An Ultipa instance can host multiple graphs, each representing a dataset of interconnected nodes and edges.
Show graphs in the database:
GQLSHOW GRAPHS
To inspect a single graph:
GQLDESCRIBE GRAPH myGraph -- DESC is a shorthand for DESCRIBE DESC GRAPH myGraph
Each graph provides the following metadata:
Field | Description |
|---|---|
graph_id | The ID of the graph. |
graph_name | The unique name of the graph. |
current | Whether the graph is the current graph. |
graph_type | The type of the graph (OPEN or CLOSED). |
node_count | The number of nodes in the graph. |
edge_count | The number of edges in the graph. |
node_label_count | The number of node labels in the graph. |
edge_label_count | The number of edge labels in the graph. |
procedure_count | The number of stored procedures in the graph. |
fulltext_index_count | The number of full-text indexes in the graph. |
trigger_count | The number of triggers in the graph. |
created_at | The creation time of the graph. |
comment | The comment of the graph. |
Most GQL queries operate on a specific graph. Use the USE statement to set the current graph:
GQLUSE myGraph -- USE is a shorthand for USE GRAPH USE GRAPH myGraph
All subsequent queries in the session will run against myGraph until another USE is issued.
Ultipa supports two main types of graphs: Open Graph and Closed Graph. This design offers both flexibility and control, supporting workflows ranging from agile exploration to production-grade applications demanding strict data integrity requirements.
Open graphs are schema-free, requiring no explicit schema definitions before data insertion. You can directly insert nodes and edges, and their labels and properties are created on the fly. This offers maximum flexibility for early-stage data exploration.
Learn more about open graphs →
Closed graphs require that any node or edge to be inserted must conform to a defined node or edge type. This ensures consistent structure, guaranteeing high data integrity and consistency.
Learn more about closed graphs →
By default, edges receive a system-generated _id and custom values cannot be assigned. The optional EDGE_ID feature lets edges carry a user-facing _id. Enable it at creation with WITH EDGE_ID:
GQLCREATE GRAPH myGraph WITH EDGE_ID
EDGE_ID can also be toggled on an existing graph.
Learn more about graphs with edge ID support →
Ontology support enables RDF/OWL-style class hierarchies and the @prefix:name label syntax. Create an ontology-enabled graph with WITH ONTOLOGY:
GQLCREATE GRAPH myOntologyGraph WITH ONTOLOGY
For details on ontology features, see Ontology.
You can use the IF NOT EXISTS clause to prevent errors when attempting to create a graph that already exists. It allows the statement to be safely executed.
GQLCREATE GRAPH IF NOT EXISTS myGraph
This creates the graph myGraph only if a graph with that name does not exist. If myGraph already exists, the statement is ignored without throwing an error.
A new graph can be created from an existing one, cloning both data and schema (if it's a closed graph):
GQLCREATE GRAPH newGraph AS COPY OF myGraph
A closed graph can be converted to an open graph, and vice versa.
Convert a closed graph to open. Existing type definitions are preserved but no longer enforced:
GQLALTER GRAPH myGraph SET OPEN
Convert an open graph to closed:
GQLALTER GRAPH myGraph SET CLOSED
After conversion, the graph has no node/edge types defined. You must add node and edge types before inserting new data. Existing data is not validated against the new types, only future inserts are checked.
Rename myGraph to newGraph:
GQLALTER GRAPH myGraph RENAME TO newGraph
Set comment for myGraph:
GQLALTER GRAPH myGraph COMMENT "This is a description"
Drop the graph myGraph:
GQLDROP GRAPH myGraph
The IF EXISTS clause is used to prevent errors when attempting to delete a graph that does not exist. It allows the statement to be safely executed.
GQLDROP GRAPH IF EXISTS myGraph
This deletes the graph myGraph only if a graph with that name does exist. If myGraph does not exist, the statement is ignored without throwing an error.
The truncating operation deletes all nodes, edges, and index data from the graph while preserving the graph itself. For closed graphs, the graph type is preserved.
You may truncate the entire graph, all nodes or edges, or nodes or edges with a specific label (open graph) or type (closed graph). Note that truncating nodes will also remove any edges connected to them.
GQL-- Truncate the entire graph TRUNCATE GRAPH myGraph -- Truncate all nodes in myGraph, all edges will also be removed TRUNCATE NODE * ON myGraph -- Truncate User nodes in myGraph, edges connected to them will also be removed TRUNCATE NODE User ON myGraph -- Truncate all edges in myGraph TRUNCATE EDGE * ON myGraph -- Truncate Follows edges in myGraph TRUNCATE EDGE Follows ON myGraph