Basic Terms

edit

This glossary covers ECE. If a term is not in this list, you might want to look in the Elasticsearch glossary.

administration console
A component of ECE that provides the API server for the Cloud UI. Also syncs cluster and allocator data from ZooKeeper to Elasticsearch.
allocator
Manages hosts that contain Elasticsearch and Kibana nodes. Controls the lifecycle of these nodes by creating new containers and managing the nodes within these containers when requested. Used to scale the capacity of your ECE installation.
availability zone
Contains resources available to a ECE installation that are isolated from other availability zones to safeguard against failure. Could be a rack, a server zone or some other logical constraint that creates a failure boundary. In a highly available cluster, the nodes of a cluster are spread across two or three availability zones to ensure that the cluster can survive the failure of an entire availability zone. Also see high availability.
beats runner
Used to send Filebeat and Metricbeat information to the logging cluster.
client forwarder
Used for secure internal communications between various components of ECE and ZooKeeper.
Cloud UI
Provides web-based access to manage your ECE installation, supported by the administration console.
cluster
Consists of one or more nodes which share the same cluster name.
constructor
Directs allocators to manage containers of Elasticsearch and Kibana nodes and maximizes the utilization of allocators. Monitors plan change requests from the Cloud UI and determines how to transform the existing cluster. In a highly available installation, places cluster nodes within different availability zones to ensure that the cluster can survive the failure of an entire availability zone.
container
Includes an instance of ECE software and its dependencies. Used to provision similar environments, to assign a guaranteed share of host resources to nodes, and to simplify operational effort in ECE.
coordinator
Consists of a logical grouping of some ECE services and acts as a distributed coordination system and resource scheduler.
director
Manages the ZooKeeper datastore. This role is often shared with the coordinator, though in production deployments it can be separated.
master node
Handles write requests for the cluster and publishes changes to other nodes in an ordered fashion. Each cluster has a single master node which is chosen automatically by the cluster and is replaced if the current master node fails. Also see node in this glossary and Node in the Elasticsearch documentation.
node
A running instance of Elasticsearch or Kibana which belongs to a cluster, typically one per host. Nodes will look to discover an existing cluster by name when they start up and try to join that cluster.
plan
Specifies the configuration and topology of an Elasticsearch or Kibana cluster, such as capacity, availability, and Elasticsearch version, for example. When changing a plan, the constructor determines how to transform the existing cluster into the pending plan.
proxy
A highly available, TLS-enabled proxy layer that routes user requests, mapping cluster IDs that are passed in request URLs for the container to the cluster nodes handling the user requests.
roles token
Enables a host to join an existing {ns installation and grants permission to hosts to hold certain roles, such as the allocator role. Used when installing ECE on additional hosts, a roles token helps secure ECE by making sure that only authorized hosts become part of the installation.
runner
A local control agent that runs on all hosts, used to deploy local containers based on role definitions. Ensures that containers assigned to it exist and are able to run, and creates or recreates the containers if necessary.
services forwarder
Routes data internally in an ECE installation.
stunnel
Securely tunnels all traffic in an ECE installation.
ZooKeeper
A coordination service for distributed systems used by ECE to store the state of the installation. Responsible for discovery of hosts, resource allocation, leader election after failure and high priority notifications.