Determining Databases: Open Source or Proprietary?
BackBy Brian Summerfield, Senior Editor —
The average lifecycle for a database, from selection to obsolescence, is somewhere between six and 10 years. When the time comes to get a new database, a few common factors must be considered, including cost, scale, functionality and performance. These days, most of these companies also will face a decision in terms of what kind of entity provides their database: proprietary or open source.
Each has its advantages and drawbacks — in fact, a benefit for one organization might be a shortcoming for another. For instance, proprietary databases usually include a bundle of support services and training provided by the vendor. Open-source databases are generally perceived as being more flexible in terms of both customization and service providers, which might sound great to certain kinds of companies and challenging to others. “(Open source) is more flexible, but the counter to that flexibility is that it sometimes requires more work on the user’s part,” said Bruce Momjian, one of the core developers for the PostgreSQL open-source database. “It becomes a more complex decision-making process for the vendor.
“For example, with something as simple as getting tech support for your database, all of a sudden you have multiple vendors that you can choose for that support, as well as multiple vendors for training,” he added. “You have to be a little more educated when you choose a vendor to do that ‘last mile’ support for their database.”
Other apparent problems frequently include the maturity of open-source databases (and other technologies) as well as their compatibility with existing systems. Although these issues can make the open-source route seem like a gamble, more and more companies are deciding that it’s worth taking a chance on. “Maybe five years ago, some very large organizations weren’t willing to take the risk of bringing open source into their operations and using it in a very mission-critical way,” Momjian said. “But what you have now is that the risk for the CIO is really rather minimal.”
Still, many companies are dipping their toes in the water rather than just diving in. For instance, they might not opt for an open-source database like PostgreSQL or MySQL until they’ve successfully implemented Linux or Open Office first. “They bring one thing in and that works well, then they bring another thing in and that works well, and the database ends up being one of those things on the checklist,” Momjian explained.