my usual “hot take” you have come to expect here at thinkingoutcloud.org. You’ll have to wait, 43 official followers of my blog.

The Existential Nightmare of Humans Creating VMs:

Where I work, we are moving away from a “traditional” VM creation model, because you know how that goes with humans:

1. An End User creates a VM Request Ticket for a server in ServiceNow, where it sits for a week in someone’s queue.
2. The human who is supposed to be making his or her way through this queue goes on a team-building retreat. You know, maybe some mini golf . . . a few trust falls . . . some activities that require a couple of those inflatable sumo wrestling suits . . .
3. Upon his or her return, the VM is created, but it’s still not ready because . . . security, which means . . .
4. More humans. . . .
5. Then the human who is supposed to be shepherding this VM along gets all existential . . . you know . . . is life really worth it? Does the happiness in life really outweigh the suffering? Is it really worth suffering through this absurd notion we call life?
6. I MEAN WHAT’S IT ALL FOR? WHAT DID NIETZSCHE MEAN BY, “DIE AT THE RIGHT TIME”? WHAT DOES THAT MEAN?
7. I MEAN, WAS HE ADVOCATING SUICIDE, OR WAS HE JUST SAYING THAT ONE’S DEATH SHOULD BE MEANINGFUL? . . . BOTH MAYBE?
8. Nascent feelings of angst and ennui overwhelm this poor human . . . and the binge drinking starts . . . with some crying in the fetal position on the kitchen floor, otherwise known as “Tuesday Night” for me . . . .
9. . . . Then, finally this poor human gets around to finalizing the VM and granting the user access.
10. The original End User, who set this existential quandary into motion in the first place, is now also trapped in his or her own existential nightmare. But does it really matter anymore? THE IRONY OF IT ALL!

This can take up to 2 weeks, give or take, depending on the VM and depending on the human workload. What we are moving to instead is a model where all of the above has been pre-approved and automated:

1. User creates a VM request in ServiceNow.
2. The Morpheus robot performs and coordinates the automation and creates the VM automatically.
3. User receives an automated email with the hostname/IP address of their new VM with access already automatically granted.
4. Profit.

This can take up to 15 minutes, give or take. And no dark existentialism along the way. Quite an improvement I must say.

Why Morpheus?

It’s a bit of a long story, but another automation platform . . . that shall not be named . . . failed us. We switched to Morpheus as a Cloud Management Platform (CMP) that orchestrates VM creation and additionally kicks off existing methods of automation we’ve created previously. What Morpheus gives us:

1. On-premises cloud-adjacent features, such as t-shirt sizing and what can best be described as a more public cloud experience.
2. The flexibility to create VMs through ServiceNow (for end users), or create VMs directly on Morpheus for custom builds (for us server infrastructure types).
3. The flexibility to create public cloud infrastructure for multi-cloud, or hybrid cloud implementations.
4. Multiple integrations for our already-built automation, such as our Ansible builds.
5. Coded workflows for customized tasks.
6. Scalable automation that also allows for more robust automation beyond compute, such as network and security automation.
7. Automated, out-of-the-box application-layer builds, like NGINX, and so on.
8. A standard REST API that lends itself very well to our IaC approach.

Morpheus also has its own integrations for “EHRMEHGERD! KURBERNERTES!” but we’re not using it for that.

I have other plans for Morpheus down the road too: HPE OneView integration and ESXi imaging are items I would like to entertain, so there is no lack of options for us, as it stands right now.

There are other features Morpheus has that the Morpheus marketing people would love for me to mention. . . .

. . . Hey there, Morpheus marketing people . . . how are you? . . . .

But I haven’t even made it to my rant yet!

The most important feature, for us, is the Morpheus API. Eventually we plan on allowing our devs to have direct access to the Morpheus API so they can create their own VMs and applications . . . on-prem, at will, and without us infrastructure types getting in their way.

This is viewed as a crazy idea by some, but programmatic access to infrastructure is what end users expect in the year 2020.

Thanks, Public Cloud!

And Finally the Aforementoned Rant

You know those “traditional” on-prem Infrastructure Engineers who lament that, “everything is going into the cloud”? Fortunately, I don’t see them where I work, but you know the type: the people who roll their eyes every time they hear the phrase, “we’re thinking of using a SaaS solution.”

These are the same people who would probably chain themselves to the last cabinet in their datacenter and yell, “to the last cab!”

The reasons for this, I would imagine, is that deep down, they perceive “the Cloud” and its various approaches as a threat to their job security, and perhaps they have an aversion to change.

It would seem that these Engineers are quite literally having an existential crisis of their own.

The great irony of these very same Engineers, I have found, is that they are commonly the same people who, against their own interests, slow processes down on-prem by wanting to control everything. They take special pleasure in being part of the GAUCON (Grand And Useless Council Of “No”).

You can’t have it both ways. It is against your best interest to both lament migrations to the cloud while also being an impediment to cloud-adjacent on-prem features like self service. Your users aren’t going to put up with that kind of behavior anymore. They will move their apps to the cloud if they can, or work around you, despite your protestations.

Please understand, I am not saying that the Public Cloud is a panacea. It most certainly is not. But you can’t deny that it has changed the expectations of Applications-owners and Devs alike. Allowing your users to have the autonomy to create their own environs (even on-prem) is what “getting on board” looks like. Wouldn’t you rather the users enjoy the experience on-prem instead of the alternative?

Maybe Morpheus could be your first step toward that?

And before you criticize me for calling out my IT Engineering peers, I have had my own existential crises. Don’t we all in this line of work? I remember building my first Docker container in 2015. I figured right then and there that my days as an on-prem vSphere Admin were numbered. But I continue to adapt and get out of my comfort zone.

. . . And binge drink.

I hope you join me for Part 2, where I show you some basics about what I’m doing with Morpheus. Hit me up on twitter @RussianLitGuy or email me at bryansullins@thinkingoutcloud.org. I would love to hear from you.