Creating Cinder Back-End with Microstack

Hi All-

Not sure if this is the correct place to be positing this, but here it goes…

I am new to both OpenStack and MicroStack. I have my first MicroStack cluster initialized and running but was a little disappointed that out of the box, using the --auto option during initializing, no Cinder backend is created.

I have read elsewhere that it is possible to add the backend as a loop device using the appropriate flags during initialization, and while that is better than nothing, I would prefer to be able to add the raw block devices to the Cinder backend the way I presume it is done in the full-fledged version of OpenStack.

Correct me if I am wrong but to do this, I will have to create a RAID group either via software (mdadm, ZFS) or using a RAID mode available in HBA. This also presumes that the loop device and boot/root partitions of my MicroStack install will share the same physical media which I was hoping to share.

Currently I have my MicroStack install on a SATADOM (where I would like to keep it), and HDD/SSD raw disks available for use as Cinder block devices.

What is the best way to proceed? Should I forgo MicroStack altogether and adopt the full version of OpenStack and if so, where do I begin. There does not seem to be a “free” version of OpenStack any longer unless I want to compile my own code, which I do not want to do.

Thanks in advance.


Hi @minorsatellite,

Thank you for the question! Microstack is intended to be a simple, efficient way to deploy a generic OpenStack that can run in as many places as possible. It’s flexible enough to accommodate specific needs, but can’t really do so automagically – there are just too many special cases to cover. (Contributions are always welcome, though!)

The OpenStack team has a deployment guide for Charmed Openstack, which is a more flexible product that does not require compiling from source. You may still need domain specific knowledge when setting up hardware and networking in the undercloud, however.

1 Like


Thanks for the link to the Deployment Guide, thats exactly what I was looking for.

I am humble enough to admit that I do not have domain specific knowledge relevant to OpenStack but I really do want to learn its internals so even though it will be more laborious and time consuming to deploy, I am going to just dive right in. After all, there is not better way to gain such knowledge.