Primary index

Prior to Server 6.0, primary index (PI) queries were called scans and had policies defined through the scan policy. See Queries for more information.

Jump to the Code block for a combined complete example.

Basic PI queries have the following features:

Filter records by set name.
Filter records by filter expressions.
Limit the number of records returned, useful for pagination.
Return only record digests and metadata (generation and TTL).
Return specified bins.

Setup

The following examples will use the setup and record structure below to illustrate primary index queries in an Aerospike database.

import aerospike
from aerospike_helpers import expressions as exp
from aerospike_helpers.operations import operations

# Define host configuration
config = {
    'hosts': [ ('127.0.0.1', 3000) ]
}
# Establishes a connection to the server
client = aerospike.client(config)

The record structure:

Occurred: Integer
Reported: Integer
Posted: Integer
Report: Map
{
    shape: List,
    summary: String,
    city: String,
    state: String,
    duration: String
}
Location: GeoJSON

Policies

See Basic Queries for query policy information.

Query a set

The following example queries the sandbox namespace and ufodata set name, while limiting the record set to only 20 records.

# Create the query
query = client.query('sandbox', 'ufodata')

# Set max records to return
query.max_records = 20

# Create callback function
def record_set(record):
    (key, meta, bins) = record
    # Do something
    print('Key: {0} | Record: {1}'.format(key[2], bins))

# Execute the query
query.foreach(record_set)

# Close the connection to the server
client.close()

Query with a metadata filter

The following example queries the same namespace and set as the example above, but also adds a metadata Filter Expression that will only return records that are greater than 16 KiB.

# Build the expression
expr = exp.GT(exp.DeviceSize(), (1024 * 16)).compile()

# Create the policy
query_policy = {'expressions': expr}

# Create the query
query = client.query('sandbox', 'ufodata')

# Create callback function
def record_set(record):
    (key, meta, bins) = record
    # Do something
    print('Key: {0} | Record: {1}'.format(key[2], bins))

# Execute the query
query.foreach(record_set, policy=query_policy)

# Close the connection to the server
client.close()

Query with a data filter

The following example queries the same namespace and set as the example above, but also adds a data Filter Expression that will only return records where the occurred bin value is in the inclusive range 20200101 to 20211231.

Take a look at secondary index queries to see how this same query can be run more efficiently with an index.

# Build the expression
expr = exp.Let(
    exp.Def('bin', exp.IntBin('occurred')),
    exp.And(
        exp.GE(exp.Var('bin'), 20210101),
        exp.LE(exp.Var('bin'), 20211231))).compile()

# Create the policy
query_policy = {'expressions': expr}

# Create the query
query = client.query('sandbox', 'ufodata')

# Create callback function
def record_set(record):
    (key, meta, bins) = record
    # Do something
    print('Key: {0} | Record: {1}'.format(key[2], bins))

# Execute the query
query.foreach(record_set, policy=query_policy)

# Close the connection to the server
client.close()

Pagination

Pagination uses a combination of a partition filter and a defined maximum records to return query results. The partition filter maintains a cursor identifying the end of the current page and the beginning of the next. Moving to the next page of results is as simple as executing the query again, with the previously defined partition filter.

Defining a maximum number of records per page to return guarantees that no page will contain more than the maximum number, but some pages may contain fewer than the maximum number. Also, if you run the same paginated query multiple times, the number of results per page may differ, depending on the order in which they are delivered by the nodes in the cluster.

The following example executes a query with an Expression Filter identifying records with more than 3 shape items in the report map, returning 10 records per page. The partition filter is set to query all 4096 partitions in the database.

# Build the expression
expr = exp.GT(
    exp.ListSize(
        None,
        exp.MapGetByKey(None, aerospike.MAP_RETURN_VALUE, exp.ResultType.LIST, 'shape', exp.MapBin('report'))
    ),
    3).compile()

# Create the policy
query_policy = {'expressions': expr}

# Create the query
query = client.query('sandbox', 'ufodata')

# Set max records
query.max_records = 10

# Set pagination
query.paginate()

# Execute the query
page = 0
while not query.is_done():
    count = 0
    records = query.results(query_policy)
    for record in records:
        (key, meta, bins) = record
        # Do something
        print('Key: {0} | Record: {1}'.format(key[2], bins))
        count += 1
    page += 1
    print('Page {0} | {1} records'.format(page, count))

# Close the connection to the server
client.close()

Code block

Expand this section for a single code block to execute a basic PI query

import aerospike
from aerospike_helpers import expressions as exp
from aerospike_helpers.operations import operations

# Define host configuration
config = {
    'hosts': [ ('127.0.0.1', 3000) ]
}
# Establishes a connection to the server
client = aerospike.client(config)

# Build the expression
expr = exp.Let(
    exp.Def('bin', exp.IntBin('occurred')),
    exp.And(
        exp.GE(exp.Var('bin'), 20210101),
        exp.LE(exp.Var('bin'), 20211231))).compile()

# Create the policy
query_policy = {'expressions': expr}

# Create the query
query = client.query('sandbox', 'ufodata')

# Create callback function
def record_set(record):
    (key, meta, bins) = record
    # Do something
    print('Key: {0} | Record: {1}'.format(key[2], bins))

# Execute the query
query.foreach(record_set, policy=query_policy)

# Close the connection to the server
client.close()