memcachio

A pure python async Memcached client with 0 dependencies with support for:

• All memcached commands
• Memcached servers serving on TCP or Unix Domain Sockets
• Memcached clusters
• SSL transport
• SASL Authentication
• Connection reuse for multiple concurrent requests
• Dynamically adjusted connection pooling
• Auto discovery with AWS ElastiCache

Installation

To install memcachio:

$ pip install memcachio

Quick start

Single Node or Cluster client

import asyncio

from memcachio import Client


async def example() -> None:

    #: basic client
    raw_client = Client(("localhost", 11211))
    #: client that decodes the byte responses
    decoding_client = Client(("localhost", 11211), decode_responses=True)
    # or with a cluster
    # cluster_client = Client([("localhost", 11211), ("localhost", 11212)], decode_responses=True)

    await raw_client.flushall()
    await raw_client.set("foo", b"1")
    await raw_client.set("bar", b"2")

    assert 2 == await raw_client.incr("foo", 1)

    # use the raw client to get a value.
    # Note the mapping returned has byte keys
    assert (await raw_client.get("foo")).get(b"foo").value == b"2"

    # get the values with the decoding client and touch their expiry to be 1 second.
    # Note the mapping and the values are strings.
    gat_and_touch_many = await decoding_client.gat("foo", "bar", expiry=1)
    assert ["2", "2"] == [item.value for item in gat_and_touch_many.values()]

    await asyncio.sleep(1)

    assert {} == await decoding_client.get("foo", "bar")


asyncio.run(example())

See Client for detailed descriptions of available options when constructing a client.

Compatibility

memcachio is tested against memcached versions 1.6.x

Supported python versions

• 3.10
• 3.11
• 3.12
• 3.13
• PyPy 3.10
• PyPy 3.11

References

• Documentation (Stable)
• Documentation (Latest)
• Changelog