As some of you may know, I am part of the Storage Networking Industry Association (SNIA) which means that I have a tendency to gravitate towards open standards as a solution. This post will be a great example of that as it will cover one of those standards.
Pretty much everyone who has selected a cloud solution as their final goal agrees that getting there isn't a binary moment. You don't go from no cloud to cloud in a single step if you want meaningful results. That means that getting there is a journey. Every vendor you speak to will tell you that the transition is a journey and they will offer to help you along it. However, when acting on behalf of customers in these conversations the initial workshops always seemed to focus on the same thing, the fundamentals of the technology and what it can do. To me this is the same as planning a holiday and focusing entirely what the plane or train is capable of. Yes, these are the tools I use to reach my destination, but they are NOT the destination itself.
When looking to deploy cloud services it's never going to be a binary transformation. There are very few organisations that have existing IT services that will just be able to transition from that current state to a full cloud solution. This is true no matter what technology is deployed as the solution, it will always be an iterative progression where gains are steadily realised.
Hopefully you have already got your environment set up in as detail in part 1, if not it's worth referring back to it as it contains some useful details about the machines we will be configuring and the network requirements for them.
The place you should be at is to have 4 VM guests installed with base + OpenSSH server of Ubuntu 14.4 LTS. You should also have a single user, in my case I used "openstack" with the same certificate on all boxes to allow password-less ssh between the boxes. This just makes life easier.
In my last post I mentioned briefly some ways of automating the installation and configuration of OpenStack which, when you start to deploy at any kind of scale, is really the only way to do it. One of those that I mentioned was TripleO or OpenStack on OpenStack and this triggered a few questions. The most abundant of them was "what is TripleO?"
A lot of the work I do around OpenStack, when not for a customer or corporation and generally when I want to kick the tyres on a new feature, is performed locally on a set of virtual machines. I've been asked by a couple of people if I could go over the setup that I have and give a quick tutorial on how I got there. Although the simplest way to do this is actually to perform the manual steps from the excellent http://docs.openstack.org documentation, I will put some details here as to the configuration I have set up. The posts that follow (including this one) will very much focus on the manual installation.
This is just a quick note to those of you who were redirected here from lucr.com. Lucr has ceased all activity in the infrastructure and consultancy line and is in the process of being wound-up due to me moving on to a new role in the not to distant future. If you have any queries, questions or comments about that, please let me know via the contact button on this site and I'll be happy to answer whatever I can.
Well here we are. Now that it's going out of fashion I've started a blog. Why? Well it's primarily for me. It's a place where I can dump thoughts and ideas as well as research that I perform as I go about my business of getting organisations, tasks and workflows into the technological present. I've discovered that this can be challenging not only from a technical perspective, where new and fragile technology tempts us and lures us into believing "we can do it" but also from a people perspective.