Eclipse Digital - Fotolia


Five hybrid IT infrastructure breaking points

A hybrid deployment of infrastructure as a service, virtualized dedicated servers, software as a service and other resources is good for the business -- and tough on IT.

Managing a mix of private cloud, virtualization and public cloud platforms is hard for IT from both a monetary perspective and a management perspective. These are five areas to invest in to get the best bang per buck in hybrid IT infrastructure.

Hybrid IT -- where an organization mixes owned IT infrastructure with public cloud platforms in the form of infrastructure, platform and software as a service -- offers much in the way of flexibility, cost savings and relatively future-proofed capabilities.

However, it also brings many problems, such as end-to-end performance, root cause analysis and overall systems maintenance.

When investing in a strategy for operating hybrid IT infrastructure, watch out for these five areas.

1. Overall visibility of the platform

If you can't see what's happening, you can't do anything about it. Therefore, you need that single pane of glass that gives you visibility into the dedicated IT infrastructure and as much of the public platform as the service provider will allow you to see.

Vendors such as CA Technologies, BMC Software and IBM continuously extend their standard systems management platforms to enable this sort of approach.

Some vendors, such as Nlyte, Future Facilities, Raritan, and Emerson Network Power, have come from the data center infrastructure management (DCIM) side to reach into public cloud monitoring. Most DCIM tools are pretty full function IT tools now. The trouble for DCIM vendors is often getting their tools into the hands of IT rather than facilities. Gaining access to enough information around a public cloud environment, DCIM tools hold enough metadata around the overall IT platform to provide not only a good view of what assets are available and how they are running, but also to be able to carry out What-If calculations around moving workloads from on premises to the cloud or other choices.

Newer platform management systems for hybrid IT infrastructure, such as those from Flexiant, EMC's Virtustream and Egenera, have started from a cloud-only management focus and moved into a more inclusive overall capability.

2. Performance monitoring

When the entire IT deployment is in your own data center, it is relatively easy to identify where performance issues lie. Is it in the user's access device; the network or the data center?

With a hybrid IT infrastructure, it becomes far more complex. Is it in your data center, your network, the service provider's data center, their network, or the global network that connects everything together?

Performance monitoring for hybrid cloud and on-premises IT is still relatively immature at the moment, but vendors such as CA, which acquired performance monitoring and management vendor Wily Technology in 2006; and Dynatrace, which acquired Gomez in 2009, with their application performance management (APM) suites have systems that can help across a hybrid platform. SolarWinds, a newer company, has a portfolio of performance monitoring tools, and companies such as SignalFx are providing APM as a service capabilities.

3. Capability to carry out root cause analysis

Things will go wrong with a hybrid IT environment -- and identifying exactly where the problem is won't be easy. With multiple independent entities now involved, generally across multiple contracts, finger pointing at each other becomes almost unavoidable.

A key component of any hybrid IT infrastructure strategy is the irrefutable capability to identify what is causing the problem, and therefore who has responsibility for fixing it.

Many of the vendors referred to above offer the capability to locate problems, and some of them also offer the capability to automatically remediate issues. However, the service providers must open up their platforms to allow IT organizations to have a good enough degree of visibility as to what is happening, otherwise it just becomes a case of "we have narrowed the problem down to something to do with you -- when can you fix it, please?"

Bear in mind that simplistic root cause identification tools may be more trouble than they are worth. It's no use to fix a problem only to find that your fix moves the problem to another part of the infrastructure. The chosen tools must be able to not only look at the basic problem, but also determine the follow-on effects of fixing that problem, on all the workloads on the platform.

4. Cloud costs

Far too many organizations move to public cloud purely as a means of saving money. Out of these, the majority then find that public cloud is far more expensive than they expected, as hidden costs suddenly appear and the cuts made to overall systems quality to meet price points become apparent.

The key is to go to cloud because it enables the organization to do things that it couldn't do before, such as respond to spikes in workloads, try new ideas and flex the overall platform to reflect market conditions. This approach to hybrid IT infrastructure may work out far cheaper than was expected.

It's critical for IT teams to identify, monitor and manage costs in a hybrid environment. Some organizations will use chargeback, while others prefer IT showback, where departments and even individual users see how much they are costing the business and receive best practice guidelines for how they can optimize these costs.

The main areas where vendors are doing a good job in cost minimization is software asset management. The likes of Snow Software, Flexera and 1E provide systems that have evolved from straight-forward license management systems to tools that monitor and control the usage of services from public clouds.

5. Security holes

A hybrid IT infrastructure approach inherently increases the security surface for an organization. Whereas a completely owned platform within one or two data centers can be fenced off securely from the outside world to a reasonable degree, a fully hybrid approach brings in the use of public networks for core, business-critical applications and data, with a higher degree of dependency on the capabilities of external service providers to manage security on the provided resources.

Hardware, application and database security are of secondary concern; the information really matters, and therefore needs the most security investment. Bring in tools such as data loss and leak prevention, from vendors such as Symantec, McAfee, Check Point and Digital Guardian, alongside digital rights management systems from vendors such as Adobe, OpenText and Microsoft to ensure that information is secured along the hybrid IT infrastructure and the value chain of suppliers and customers.

Indeed, when looking at how information is shared across this value chain, greater control is needed. Vendors such as Tresorit, FileOpen, Intralinks and others make sharing information easier while maintaining high levels of security.

The move to a hybrid IT infrastructure is pretty much unstoppable. The rise in usage of public cloud from commercial service providers is growing. Plan for the best level of control of this new platform as close to the beginning of its adoption as possible

Next Steps

Building an effective hybrid cloud management strategy

Management tools for hybrid cloud environments

Is the single pane of glass a reality for hybrid cloud?

Dig Deeper on Managing Cloud-Native Applications