Important: This feature is in limited release. Before using the low-latency match (LLM) service, contact Tamr Support at [email protected] to discuss your use case and for configuration assistance.
You can tune the throughput and latency of the LLM service.
- Throughput is the total time to match a set of records.
- Latency is the total time between request and first response.
Before You Begin:
Parts of the LLM service use port
9170. To use the service, open this port.
To configure throughput and latency for the LLM service:
- Set the value of the configuration variable
TAMR_LLM_BATCH_SIZEto the expected largest batch size using the administration utility. See Creating or Updating a Configuration Variable.
The default value is
- A lower value reduces latency while decreasing throughput.
- A higher value increases latency while increasing throughput.
- Restart Tamr Core and its dependencies. See Restarting Tamr Core
Updated 3 months ago