Configuring Specific Node Roles

If a specific area of the performance within LogScale has been identified, increasing of the number of nodes within the cluster will expand the capacity across all the different vectors.

LogScale has a combination of roles governed through a combination of the NODE_ROLES variable, and the configuration of the digest and storage roles that configure which roles in the system take one on which responsibility:

  • NODE_ROLE=all is the default logical role for all nodes, including Digest and Storage nodes. Whether a node is used for digest or storage is determined by the Digest Rules and Storage Rules

  • NODE_ROLE=httponly is a logical role that handles the UI, acts as a query coordinator and accepts ingest and parsing responsibilities.

  • NODE_ROLE=ingestonly is a logical role that handles only ingesting and parsing.

To target a specific area, identify whether a specific role will will have an impact on the desired capacity vector:

  • Adding ingest nodes

    Improves the rate at which data can be accepted and parsed into the system before the data is passed on to the Kafka cluster.

Important

Changing the NODE_ROLES of a host is forbidden after it has been added to the cluster. A host will crash if the role it is configured to have doesn't match what is listed in the global database for that host. If you want to change the role of a host in a cluster you need to remove that host from the cluster by unregistering it, wipe the data directory of the host, and then boot the node back into the cluster as if it were a completely new node. The node will then be assigned a new vhost identifier.