Hybrid Cloud: The new buzzword that many vendors want to use but rarely deliver on
So Public and Private Clouds are pretty easy to understand and both have their pros and cons. But in cloud terms, the Holy Grail is the ability to move application workloads from your on-site computer room to a secure data centre either locally or in another region altogether. This might be due to a Disaster Recovery event, expansion of the business to a new overseas territory or during a peak sales event or end of year data processing task.
Hybrid Cloud technology is used to facilitate the movement of the compute requirement between different platforms. Sounds simple, but what makes Hybrid Clouds difficult to define is no one customer’s compute and data transportation requirements are the same. This would be best explained with some examples.
Say for an instance you run a large online department store. For 11 months of the year you know what your traffic and transaction volumes and have purchased a cost effective managed Private Cloud to handle the average workload. However your problem is the December sales period. Here you experience volumes that are ten times greater than a normal month.
So to augment your Private Cloud, you setup additional servers in a Public Cloud (NTT ICT’s I hope!) and direct traffic to them during the sales event. Because they are not active for most of the year they only attract a small storage fee.
Your customers still get the same user experience as they normally would and you get sales that would have normally gone missing if the site had of crashed. Sounds good doesn’t it? To accomplish this, there has to be a lot of smarts built into your solution. The Load Balancer needs work out if the new servers are up or not, how much traffic to direct to them, and both environments has to be able to talk to each other as if they are on the one network.
The best performing IT department generally isn’t the one that keeps their services available most of the time, but it's the one that gets their services back up and running the quickest after a disaster or major event. Traditionally this has meant having a like-for-like replica of the production environment at another site. A hybrid cloud solution changes this approach somewhat.
Instead of having identical servers in another data centre, a hybrid cloud solution allows you have a remote virtual data centre where compute resources are only consumed during a DR event. Copies of production VMs can be kept in sync with vSphere vReplication or Veeam replication, allowing IT departments to offer their users sub 5 min Recovery Point Objectives and Recovery Time Objectives at a fraction of a cost of traditional DR strategies.
The maturing of Infrastructure clouds in the market now gives customers the ability to choose a solution that not only closely matches their IT needs, but also matches their financial requirements as well.