The discussion around open source and saving money has been going on for as long as open source has existed. While there are definite benefits that open source can provide in terms of controlling your data and fully understanding the code that is in place, cost saving are often seen as the biggest reason to move from proprietary software. However, how can those cost savings be achieved in practical terms, and why are they still possible so many years after open source was first developed?
One of the greatest challenges is understanding and quantifying the impact of software licensing for proprietary software, and how this can lead to problems over time. The issue is not whether suppliers should be paid for their support services, or be able to license their software in the way that suits them. Instead, problems occur through lack of clarity around historical support contracts. This is where open source solutions can provide immediate savings.
Where do the problems start?
Let us look at a hypothetical example, showing how a company might find itself affected by this problem. The IT team at Company X has a project in place that needs a particular product or service, and so they build a case to buy that product. In the past, this would be a capital expense plus money for support over time. For companies with large IT deployments, an enterprise license that acts like an ‘all you can eat buffet’ can be a great way to save on costs. However, this can lead to neglect when managing the use of software, such as databases, over time.
The thinking here is that everything is already bought and paid for, so why not make full use of that service? Any new additions are effectively ‘free’ to IT, as their license covers everything. This can encourage teams to innovate and move faster, as they don’t have to go through financial arguments and cost justifications for each and every project.
To make this easier, let’s provide some numbers. For a capital spend of £500,000 with support at 22 per cent, the cost of support per year is £110,000. So far, so good on both sides - Company X has the technology it needs, and the vendor is fairly compensated for its product. However, as time goes by, technology needs change and evolve.
A new contract will be needed, or additional products will be added. It’s here that problems start to arise. Typically, all the previous software support will be bundled together into an overall item to be used within the new negotiations. Reducing that set of licenses, products, and services to one line item simplifies the negotiation process, but stores up trouble for the future.
Again, let’s look at some numbers. The new contract might be £1million, with support coming in at £220,000. However, the previous support line will also have to be added, bringing support up to £330,000 per year. If this happens again, the new line item for past support would be £330,000, on top of whatever other support is needed.
Over time, this can snowball and increase the overall costs around support to high levels. This is particularly true when you have software or applications that are critical and reliable, such as databases. These tend to be in place for many years as they support services that create revenue or run operations, and they have to be supported.
Breaking down the problem
The typical response here would be to think, “well, we can just take off what we don’t use, right?” However, there are two problems here. The first is that it can be difficult for technology teams to track the applications, databases, or services that they use over time. They often don’t have an accurate list of everything that is still running and required by the business.
The second challenge is that all those past software licenses, agreements, and negotiations are now all bundled into one line item. So, instead of being able to pick and choose - and more importantly, remove costs that are no longer needed - all these licenses are being paid for regardless. In effect, each license cost will be compounded over time. If the team wants to cut those costs, everything has to be stopped.
This has a massive effect on IT teams, which have to keep operations running. The prospect of stopping all of their operations due to software licensing issues is not realistic for the business, so many companies carry on swallowing the extra costs over time, even though they know that they are no longer using the services they are paying for. Fixing the issues becomes too difficult, so it can linger for years.
Eventually this does lead to more financial pain than can be borne. The reality of vendor lock-in bites. Companies are not solely locked-in to a vendor’s technology, they are often also locked into a massive and potentially oppressive annual support obligation.
On the surface it seems that the only way to escape this spiral is to completely discontinue the use of all products from the vendor. This can also be an expensive and time-consuming process. It is unsurprising that many organisations in similar situations are looking at possible options. But, what does vendor lock-in extraction look like? And how easily can it be attained?
The value of open source
It’s here where open source approaches can have the most impact. At the very least this can help you improve your leverage with existing vendors. At best, it can provide you with concrete plans for managing your migration, whether you want to move to a new database, into the cloud, or to another platform of your choice.
Getting independent advice in advance of any new support contract being signed can help you guard against future rolling-up of service and support costs and avoid paying for services that you no longer use. Equally, this approach can provide peace of mind as and when any major migration is planned.
Rather than being mired in unclear agreements, open source provides clarity that IT teams require around what is licensed, and for what purpose. If and when IT teams decide to remove implementations, the costs are clear, and the budget impact can be felt immediately. Using open source in this way can help make lock-in easier to avoid. For those already dealing with the challenges around costs and support, breaking out lists of existing services that should remain in place and be supported again alongside what has been decommissioned or replaced will be essential. This can provide an opportunity to keep what works and and identify real opportunities for leveraging open source replacements for the things that don't work.
This will require some planning and preparation so that any migration is effective and there are no missed licenses. It is worth getting advice on migration approaches in advance.
Avoiding the compound effect of support lock-in
It always pays to carefully read the contract and leave enough time to understand the impact that your decisions may have over time. Avoiding excessive support costs is possible through better planning, and evaluating your options. By keeping tighter control over your support and services costs, you can ensure your budget goes further.
Brian Walters, Director of Solutions Engineering, Percona