Many suppliers have tried to ride the wave of cloud computing by simply shoehorning existing product portfolios into a cloud-ready pigeonhole.
Quocirca has seen numerous examples of this, whereby companies take existing applications, host them in a standard (or semi-virtualised) way and label it “cloud” just because the apps are accessible over the internet.
According to the National Institute of Standards and Technology (Nist), a true cloud service has to satisfy a basic set of features. It needs to be have elasticity, broad network accessibility, self-service capabilities, resource pooling and provide a measured service.
One thing Nist covers is what constitutes a cloud workload. Moving an cloud application off-premise may appear to be cloud, but it fails to provide the optimum value that this IT delivery model promises.
Startups versus enterprise
If you look at modern, green field organisations that have seemingly cropped up overnight and become household names, they have generally taken a very different approach to IT than traditional enterprises.
Taxi-booking app Uber is a good example. Instead of turning to the likes of Oracle and SAP for its enterprise resource planning or customer relationship management systems, it has mashed together cloud services from Twilio and Braintree to create a system better suited to its business needs.
This is a strategy many other companies have adopted, deciding to eschew building their own IT platforms to bring together existing functions from other clouds for business agility reasons.
However, it’s an approach that requires a high degree of monitoring as the resultant platform evolves. If not, the systems might not work well together and customers will desert the company.
Compare the cloud monitoring market
So, what is available out there to manage such a hybrid cloud system?
BMC, coming out from a very quiet period of self-imposed purdah since taking itself private in 2013, has updated its TrueSight Operations Management, which offers a pretty comprehensive means of monitoring and managing these types of workloads.
CA Technologies, meanwhile, has a comprehensive portfolio of offerings, including CA Business Service Insight, CA Application Performance Management, CA App Synthetic Monitor and others. Its strength, though, is also its biggest weakness. Having such a broad portfolio can make it difficult for prospective customers to understand just what they need from CA.
IBM has its Tivoli portfolio and can draw on its 2013 acquisition of SoftLayer, which has a host of built-in monitoring and management capabilities.
Cisco Intelligent Automation for Cloud is an option for those wanting to utilise the networking giant’s cloud aggregation model. This enables disparate (but OpenStack-based) clouds to interoperate and manage workloads across them.
As a result of its 2013 acquisition of Enstratius, Dell is another player now capable of managing workloads in a hybrid environment.
Having decided to position itself as a cloud service broker, the company can combine its Dell Cloud Manager with its acquisitions of Boomi in 2010 and StatSoft in 2014 to provide an intelligent cloud management system that fully understands the various needs of different workloads.
Storage giant EMC acquired cloud platform provider Virtustream, which comes with an advanced workload management capability in xStream that should play well with its parent’s cloud aspirations.
Hewlett-Packard, having dipped in and out of various cloud approaches, is strengthening its Helion portfolio to offer hybrid cloud workload management capabilities too.
Then there are the smaller players – the likes of Stratus Technologies, Accelerite, StackIQ, Flexiant and Novell, all fighting to gain a significant portion of what is going to be a large market.
Some of these have differentiated offerings. For example, StackIQ has an approach to managing provisioning using packages and pallets to provide “warehouse-grade” workload management.
Making the right choice
All this choice can be confusing for companies, and the consequences to the business if the wrong choice is made can be dire. There may not be sufficient functionality available, for example, or the supplier could go out of business, leaving your company high and dry.
It is, therefore, important to understand exactly what your company is trying to do. A base-level tool from an incumbent supplier (maybe an existing systems management supplier or the majority hardware provider of your company’s private infrastructure) may give enough functionality for dealing with day-to-day semi-automated workload management. For some, this may be enough, as third-party clouds mature and change.
Others may require dynamic workload management, which brings two choices.
One is to take the “safe” route of a base-level tool and then layer extra best-of-breed capabilities over the top of it. This could, however, lead to problems, as the base product or best-of-breed tools evolve and interoperability problems emerge.
The other option is to use a smaller supplier that provides more of what is believed to be required and integrates well with the company’s existing management platforms.
Regard this as a tactical investment and ensure the value you are going to get from the use of the system outweighs the costs of licences and implementation (if it is not already a software-as-a-service-based product).
Keep an eye on the markets and if the provider you opt for goes out of business or is acquired by a supplier you don’t like, implement plan B and replace what you have with a different system.
Effective workload management across a hybrid platform is going to be big.
Those that end up with manual or semi-automated systems will struggle in the new economy. Those that get it right will find the IT platform becomes just what it should be – a business facilitator, rather than a constraint on the business’s capabilities.
