I have been talking a lot lately about our Axellio Architecture for edge computing and analytics on streaming big data at the edge. The question that comes up is, “where is the edge?”
Most of us are aware and understand the role that Content Distribution Networks play in making access to websites and web applications faster and less expensive. CDNs cache data at the edge, so that getting to that data does not always require traversing wide area networks to get all the way to the data’s original source. Edge computing is somewhat analogous to CDNs, except edge computing deals with data coming in that must be ingested, stored, processed and acted on rather than dealing only with data being accessed as with CDNs. The analogy to CDNs helps us think about the edge in the context of something that is by now very familiar to many. When there is a lot of data coming in at very high rates that must be ingested, stored, and processed in near real time, moving it across the internet or wide area networks to a data center or to the cloud would take too long and cost too much. Where that data is created and captured is where “the edge” is.
Axellio is uniquely suited for ingesting big data fast and storing a lot of it while simultaneously processing it to do things like comparing the most recently ingested data against the stored data set to detect anomalies in real time. Axellio’s performance makes it practical to simultaneously run forensic analysis against the entire data set in place to explore whether detected anomalies represent a current threat, all the while continuing to ingest data without slowing down or dropping anything. Problems like this occur in applications like cybersecurity, detecting fraudulent or erroneous activity in financial markets, or identifying that something is going awry with operations of critical infrastructure or industrial processes.
For insights and actions to come soon enough to be truly useful, big data must often be ingested, stored, and processed right there at the edge where it is created. That is what FabricXpress™ Architecture is uniquely designed to do.