KRvdld—VocDelta load
This event is logged when domain language models or wordset files are loaded.
VocDelta files consist of dynamic configuration items that customize the recognition at runtime; for example, JSON wordsets, compiled wordsets, and domain language models. A single load command can define any number of the above items; items are handled in parallel. All items are specified via URLs. JSON wordsets can also be specified inline.
In addition to the Tokens used for every event, this event has the following tokens:
| Token | Meaning |
|---|---|
| FETCHES | Number of HTTP fetches required to complete the Load command. |
| IDx | ID associated with a resource, where x is an integer enumerating the domain language model or wordset files are loaded (see Note below). |
| IFBYTESx | Total bytes fetched. |
| IFCPU | Total CPU milliseconds (user and system) consumed during fetches (aggregate measure covering the entire time span of all fetch activity). |
| IFTIMEx | Total clock-time milliseconds to complete the fetch. |
| LDCPU | Total CPU milliseconds (user and system) consumed during the entire Load command. |
| LDTIME | Total clock-time milliseconds to complete the entire Load command. |
| MEMHITS | HTTP client cache hits for this load operation. |
| MEMMISS | HTTP client cache misses for this load operation. |
| RC | Return code for the API call. |
| TYPEx | MIME type for the load.
Mime type of the URI as specified in the Load command request or, if not specified, value determined from the server response. |
| URIx | URI for the load, as specified in the Load command. Note: In-line bodies for wordsets are conveyed as |
Note: The IDx, IFBYTESx, IFTIMEx, TYPEx, and URIx token sequence will appear with 0-based indices, as many times are there are items specified in the load. IFTIME and IFBYTES values will be 0 for cache hits.
The Krypton engine logs the URIs of all active domain language models or wordset files at KRvdld. The URI of the files are enumerated and stored in a field called URIx, where x is an integer enumerating the domain language model or wordset files loaded, starting at 0.
