90 lines
3.0 KiB
Python
90 lines
3.0 KiB
Python
"""Consistent hashing ring with virtual nodes.
|
|
|
|
A minimal, dependency-free implementation suitable for distributing keys
|
|
across a dynamic set of nodes with minimal rebalancing.
|
|
"""
|
|
import bisect
|
|
import hashlib
|
|
from typing import Any, Iterable, List, Optional
|
|
|
|
|
|
class ConsistentHashRing:
|
|
def __init__(self, num_vnodes: int = 100):
|
|
if not isinstance(num_vnodes, int) or num_vnodes < 1:
|
|
raise ValueError("num_vnodes must be a positive int")
|
|
self.num_vnodes = num_vnodes
|
|
self._ring: dict[int, Any] = {} # position -> node
|
|
self._sorted: List[int] = [] # sorted positions
|
|
self._nodes: set = set()
|
|
|
|
@staticmethod
|
|
def _hash(key: Any) -> int:
|
|
# 128-bit digest collapsed to int; deterministic across processes.
|
|
return int(hashlib.md5(str(key).encode("utf-8")).hexdigest(), 16)
|
|
|
|
def add_node(self, node: Any) -> None:
|
|
if node in self._nodes:
|
|
return
|
|
self._nodes.add(node)
|
|
for i in range(self.num_vnodes):
|
|
pos = self._hash(f"{node}#{i}")
|
|
# handle (extremely unlikely) collision by nudging
|
|
while pos in self._ring:
|
|
pos = (pos + 1) & ((1 << 128) - 1)
|
|
self._ring[pos] = node
|
|
bisect.insort(self._sorted, pos)
|
|
|
|
def remove_node(self, node: Any) -> None:
|
|
if node not in self._nodes:
|
|
return
|
|
self._nodes.discard(node)
|
|
# rebuild positions belonging to node
|
|
to_delete = [p for p, n in self._ring.items() if n == node]
|
|
for p in to_delete:
|
|
del self._ring[p]
|
|
idx = bisect.bisect_left(self._sorted, p)
|
|
if idx < len(self._sorted) and self._sorted[idx] == p:
|
|
del self._sorted[idx]
|
|
|
|
@property
|
|
def nodes(self) -> List[Any]:
|
|
return sorted(self._nodes, key=lambda x: str(x))
|
|
|
|
def get_node(self, key: Any) -> Optional[Any]:
|
|
if not self._sorted:
|
|
return None
|
|
pos = self._hash(key)
|
|
idx = bisect.bisect_left(self._sorted, pos)
|
|
if idx == len(self._sorted):
|
|
idx = 0
|
|
return self._ring[self._sorted[idx]]
|
|
|
|
def get_nodes(self, key: Any, n: int) -> List[Any]:
|
|
if n < 1:
|
|
raise ValueError("n must be >= 1")
|
|
if not self._sorted:
|
|
return []
|
|
result: List[Any] = []
|
|
seen: set = set()
|
|
pos = self._hash(key)
|
|
idx = bisect.bisect_left(self._sorted, pos)
|
|
if idx == len(self._sorted):
|
|
idx = 0
|
|
start = idx
|
|
# walk ring collecting distinct nodes
|
|
i = 0
|
|
total = len(self._sorted)
|
|
while len(result) < n and i < total:
|
|
node = self._ring[self._sorted[(start + i) % total]]
|
|
if node not in seen:
|
|
seen.add(node)
|
|
result.append(node)
|
|
i += 1
|
|
return result
|
|
|
|
def __len__(self) -> int:
|
|
return len(self._nodes)
|
|
|
|
def __contains__(self, node: Any) -> bool:
|
|
return node in self._nodes
|