Connecting to Elasticsearch

Prerequisites

• PuppyGraph version 1.3.0 or higher.
• Every configured Elasticsearch host is accessible over the network from the PuppyGraph instance.

Configuration

Configuration	Explanation
Username	The username used to authenticate with the Elasticsearch cluster
Password	The password of the user
Server Hosts	One or more Elasticsearch HTTP endpoints, e.g. http://es-node1:9200. Add one entry per node, or a single entry pointing at a load balancer in front of the cluster. Both http and https endpoints are supported.

Authentication

PuppyGraph authenticates to Elasticsearch with basic authentication (username and password). Other Elasticsearch authentication mechanisms, such as API keys, service account tokens, and client certificates (PKI), are not currently supported.

Leave Username and Password empty only if the cluster runs with security disabled.

TLS

To connect over TLS, use https:// endpoints in Server Hosts. No further TLS configuration is required: PuppyGraph does not require the cluster's certificate to be issued by a publicly trusted authority, so clusters using self-signed or private-CA certificates work without importing a CA certificate into PuppyGraph.

Mutual TLS (client certificate authentication) is not supported; see Authentication for the supported mechanism.

How Indices Are Mapped

Elasticsearch has no databases or tables, so PuppyGraph presents every index of the cluster as a table under a single database named default_db. Each document becomes a row, and each mapped field becomes a column. When selecting data for nodes and edges, use default_db as the schema and the index name as the table.

Demo

See Querying Elasticsearch Data as a Graph for a complete demo for Elasticsearch. The demo includes setting up the connection to an Elasticsearch instance.

Data Type Mapping

Elasticsearch Type	PuppyGraph Type	Description
boolean	Boolean	True or false
byte	Byte	8-bit signed integers
short	Short	16-bit signed integers
integer	Int	32-bit signed integers
long	Long	64-bit signed integers
unsigned_long	HugeInt	Unsigned 64-bit integers
half_float	Float	Half-precision 16-bit floating point
float	Float	32-bit IEEE 754 floating point
double	Double	64-bit IEEE 754 floating point
scaled_float	Double	Floating point scaled by a fixed factor
keyword	String	Exact-value strings for filtering and identifiers
text	String	Analyzed full-text strings
wildcard	String	Strings optimized for wildcard matching
ip	String	IPv4 and IPv6 addresses
date	DateTime	Timestamps
object	JSON	Nested JSON objects
nested	JSON	Arrays of objects indexed as separate documents
geo_point	String	Latitude and longitude pairs

Use keyword fields (rather than text) for document identifiers and the fields that connect nodes and edges, since these are matched by exact value.