Stroom Proxy

Stroom-Proxy’s primary role is to act as a front door for data being sent to Stroom. Data can be sent to Stroom-Proxy in small chunks and it will aggregate the data into larger chunks (grouped by Feed and Stream Type ) so that Stroom doesn’t have to process lots of small Streams . It also provides a separation between the client and Stroom, so Stroom can be taken offline while data is still being accepted by Stroom-Proxy.

See Architecture for an example of how Stroom-Proxy is typically deployed.

API

Stroom-Proxy presents an identical HTTP POST /datafeed API to Stroom, so clients can send the same data in the same way to either Stroom or Stroom-Proxy. For more detail on sending data into Stroom-Proxy, see Sending Data.

It also presents a number of other APIs for administration and communication with other proxies. For more detail on Stroom-Proxy’s other APIs, see Proxy API.

Functions

Stroom-Proxy has a number of key functions:

• Receipt Filtering - The process of filtering the incoming data based on the HTTP headers. Data can either be Received, silently Dropped or Rejected with an error.
• Splitting - Splitting received ZIP files by Feed and Stream Type .
• Aggregation - Storing received data locally and forwarding it when the aggregation limits have been reached.
• Forwarding - Forwarding the received/aggregated data to one or more forward destinations.
• Instant Forwarding - Data is streamed to a single HTTP forward destination (i.e. Stroom or another Stroom-Proxy) as the data is received. This function does not support multiple forward destinations or aggregations.
• Directory Scanning - Periodically scanning one or more directories for ZIP files in Stroom ZIP Format.
• Event Store - Stroom-Proxy presents an API for receiving individual events. This is to support applications that want to log events directly to Stroom-Proxy rather than writing them to rolled files locally.

For a more detailed explanation of each function, see Proxy Functions.