One of the most frustrating challenges a data professional can encounter is to be blindsided by the news that the database system is about to run out of storage space. This unwelcome situation can have serious consequences for the business by slowing transactions or bringing operations to a halt.
Accurately predicting storage space is a difficult problem to solve because of the highly variable nature of storage usage history. For example, an e-commerce company might have somewhat predictable seasonal e-commerce spikes in database transactions as well as unanticipated surges—say, if Beyoncé tweeted about the new shoe line. If a storage planning tool doesn’t account for periodic workload spikes and doesn’t help with long-range capacity planning, data professionals are left in the dark—or worse, lulled into thinking they have enough storage capacity to handle the organization’s workload when they don’t.
The great news is that SentryOne can now address this nightmare problem for your customers with Storage Forecasting, a new capability in the SentryOne monitoring platform that applies machine learning algorithms to forecast daily usage for all storage volumes across servers. This approach is far more accurate than other storage planning tools available because the Storage Forecasting system continuously learns how your customers are using storage space and adjusts usage predictions accordingly. The more it’s used, the more accurate it becomes.
In addition to daily usage forecasts, Storage Forecasting provides a Windows disk forecast report that shows all volumes in the monitored environment that are predicted to run out of space within 180 days. This view takes much of the guesswork out of capacity planning and helps your customers stay in control of their data environments.
Storage Forecasting is just one of several innovations that will use our expertise in machine learning to make database performance management easier and more proactive for you and your customers. If you want to look under the hood of this new capability, check out the blog post “Storage Forecasting and Machine Learning with SentryOne,” by data scientist Fred Frost. In this post, he discusses how Storage Forecasting builds on Microsoft Machine Learning Services (ML Services) available in SQL Server 2016 and later. Another great resource you can pass along to your customers is our on-demand webinar, “How to Predict Your Future Storage Needs,” presented by Steve Wright, SentryOne director of data analytics.
Ready to help your customers solve one of the most challenging—and potentially debilitating—data performance problems? Contact us to learn how to unleash this new capability for your customers.