Retry documentation. (#2166) (#2456)

* Retry documentation. (#2166)

* Fixed typo.

* Update docs/retry.rst

Co-authored-by: Nermina Miller <[email protected]>

* Update docs/retry.rst

Co-authored-by: Nermina Miller <[email protected]>

* Update docs/retry.rst

Co-authored-by: Nermina Miller <[email protected]>

* Update docs/retry.rst

Co-authored-by: Nermina Miller <[email protected]>

* Update docs/retry.rst

Co-authored-by: Nermina Miller <[email protected]>

* Update docs/retry.rst

Co-authored-by: Nermina Miller <[email protected]>

* Update docs/retry.rst

Co-authored-by: Nermina Miller <[email protected]>

* Update docs/retry.rst

Co-authored-by: Nermina Miller <[email protected]>

Co-authored-by: Nermina Miller <[email protected]>

Author: lucacillario

docs/backoff.rst

@@ -1,3 +1,5 @@
+.. _backoff-label:
+
 Backoff
 #############
 

docs/exceptions.rst

@@ -1,4 +1,4 @@
-
+.. _exceptions-label:
 
 Exceptions
 ##########

docs/retry.rst

@@ -2,4 +2,69 @@ Retry Helpers
 #############
 
 .. automodule:: redis.retry
-    :members: 
+    :members:
+
+
+Retry in Redis Standalone
+**************************
+
+>>> from redis.backoff import ExponentialBackoff
+>>> from redis.retry import Retry
+>>> from redis.client import Redis
+>>> from redis.exceptions import (
+>>>    BusyLoadingError,
+>>>    ConnectionError,
+>>>    TimeoutError
+>>> )
+>>>
+>>> # Run 3 retries with exponential backoff strategy
+>>> retry = Retry(ExponentialBackoff(), 3)
+>>> # Redis client with retries on custom errors
+>>> r = Redis(host='localhost', port=6379, retry=retry, retry_on_error=[BusyLoadingError, ConnectionError, TimeoutError])
+>>> # Redis client with retries on TimeoutError only
+>>> r_only_timeout = Redis(host='localhost', port=6379, retry=retry, retry_on_timeout=True)
+
+As you can see from the example above, Redis client supports 3 parameters to configure the retry behaviour:
+
+* ``retry``: :class:`~.Retry` instance with a :ref:`backoff-label` strategy and the max number of retries
+* ``retry_on_error``: list of :ref:`exceptions-label` to retry on
+* ``retry_on_timeout``: if ``True``, retry on :class:`~.TimeoutError` only
+
+If either ``retry_on_error`` or ``retry_on_timeout`` are passed and no ``retry`` is given,
+by default it uses a ``Retry(NoBackoff(), 1)`` (meaning 1 retry right after the first failure).
+
+
+Retry in Redis Cluster
+**************************
+
+>>> from redis.backoff import ExponentialBackoff
+>>> from redis.retry import Retry
+>>> from redis.cluster import RedisCluster
+>>>
+>>> # Run 3 retries with exponential backoff strategy
+>>> retry = Retry(ExponentialBackoff(), 3)
+>>> # Redis Cluster client with retries
+>>> rc = RedisCluster(host='localhost', port=6379, retry=retry, cluster_error_retry_attempts=2)
+
+Retry behaviour in Redis Cluster is a little bit different from Standalone:
+
+* ``retry``: :class:`~.Retry` instance with a :ref:`backoff-label` strategy and the max number of retries, default value is ``Retry(NoBackoff(), 0)``
+* ``cluster_error_retry_attempts``: number of times to retry before raising an error when :class:`~.TimeoutError` or :class:`~.ConnectionError` or :class:`~.ClusterDownError` are encountered, default value is ``3``
+
+Let's consider the following example:
+
+>>> from redis.backoff import ExponentialBackoff
+>>> from redis.retry import Retry
+>>> from redis.cluster import RedisCluster
+>>>
+>>> rc = RedisCluster(host='localhost', port=6379, retry=Retry(ExponentialBackoff(), 6), cluster_error_retry_attempts=1)
+>>> rc.set('foo', 'bar')
+
+#. the client library calculates the hash slot for key 'foo'.
+#. given the hash slot, it then determines which node to connect to, in order to execute the command.
+#. during the connection, a :class:`~.ConnectionError` is raised.
+#. because we set ``retry=Retry(ExponentialBackoff(), 6)``, the client tries to reconnect to the node up to 6 times, with an exponential backoff between each attempt.
+#. even after 6 retries, the client is still unable to connect.
+#. because we set ``cluster_error_retry_attempts=1``, before giving up, the client starts a cluster update, removes the failed node from the startup nodes, and re-initializes the cluster.
+#. after the cluster has been re-initialized, it starts a new cycle of retries, up to 6 retries, with an exponential backoff.
+#. if the client can connect, we're good. Otherwise, the exception is finally raised to the caller, because we've run out of attempts.