Splunk, Java and “The Internet of Things”

Spending some time in Asia this week has only further reminded me of how many machine data generating devices permeate our modern lives.Mobile devices, CCTV cameras, Car computers, Hotel smart TVs, Traffic controllers, Payment terminals, Public transport tap n go passes,Wifi access points, Automated laser and light shows etc….All connected systems that generate massive amounts of data. And this is just what you can see on the surface.There are a myriad of embedded systems and controllers running quietly under the covers,  churning away, powering the very lifeblood  of the city.

A recent article on had some interesting quotes that got my data wrangling adrenaline flowing :

“….there are over three billion embedded devices out there powered by Java, and that number is expected to exceed one hundred billion by the end of the decade….”

“….with about 97 billion more embedded applications predicted to be developed in the next handful of years…”

Oracle had their JavaOne conference recently, and this year included a dedicated Embedded program.

As the new commercial stewards of Java, I found it particularly interesting that Oracle are really taking the Embedded Java space so seriously now.This is great for Big Data platforms like Splunk that are a perfect compliment for handling the variability and velocity of data that embedded devices create.Furthermore, embedded devices are typically rather resource constrained (storage, CPU) , so you can simply stream your data over the network to Splunk and let Splunk be the quilter extraordinaire that stitches together your data fabric into a stunning technicolor datacoat. That’s all my fabric related references used up in one sentence by the way.Ok, moving right along….

A key component of the Oracle Embedded Java offering is bundling the Glassfish Application server into the Embedded Suite.Wait a moment ? Did you just say Application Server on the Embedded client side ? Yes I did, and I haven’t been drinking.Things really are coming full circle , after all in the early 90’s Java starting of as a technology for embedded and consumer devices, before cementing its foothold in the server side space today.But you know , it’s all about lowering the barrier of entry for developers and making it as easy as possible for them to get up and running and being productive with creating their embedded solutions, just like what we are doing at Splunk in developing our various language SDKs to enable developers to integrate and create big data applications as simply as possible.
We already have a very comprehensive Java SDK for the Splunk platform that will allow Java Embedded systems to easily get their machine data into Splunk.Furthermore the embedded devices could run Splunk searches, offloading that computational workload to Splunk, process the results from the operational insights returned from Splunk that might correlate the machine data from the embedded devices across other related data that Splunk has indexed and even make operational decisions in the embedded application based on the outputs of the Splunk searches.

The possibilities are as vast as the data itself.

Damien Dallimore

Posted by


Show All Tags
Show Less Tags