Exponential with jitter backoff (#3550)

Author: JimNero009

redis/backoff.py

@@ -110,5 +110,20 @@ def compute(self, failures: int) -> float:
         return self._previous_backoff
 
 
+class ExponentialWithJitterBackoff(AbstractBackoff):
+    """Exponential backoff upon failure, with jitter"""
+
+    def __init__(self, cap: float = DEFAULT_CAP, base: float = DEFAULT_BASE) -> None:
+        """
+        `cap`: maximum backoff time in seconds
+        `base`: base backoff time in seconds
+        """
+        self._cap = cap
+        self._base = base
+
+    def compute(self, failures: int) -> float:
+        return min(self._cap, random.random() * self._base * 2**failures)
+
+
 def default_backoff():
     return EqualJitterBackoff()

tests/test_backoff.py

@@ -0,0 +1,18 @@
+from unittest.mock import Mock
+
+import pytest
+
+from redis.backoff import ExponentialWithJitterBackoff
+
+
+def test_exponential_with_jitter_backoff(monkeypatch: pytest.MonkeyPatch) -> None:
+    mock_random = Mock(side_effect=[0.25, 0.5, 0.75, 1.0, 0.9])
+    monkeypatch.setattr("random.random", mock_random)
+
+    bo = ExponentialWithJitterBackoff(cap=5, base=1)
+
+    assert bo.compute(0) == 0.25  # min(5, 0.25*2^0)
+    assert bo.compute(1) == 1.0  # min(5, 0.5*2^1)
+    assert bo.compute(2) == 3.0  # min(5, 0.75*2^2)
+    assert bo.compute(3) == 5.0  # min(5, 1*2^3)
+    assert bo.compute(4) == 5.0  # min(5, 0.9*2^4)