Splunk has the ability to read all kinds of data—in any format—from any device or application. Its power lies in its ability to turn this data into operational intelligence (OI), typically out of the box and without the need for any special parsers or adapters to deal with particular data formats.
Splunk uses internal algorithms to process new data and new data sources automatically and efficiently. Once Splunk is aware of a new data type, you don't have to reintroduce it again, saving time.
Since Splunk can work with both local and remote data, it is almost infinitely scalable. What this means is that the data that you are interested in can be on the same (physical or virtual) machine as the Splunk instance (meaning Splunk's local data) or on an entirely different machine, practically anywhere in the world (meaning it is remote data). Splunk can even take advantage of Cloud-based data.
Generally speaking, when you are thinking about Splunk and data, it is useful to categorize your data into one of the four types of data sources.
In general, one can categorize Splunk data (or input) sources as follows:
Files and/or directories: This is the data that exists as physical files or locations where files will exist (directories or folders).
Network events: This will be the data recorded as part of a machine or environment event.
Windows sources: This will be the data pertaining to MS Windows' specific inputs, including event logs, registry changes, Windows Management Instrumentation, Active Directory, exchange messaging, and performance monitoring information.
Other sources: This data source type covers pretty much everything else, such as mainframe logs, FIFO queues, and scripted inputs to get data from APIs and other remote data interfaces.