Best Practices For IoT Workloads With Oracle Database

Maria Colgan
Jul 26, 2017 · 3 min read

Over the last few years there has been a rapid surge in the adoption of smart devices. Everything from phones and tablets, to smart meters and fitness devices, can connect to the Internet and share data. You only have to follow @MarkRittman and his experiences with getting his kettle to boil remotely to see just how many devices within your own home can connect to the internet.

With all of these smart devices, comes a huge increase in the frequency and volume of data being ingested into and processed by databases. This scenario is commonly referred to as the Internet of Things or IoT.

Some people assume that a NoSQL database is required for an IoT workload because the ingest rate required exceeds the capabilities of a traditional relational database. This is simply not true.

However, if we look at some more industrial examples, you quickly realize that these types of workloads have actually been around a really long time.

Telecoms process tens of millions of Call Detail Records (CDRs) per second or connected manufacturing equipment, where every machine in a production line sends information about which component is currently being worked on. And if we look at how the industry leading company in each of these fields manages their workloads, we find they have all been using the Oracle Database.

With that in mind, I began working with Cecilia Gervasio Grant, Weifeng Shi and my boss Juan Loaiza to define a set of best practices or management techniques to facilitate IoT workloads. We drew on our years of experiences with leading telecoms, financial institutions, retailers etc. to make realistic recommendations and to share real-world code examples, as well as performance numbers.

Throughout the paper we use a simple analogy of someone going grocery shopping to help explain the reasoning behind each of the tuning techniques recommended. For example, if you know you are going shopping for a lot of stuff, you get a big cart right away rather than selecting a small hand basket. The same is true in the database. You want to allocate a BIGFILE tablespace with large initial extents, so that you can ingest and store large volumes of data efficiently.

You would also never select one item at a time and pay for it before you select the next item on your list, which is akin to a single row insert followed by a commit within the Oracle Database. Instead, you would walk through the store, collecting all of the items on your list and then go to the check out once, which is why we recommend using array insert followed by a commit within the database.

We hope you find the recommendations outlined in the paper useful and would love to hear if you have used these or other techniques in your IoT or high-volume ingest workloads.

Oracle Developers

Aggregation of articles from Oracle & partners engineers, Groundbreaker ambassadors & the developer community on all things Oracle Cloud and its technologies. The views expressed are those of authors solely and do not necessarily reflect Oracle's. Contact @jimgris or @brhubart

Maria Colgan

Written by

Oracle Developers

Aggregation of articles from Oracle & partners engineers, Groundbreaker ambassadors & the developer community on all things Oracle Cloud and its technologies. The views expressed are those of authors solely and do not necessarily reflect Oracle's. Contact @jimgris or @brhubart

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade