Performance improvements due to extra parameter handling changes
To significantly improve throughput and latency, the event protocol previously used to handle extra parameters has been redesigned to separate schema data from values. Values of extra parameters are now contained in the market data events, which removes the need for a second event.
For book based market data (Orderbook, Depth, Quotebook and AggregatedBook), the schema (type and description data) for any extra parameters is sent upon the initial connection. Each level or entry now has its own set of extra parameter values. This removes the need for extra parameters to be encoded into a top-level extra parameter value. Also, the schema can now be delta updated, which removes the need for the large snapshot events that were sent in previous releases.
Two new EPL interfaces provide access to extra parameters:
com.apama.md.client.EPSchemaInterface — Each market data interface now provides the
getEPSchemaInterface() action, which you can call to access the schema for any extra parameters.
com.apama.md.client.EPValuesInterface — Each market data interface and each book entry interface now provides the
getEPValuesInterface() action, which you can call to access the values of any extra parameters.
The AggregatedBook actions and interfaces are slightly different as they offer access to the extra parameters from the underlying datasources of the Aggregator/FTSC. They require the symbol and source id (known as the XEPKey) of the underlying datasource to access the values and schema.
See the ApamaDoc for details about using these interfaces.
For an example of using the new interfaces, see
Accessing extra parameters .