My Azure Virtual Machine Won’t Start

To benefit from elasticity on Microsoft Azure, we regularly leverage strategies that utilize Azure Automation to shut down and de-allocate Virtual Machines when they are not needed.

This strategy is excellent and usually works on pretty well. Due to Azure’s highly dynamic nature, the allocation of Large Virtual Machines can sometimes pose a challenge in the form of “The server encountered an internal error. Please retry the request.” or “Failed to produce an allocation for the service.“.

Can’t Delete an Azure Resource Group

Recently, I hit a wall. I wasn’t able to delete my Azure Resource Groups and got quite frustrated… I’m writing this post, with hopes that you may save some time, frustration and energy with your ongoing challenge.

Provisioning a New Data Disk

Working with Linux on Azure has brought me back to the basics. This post is all about adding a new Data Disk to an existing Azure CentOS Virtual Machine. If you don’t have a Virtual Machine handy, I wrote a short walkthrough to help you get going.

Disk UUIDs and Linux on Azure

I recently ran a few tests to figure out how UUIDs behave on Azure when we capture VM Images, copy VHDs, move Virtual Machines and re-provision them in different data centers.

Moving to a New Azure Datacenter

From time to time, I face interesting challenges. Azure is an exciting platform, because it’s pushing me to learn about things that I wouldn’t of dreamed of a few years back.

This post is all about moving a CentOS Virtual Machine that has a RAID 0 to a new Microsoft Azure Datacenter.

Share this:

While I was playing around with the Azure Resource Manager Copy Operation, I started thinking about what I could do with it. The first wild idea that popped into my head was, to use it to deploy multi-geo environments from a single ARM Template.

Alright, some of you might think that it’s not such great idea, and I can appreciate that. But I’m just too curious, so let’s give this a chance.

Creating a CentOS VM Using ARM

Docker and Open Source projects are getting lots of attention, so I decided that it was time for me to build a Linux Virtual Machine on Microsoft Azure. This post is all about creating an Azure Resource Manager Template for a CentOS Virtual Machine with two stripped Data Disks. This template should be used as a starting point and may require some tweaking to meet your needs. Feel free to share your thoughts by using the comment section.

A Monster VM Azure Resource Manager Template

In April I wrote a post about building a monster Virtual Machine using PowerShell on Microsoft Azure. Since then, Microsoft has released version 2 of the Azure Resource Manager (ARM). This version allows us to define a Virtual Machine, its data disks and its Desired State Configuration (DSC) VM Extensions as a template. Seeing this as a great opportunity, I decided to convert my first PowerShell script to an ARM template that would create a Virtual Machine and striped data disk.

The Target Virtual Machine Configuration

16 Cores
112 GB of RAM
800 GB of local SSD for temp disk
32 TB for the data disk
50,000 IOPS for the data disk
512 MB per second for the data disk

DSC Module to Prepare a Stripe Volume

This post marks my first adventure with PowerShell Desired State Configuration. Over the past few weeks, there have been a few announcements around Microsoft Azure. Once of these, is an extended version of the Azure Resource Manager which brings us the ability to run PowerShell DSC on Virtual Machines. This update changes everything and I will delve into it in an upcoming post.

Using the xDSCResourceDesigner enabled by installing Windows Management Framework 5.0 Preview April 2015, I was able to use the following commands to scaffold a Customer PowerShell Desired State (DSC) Module that allows me to stripe data disks on an Azure Virtual Machine.

Building a Monster VM

Microsoft Azure is all about the opportunity to push back on known boundaries. In the last couple of months, I dealt with some scenarios where on-premises Data Centers ran out of capacity. Consequently, we were not able to push our tests as far as we would have liked. Taking the work loads to Microsoft Azure gave us the opportunity to stretch workloads to their limits.

In one of these scenarios, the workload was limited to 16 aging physical machines, and the time required to process the data was not acceptable. We needed to find ways to reduce the compute time, and had means to accomplish this on-premises. Pushing this workload to its limits, we deployed it on Microsoft Azure and provisioned it with well over 256 cores. We ended up processing workloads so fast that we now had the opportunity to run them multiple times a day. This newly discovered agility gave us the ability to refine the workload processes without disturbing ongoing business activities.

The first scenario was all about the lack of compute resources necessary to push a workload to its full potential. In a second scenario, we needed an impressive amount of resources within a single Virtual Machine. This blog post is all about how we created this Monster VM.

The Requirement

16 Cores
112 GB of RAM
800 GB of local SSD for temp disk
32 TB for the data disk
50,000 IOPS for the data disk
512 MB per second for the data disk

Let’s think about that spec for a second. That’s monstrous! And we’re going to build it!