Throttling the Automated Clients

The automated clients are each configured by default to use all of the cores on the system minus 1 (to leave room for the OS). This configuration is designed to maximize the CPU when each automated client is deployed on a separate machine/VM.

When deploying multiple or all ancoraDocs automated clients onto a single machine/VM it is recommended to throttle the number of cores each automate clients can consume.

To determine the appropriate number workers/CPU cores for each service see the ancoraDocs Sizing Calculator

OCR Service

  1. Open the file SSiOcrClient.xml in the root of the “ancoraDocs Automated Clients” installation directory
  2. Change the value in the MAXWorkerCount from -1 to the desired number cores OCR should consume.

Classification Service

  1. Open the file SSiClassificationClient.xml in the root of the “ancoraDocs Automated Clients” installation directory
  2. Change the value in the MAXWorkerCount from -1 to the desired number cores Classification should consume.
  3. Change the value in the MAXOCRWorkersPerClient from 4 to the desired number of sub worker processes Classification should consume.

For each Worker, the Classification service will spawn separate sub processes “OCR Workers” to analyze the OCR content from each page. Each of these “OCR Workers” will consume 100% of a single core’s throughput.

Data Capture Service

  1. Open the file SSiDataCaptureClient.xml in the root of the “ancoraDocs Automated Clients” installation directory
  2. Change the value in the MAXWorkerCount from -1 to the desired number cores Data Capture should consume.

Export Service

  1. Open the file SSiExportClient.xml in the root of the “ancoraDocs Automated Clients” installation directory
  2. Change the value in the MAXWorkerCount from -1 to the desired number cores Export should consume.