Primordial Software Group has many years of experience building and operating general-purpose virtualized computing environments (aka. virtual datacentres). Often described as Infrastructure as a Service ("IaaS"), these facilities allow for highly flexible and rapidly deployable and adaptable computing, storage, and networking resources.
Running an IaaS-like system in-house is often called a "private cloud", because you can have all the benefits of cloud-based deployments, but keep it in-house for total control and adaptablity above and beyond what the big players will let you do.
Some of the benefits of operating a virtualized datacentre include:
- Fully redundant infrastructure
- Rapid/Flexible deployment
- Integrated monitoring and notification
- Independence and fine-grained control
The physical servers, storage, and networking that make up the virtualization cluster is 100% redundant. With multiple servers, distributed storage, multiple network links to every system, and redundant network switches, there is no single-point-of-failure within the system. Any server, storage node, network cable, or network switch can undergo maintenance with no impact on the operation of the virtualized systems.
In the event of an unplanned outage of a server, any virtual machines that were running there will be immediately restarted on another node. Readily-distributable services, like load-balancers, web servers, and databases, would already be running multiple instances across the cluster, thus service availability would be maintained at all times.
With ready-made virtual server images, new systems can be brought online within moments, with minimal setup. No lead-time for new hardware purchasing and acquisition, no physical installation, power budgeting, or asset management processes required.
Specify the VM's CPU, RAM, and storage capacities exactly as required. There is no restriction to generic "small", "medium", "large" node definitions like with some providers. And many aspects of the system specification can be changed on-the-fly with no downtime required. Disks can be grown, memory added, even CPU cores added to a live system.
Integrated system monitoring can alert you to problems with your virtual machines, including abnormal CPU or network usage, automatic restart or migration events, etc.
Give your system operators and even end-users control of the parts of the system that they need, without having to adopt an open-door datacentre. Every VM or pool of VMs, virtual network, storage pool, disk, and cluster node is restricted by a set of fine-grained permissions that may be granted to any user or group of users. You could let each of your developers have total control of a VM for their dev/test needs while still restricting which virtual networks they can connect it to, for example. Or you could allow junior sysadmins to only control less-critical virtual servers, while allowing only the senior sysadmins to manage virtual machines that service the finance and HR departments.