Configure Nodes and Assign Server Tags

Configure the nodes you need and assign functions to them.

  2 minute read  

Once installed, we suggest that you configure several nodes and assign functions using server tags. Documentation on how to do so can be found in the cluster configuration section of the documentation.

Prior to running any tests, we also recommend you configure unified logging within Elastic as described in the Unified Logging section of the documentation.

An example use case would be to provision five (5) servers, each with at least 32 gigabytes (GB) of memory (the more, the better).

Assign and configure nodes

Assign one node each to the following:

  1. Search
    1. Allocate 24 GB of memory to the application server instance running Nuix REST. Reserve 8 GB for the OS and other processes.
    2. Designate this node as a producer node.
  2. Ingestion
    1. Allocate 8 GB of memory to the application server instance running Nuix REST. Reserve other heap space for Nuix workers and other processes. For example, 2 GB each for eight (8) workers leaves 8 GB free.
    2. Designate this node as a consumer node.
  3. OCR
    1. Allocate 8 GB of memory to the application server instance running Nuix REST. Reserve other heap space for Nuix workers and other processes. For example, 1 GB each for 16 workers leaves 8 GB free.
    2. Install the Nuix OCR Plugin and configure a very fast disk for cache.
    3. Designate this node as a consumer node.
    4. Add a server tag designating this node for “ocr”.
  4. Export
    1. Allocate 8 GB of memory to the application server instance running Nuix REST. Reserve other heap space for Nuix workers and other processes. For example, 1 GB each for 16 workers leaves 8 GB free.
    2. Ensure a very fast disk for cache and that your export directory file system is tuned for a large number of IOPS with sufficient storage capacity.
    3. Designate this node as a consumer node.
    4. Add a server tag designating this node for “export”.

Issue all calls to the producer node

Using Postman or another REST client, issue all calls to the producer node, referenced above in item 1 (Search). Do the following:

  1. Create several Nuix cases.
  2. Create multiple long-running ingestion jobs for each of these cases.
  3. Monitor the asynchronous ingestion functions.
  4. As ingestion jobs complete, invoke search and utilize server tags to submit OCR and Export requests to the cluster.
  5. Ensure each function is triaged to the appropriate node within the cluster. Verify the unified logging mechanism or async function payloads using Kibana.

Configure REST to ship logs to Elasticsearch

We recommend you configure unified logging with Elasticsearch as described in the Unified Logging section of the documentation.