One attribute of AutoNOC is that it works like a black box, or a flight recorder storing a detailed by minute history of everything happening across the network in all systems and applications. When a problem occurs, troubleshooting it is little more than looking back in time to see what, where, and when it happened. In many cases the recorded, categorized, and hierarchically organized information will provide enough clues into why and how the problem occurred!

The following screenshot shows an example high-resolution archive of traffic data for the internal AutoNOC network.

The Recoiling Database (RDB)
Managing, archiving, and accessing the vast volumes of information AutoNOC acquires is certainly a non-trivial exploit. Legacy management solutions have used SQL servers, text files, and other third party forms of storage to try and maintain the vast quantities of data a long term historical record of network activity requires.

AutoNOC's compressing recoiling database technology solves the data warehousing problems customers often have to deal with when using legacy systems. Recoiling database technology collects data in memory and than, based on an interval "recoils" the database. A spring stretches out over time and when released it snaps back into a tight coil. Recoiling database technology works in much the same way. It collects the typically sparse data over time and then "tightens the coil" periodically.

This automated, background process works extremely well for the type of data that AutoNOC acquires (sparse with all records growing simultaneously).

AutoNOC can typically store about 1 year of high resolution probe data in about a megabyte of disk storage. The database is indexed and provides high performance windowed access to the information stored within the recoiling database. This architecture has proven ideal for the types of reads and writes that occur within real world operations management scenarios.