In case you haven’t seen our previous post, here at InMotion we are reshaping OpenStack a bit from the traditional deployment. We have always worked with small and medium sized companies. No enterprises on our customer roster and just a few large companies.
So we have a different goal – enable the “non-enterprise” to easily provision OpenStack based infrastructure at a cost that enterprises can get due to their unique scale. We have released the first of our marketing content on our Infrastructure as a Service section of the site and we expect to be provisioning around November 17th.
With this macro goal in mind, in my last post I shared we have been aiming at these goals:
- Production grade OpenStack and Ceph on a 3 server footprint.
- Spun up in 20 minutes.
- Add additional servers in another 20 minutes.
- Private control plane.
- Charge by the hour.
- Keep the costs low and, absolutely, absolutely, beat the TCO of running on the mega clouds including your team costs if you manage your own private cloud with us.
Up until this release, the Team has been presenting individual parts of our overall system. As mentioned before, we are an “API first” company – so, the theory is that when we bring it together, it will just work, right?
So, Release 0.9.3!
Results in bold
Production grade OpenStack and Ceph on a 3 server footprint.
YES! We have Hyper-Converged OpenStack and Ceph on both a “Small” and a “Standard” setup.
Spun up in 20 minutes.
ALMOST! 37 minutes and there are quite a few health checks in the system for Dev purposes still. We do think that trying to get to 20 minutes may not be worth some of the trade-offs and some of the extra checks should remain. So we might just be satisfied with spinning up a fully production ready OpenStack in 30 minutes.
Add additional servers in another 20 minutes.
ALMOST! We are pretty confident we will be way below this goal, think a couple minutes, eventually.
Private control plane.
YES! A private control plane is required to truly have private infrastructure that a customer can modify as they see fit. Need another OpenStack service? Add it. Want to configure something exactly your way? Sure. Want to automate the control plane? Well, it is yours, so yes.
Charge by the hour.
YES! This can be tricky with full hardware as decommissioning the hardware means it automatically is “cleaned”. We are working on a balance – we don’t want people to accidently return an active server since all data will be removed completely as soon as that happens.
Keep the costs low and, absolutely, absolutely, beat the TCO of running on the mega clouds including your team costs if you manage your own private cloud with us.
YES! And our pricing calculator for our Ashburn/Washington DC data center is available here.
Next week, I will post an update on how we are doing. Many things are close and we do not anticipate not meeting the major goals, but it is always tough at the final stages to get everything to align.
If you really want to dig into what we are doing, see Why is on Demand OpenStack important for SMBs.
Also, if you believe in Open Source and your company is using a “Mega Public Cloud”, please consider our OpenMetal Cloud with a core Hyper-Converged Private Cloud. Built on Open Source, no lock in, better costs, and private from the start.