httpClient.executorMaxThreads
Maximum number of threads created by individual Dragon Voice services.
Value |
Integer from 0 to Java max int |
Default |
400 |
How to set |
In Management Station, set on the Natural Language Processing service. If not using Management Station, set in the NLP service configuration file (User-nlpsxx.properties, see Configuration roadmap). |
Usage |
Rarely changed. Consider the default as a high value for a system operating under heavy loads. (Historically, Nuance used a value less than 100.) There is little known cost to memory or CPU. |
Specify the same value for each NLP service instance in your network. (Each services gets the same value. The services do not share a single value.) NLP service acts as an HTTP client as follows:
- To fetch (http GET) the manifest file for Dragon Voice activities.
- To allocate (http GET .../getResource) the Krypton and NLE engines from the Resource Manager.
- To initiate websocket connections with the allocated engines.