The term hybrid cloud has been around since about 2008. It’s one of those terms that never really got any traction in the over hyped world of cloud computing when everyone was focused on outsourcing services to the cloud. The problem was that hype over took reality. Most organizations whether they be at the Enterprise level or at the SMB level consume data from both on premise and cloud platforms. The mix varies from company to company. I know of no company that runs a completely cloud based operation. They may be out there but I have never come across one. The reality is it’s a mix of on premise and cloud – a hybrid model.
If you take a simple example from the logistics space. A logistics company may use a cloud based TMS package for fleet and load management, and an in house legacy WMS with a on premise accounting package. The legacy WMS and the accounting software were probably made to talk to each other years ago through custom programming. The issue now is how to connect the cloud based TMS to the on premise solutions ? Well there are two ways.
The first would be to outsource all the integration work to a hosted integration provider. If the on premise solutions have web service based APIs this is a piece of cake. If the data access point is through database staging tables it then becomes a little more complicated. Its not impossible but it does require solid infrastructure and a good Internet pipe. If the in house platforms have limited import export capabilities and simply require a file drop and pick up, it’s also doable. The more complex the data flow is, and the more complex the data touch points are, this is where the hybrid cloud comes into play.
So what exactly is the hybrid cloud? A hybrid cloud is a scenario where part of the integration and translation workload is placed in the cloud and part of it is in house. Workloads are chosen selectively based on a properly formulated data integration strategy that must include a complete assessment of the data touch points.
So why is hybrid so important. The reality as I outlined in my very simple example above, is that companies will for some time to come consume both on premise and cloud based data. Corporate IT strategies and ultimate spend will drive cloud adoption but the bigger driver will be the ability to integrate. Because there is no standardization in APIs, data integration is a critical issue. In an earlier blog post Randy Smith discussed how EDI hadn’t really changed much since the early 80’s, I would argue that the current API landscape is reminiscent of the pre ANSI X12 era in the world of standards based EDI. Everyone did his or her own thing. Just take a look at Programmable Web and see the huge number of APIs and the different formats, approaches and communications methods.
So how does a business cope with this reality? I have 3 points I always tell prospective customers.
- Have a solid data usage strategy. This means understanding how data flows into and out of your organization. Do an organizational 360 of data flows and usage. Map them out, and then map out the touch points
- Find the commonalities. A purchase order is a purchase order regardless of the data format it arrives in. Too many companies still look at data in siloes. Think holistically
- Finally determine which integrations are best handled with a solid middle ware platform and which ones are best out sourced to the cloud either because of their complexity or because of the uniqueness. A number of factors go into this decision.
Once you have armed yourself with answers to these 3 steps ensure that whatever you use in-house or in the cloud can be easily integrated. What does that mean ? Does the application have an API, can you get data in and out easily without custom development. If you can’t and the application is closed scrap it, it will hinder your business and cost you a lot in lack of flexibility.
Finally, ensure that what you choose as your in-house middleware has the capability to handle your on premise and cloud-based data integration requirements. In my next article I will take a look at how we at VL have used the ECS and Delta cloud connectors to create a hybrid solution for our EDI Desktop software. On premise pick and pack software connected to the cloud for data translation.
In the mean time take a read of our ebooks if you need help in formulating your data strategy.