In my previous blog, I discuss the emergency preparedness techniques all companies must know in order to survive an unexpected disaster. I continue the discussion here.
Regular execution and testing of emergency plans
If you put together an emergency plan and then don’t test it regularly to make sure that it works and still meets your current risk requirements, then when a real incident happens, all that planning was for nothing.
Your private cloud should be engineered with redundancy built-in, and it should allow you to set the level of protection and recovery that makes sense for you. At a base level, your Dedicated Private Cloud (DPC) clusters should replicate all data in any node to at least two other nodes, providing redundancy of data. This can be combined as desired by DPC users with protection of VMs, providing protection of critical workloads and data even in the event of an underlying node’s hardware.
For customers who want to enable High Availability across multiple clusters, they should seek options to implement private clouds in multiple data centres, providing the ability to create geographic redundancy in active-active or active-passive configurations, across either a single metro region or nationally.
Beyond this, DPC should be connected to your provider’s backup services and pre-enabled with backup software, providing your company with on-demand access to an additional layer of data backup outside the cluster.
Your service provider should be able to offer you many other options for other layers of redundancy or resiliency as it relates to their cloud platform, and they should be able to work with you to design a solution that meets your specific requirements in each case.
Lastly, the cloud related services that your company is receiving should be built on a redundant network fabric and should include redundant 10Gbps connectivity out of the fabric to prevent connectivity failure. In addition, the service provider’s private cloud should connect to a separate out of band (OOB) management network, which provides an additional method to connect to resources as required.
I can’t think of a modern business out there that can exist long without ongoing IT operations, if for no other purpose than to be able to take and make payments and keep the books balanced.
IT is not optional in business operations. Recent research shows the cost of data centre outage for the average Canadian business is around $6000 per minute of outage. And even after an outage, the impact to the business’ long-term viability and credibility can be permanent. The data is pretty clear: IT outages at best cost a business dearly.
But with the technologies available today and the right partners, it’s a risk that can be managed better than ever before.
The future or emergency preparedness
Of course it’s always better to avoid risk than to mitigate it. This comes down to reporting and telemetry again. With the right performance data and the right intelligence, at both the network and the infrastructure level, you can apply predictive analytics to try and avoid emergencies before they happen.
There’s a lot of growth and innovation in this space. Two of the areas that I think show a lot of immediate potential are in automated threat management and predictive analytics. Especially when used in conjunction, these new approaches to risk prevention and minimization are providing exciting new ways to minimize IT risk in a constantly changing threat landscape.
The redundancy and ability to scale IT and communications infrastructure as demanded are primary factors in reducing impact and minimizing the recovery time from the event.