Skip to main content
Version: 2.10.1

Inspecting queries

Before a Cypher query is executed, it is converted into an internal form suitable for execution, known as a plan. A plan is a tree-like data structure describing a pipeline of operations which will be performed on the database in order to yield the results for a given query. Every node within a plan is known as a logical operator and describes a particular operation.

Because a plan represents a pipeline, the logical operators are iteratively executed as data passes from one logical operator to the other. Every logical operator pulls data from the logical operator(s) preceding it, processes it and passes it onto the logical operator next in the pipeline for further processing.

Using the EXPLAIN clause, it is possible for the user to inspect the produced plan and gain insight into the execution of a query.

Operators

OperatorDescription
AccumulateAccumulates the input it received.
AggregateAggregates the input it received.
ApplyJoins the returned symbols from two branches of execution.
CallProcedureCalls a procedure.
CartesianApplies the Cartesian product (the set of all possible ordered combinations consisting of one member from each of those sets) on the input it received.
ConstructNamedPathCreates a path.
CreateNodeCreates a node.
CreateExpandCreates edges and new nodes to connect with existing nodes.
DeleteDeletes nodes and edges.
EdgeUniquenessFilterFilters unique edges.
EmptyResultDiscards results from the previous operator.
EvaluatePatternFilterPart of the filter operator that contains a sub-branch which yields either true or false.
ExpandExpands the node by finding the node's relationships.
ExpandVariablePerforms a node expansion of a variable number of relationships
FilterFilters the input it received.
ForeachIterates over a list and applies one or more update clauses.
LimitLimits certain rows from the pull chain.
LoadCsvLoads CSV file in order to import files into the database.
MergeApplies merge on the input it received.
OnceForms the beginning of an operator chain with "only once" semantics. The operator will return false on subsequent pulls.
OptionalPerforms optional matching.
OrderByOrders the input it received.
ProduceProduces results.
RemoveLabelsRemoves a variable number of node labels.
RemovePropertyRemoves a node or relationship property.
ScanAllProduces all nodes in the database.
ScanAllByIdProduces nodes with a certain index.
ScanAllByLabelProduces nodes with a certain label.
ScanAllByLabelPropertyProduces nodes with a certain label and property.
ScanAllByLabelPropertyRangeProduces nodes with a certain label and property value within the given range (both inclusive and exclusive).
ScanAllByLabelPropertyValueProduces nodes with a certain label and property value.
SetLabelsSets node labels of variable length.
SetPropertySets a node or relationship property.
SetPropertiesSets a list of node or relationship properties.
SkipSkips certain rows from the pull chain.
UnwindUnwinds an expression to multiple records.
DistinctApplies a distinct filter on the input it received.

Example plans

As an example, let's inspect the plan produced for a simple query:

EXPLAIN MATCH (n) RETURN n;
+----------------+
| QUERY PLAN |
+----------------+
| * Produce {n} |
| * ScanAll (n) |
| * Once |
+----------------+

The output of the query using the EXPLAIN clause is a representation of the produced plan. Every logical operator within the plan starts with an asterisk character (*) and is followed by its name (and sometimes additional information). The execution of the query proceeds iteratively (generating one entry of the result set at a time), with data flowing from the bottom-most logical operator(s) (the start of the pipeline) to the top-most logical operator(s) (the end of the pipeline).

In the example above, the resulting plan is a pipeline of 3 logical operators. Once is the identity logical operator which does nothing and is always found at the start of the pipeline; ScanAll is a logical operator which iteratively produces all of the nodes in the graph; and Produce is a logical operator which takes data produced by another logical operator and produces data for the query's result set.

A slightly more complicated example would be:

EXPLAIN MATCH (n :Node)-[:Edge]-(m :Node) WHERE n.prop = 42 RETURN *;
+--------------------------------+
| QUERY PLAN |
+--------------------------------+
| * Produce {m, n} |
| * Filter |
| * Expand (m)-[anon1:Edge]-(n) |
| * ScanAllByLabel (n :Node) |
| * ScanAllByLabel (m :Node) |
| * Once |
+--------------------------------+

In this example, the Filter logical operator is used to filter the matched nodes because of the WHERE n.prop = 42 construct. The Expand logical operator is used to find an edge between two nodes, in this case m and n which were matched previously using the ScanAllByLabel logical operator (a variant of the ScanAll logical operator mentioned previously).

The execution of the query proceeds iteratively as follows. First, two vertices of type :Node are found as the result of the two scans. Then, we try to find a path that consists of the two vertices and an edge between them. If a path is found, it is further filtered based on a property of one of the vertices. Finally, if the path satisfied the filter, its two vertices are added to the query's result set.

A simple example showcasing the fully general tree structure of the plan could be:

EXPLAIN MERGE (n) RETURN n;
+------------------+
| QUERY PLAN |
+------------------+
| * Produce {n} |
| * Accumulate |
| * Merge |
| |\ On Match |
| | * ScanAll (n) |
| | * Once |
| |\ On Create |
| | * CreateNode |
| | * Once |
| * Once |
+------------------+

The Merge logical operator (constructed as a result of the MERGE construct) can take input from up to 3 places. The On Match and On Create branches are "pulled from" only if a match was found or if a new vertex has to be created, respectively.

Where to next?

To learn more about Memgraph's functionalities, visit the Reference guide. For real-world examples of how to use Memgraph, we strongly suggest going through one of the available Tutorials.