Micro Focus is now part of OpenText. Learn more >

You are here

You are here

How to choose: Open-source vs. commercial cloud management tools

public://pictures/davidl.jpg
David Linthicum Chief Cloud Strategy Officer, Deloitte Consulting
 

With so many open-source and commercial tools available for cloud management, how do you, as an IT operations management professional, decide which is the best fit for your team's needs? Here's a quick rundown of the general strengths and weaknesses of open-source versus commercial tool options, and when it makes sense to have one, the other, or a mix.

You need a plan for where free and open-source tools fit into your overall cloud management strategy, and where you should consider commercial options to complete the picture and meet your overall needs.

Here's how to get started.

Open-source is an option, not a religion

People like to get by with open source, and they also like the fact that they can have a greater influence on the maturation of open-source tools. Many companies' dedication to the open-source concept is almost religious. 

This is understandable. Enterprise IT has been held hostage for years by large vendors selling proprietary, and expensive software. Vendor software remains embedded within enterprise IT, but staffers have worked diligently to displace it with open source and, more recently, with cloud-based open-source technology. 

But IT is missing out on something when it chooses to go exclusively on the open-source route. Organizations that have a religious commitment to open source risk having the pendulum swing to far in that direction. Open source doesn't solve all problems, so it's important to keep traditional software options in the mix when considering your cloud management tool needs. 

Back to basics: The purpose of cloud management technology

While cloud management and monitoring platforms provide very different types of capabilities, a good depiction of common patterns appears in Figure 1 below. Keep in mind that humans sit at the highest level in this stack, since they provide core monitoring services. However, there is also automation at this level, meaning that you can set up rules and policies that can take automatic corrective action. 

One example is the ability to automatically provision more virtual storage when a database is about to run out of space. Another is the ability to provision more machine instances when performance starts to lag.

 

 

Figure 1: Typical cloud management systems let you look at the applications and data, platform, and infrastructure. They can monitor and manage most of the cloud-based resources across private and public cloud platforms using both manual and automated processes.

The requirements that you work from will emphasize certain functions specific to your needs. For instance, the ability to manage the network is more important than is managing compute and storage. Also, application management may not be as important if you leverage packaged applications that don't provide good management hooks. 

Finally, security may be a top priority, in which case a cloud management system that works and splay well with existing and newer cloud security subsystems will provide more value. 

The assumption here is that you get the requirements right for your needs. There are general patterns of cloud management tools, open source or commercial, that you should consider based on your requirements. However, you need to define your own needs before you can back the right technology into your requirements. In the end, it doesn't matter if the technology you choose is open or commercial. 

A typical short list of requirements should include:

  • Automation mechanisms, either policy-based, rule-based, or both
  • Links with service and resource governance
  • Links with security subsystems
  • Cost governance and monitoring
  • Types and granularity of data gathered, and how it's analyzed
  • Application-level management and monitoring
  • Infrastructure-level management and monitoring (storage, compute, network, etc.)
  • Connection to cloud and non-cloud components (agents)
  • Clustering and scaling
  • Orchestration and scheduling 
  • A "single pane of glass" console
  • Integration with existing non-cloud system management tools 
  • Integration with DevOps tools and automation 
  • Budget/cost requirements, and business case 

Yes, it's complicated. That’s why many enterprises choose not to go through a requirements and selection process. But if you don’t go through it, you'll probably end up with a substandard, ineffective solution. It may work, but the shortcomings will likely cost you millions of dollars over the years, cause problems for end users and create a lack of confidence—and those issues are not easy to fix. 

Commercial or open source?

It’s helpful to look at the realities of both types of software, open source and commercial, when it comes to cloud management tools. Both approaches have pros and cons.

Open source: Free, with caveats

The upside is that it's free, and can be had for the price of a download from GitHub or another open-source download site. Moreover, you don't need to go through hours and hours of meeting with vendors, as well as the drudgery of enterprise software negotiations. You need it, you download it, you install it, and you're ready to test or deploy. 

Another pro is that you have a say in how the product matures over time. These are community-driven projects, and you can contribute to the code tree and make recommendations as to the software's direction. The downside is that you have to pay people to do that on behalf of your business, and there's an opportunity cost to that. Is there a return on investment for your business to advance the software or add a key new feature?

On the downside, open source, generally speaking, is not as mature as commercial offerings. In general, you'll experience more problems with open-source cloud management software, and you’ll have to deal with those problems yourself, working with the community—or by engaging a company that provides consulting services around its distribution of an open source cloud management system.

Second, bug fixes and new features may take months—or years—to show up in the core code tree. This is due largely to the fact that volunteers make the fixes and/or enhancements. This lack of a profit motive can significantly delay releases compared to the commercial world, where money talks and motivates the cloud management provider to make the fixes and improvements you've requested as soon as possible. 

Going commercial: Stability and maturity

The commercial cloud management tools tend to be more mature and in more stable states. Considering the profit drivers, it's not advantageous for vendors to release code that will cause their customers problems, in terms of stability. 

Bad commercial releases of cloud management tools can damage a vendor's brand reputation, lose them deals or even result in lawsuits. They're motivated to deliver quality software because doing otherwise is bad for business.

Open-source cloud management tools, in contrast, make users agree to terms that essentially hold open-source providers harmless. It's one of those "you get what you pay for" types of situations. 

Another pro with commercial tools is the level of support. While you'll find good community support in the open-source world, expertise is harder to find there, and the support techs are, in essence, supporting a third-party technology, not something they have built in its entirety. 

In the commercial world, vendors support their own technology, and they have more incentive to see it be successful. Moreover, they have direct access to project managers, developers, and other experts who can fix problems directly or provide answers to tough technical questions. 

But you pay for that. Commercial cloud management systems are not free. They can cost you dearly when you add up license fees, support agreements, consulting, and so on. In the end, however, both have costs. It's just a matter of realizing that you'll either pay up front now or down the road. 

And the winner is …

First you need an accurate list of your own requirements, and then it's time to consider the true costs of both types of cloud management solutions: open source and commercial. 

What does "true cost" mean? It's the actual cost of leveraging open source vs. commercial. On the surface, open source usually appears to be the cheaper solution because it's free for the taking.

However, you need to add in the expense of installation, support, fixes, upgrades, and other hidden costs. Your internal labor force will incur most of those costs as they work to learn the software, get it up and running, and resolve issues. Many enterprises fail to consider these costs in the planning and procurement stages, but in some instances, the hidden costs, in terms of tying up labor that could be working on other things, could well exceed the cost of going with commercial cloud management alternative. 

So think about your cloud management system requirements. Open-source cloud management tools lack some features and functions that will be on your must-have list. 

Unfortunately, to make an open-source square peg fit into a round hole, many organizations minimize or overlook these requirements. But for them, the cost of not meeting those requirements could actually exceed the total cost of ownership of a commercially sold cloud management system. And the latter might not only better meet the requirements, but deployment and support expectations as well. 

It's possible that a commercial cloud management system could actually end up costing half as much as an open-source system when you consider all of the hidden costs outlined above. That's why in my practice I've seen many enterprises pivot from open source to commercial when the metrics became better known. While that's better than continuing down the wrong path, having to make such a pivot is a waste of time and money, and adds significant risk. 

That said, the best choice is not always commercial; there are many good reasons to go with open-source cloud management system. The point is to start by focusing on your own specific requirements, not what people perceive to be the right direction. 

Keep learning

Read more articles about: Enterprise ITCloud