1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
|
from __future__ import annotations
import logging
from typing import TYPE_CHECKING, Any, NewType
from dask.base import tokenize
from dask.highlevelgraph import HighLevelGraph
from dask.layers import SimpleShuffleLayer
logger = logging.getLogger("distributed.shuffle")
if TYPE_CHECKING:
import pandas as pd
from dask.dataframe import DataFrame
# circular dependency
from distributed.shuffle._worker_extension import ShuffleWorkerExtension
ShuffleId = NewType("ShuffleId", str)
def _get_worker_extension() -> ShuffleWorkerExtension:
from distributed import get_worker
try:
worker = get_worker()
except ValueError as e:
raise RuntimeError(
"`shuffle='p2p'` requires Dask's distributed scheduler. This task is not running on a Worker; "
"please confirm that you've created a distributed Client and are submitting this computation through it."
) from e
extension: ShuffleWorkerExtension | None = worker.extensions.get("shuffle")
if extension is None:
raise RuntimeError(
f"The worker {worker.address} does not have a ShuffleExtension. "
"Is pandas installed on the worker?"
)
return extension
def shuffle_transfer(
input: pd.DataFrame,
id: ShuffleId,
npartitions: int,
column: str,
) -> None:
try:
_get_worker_extension().add_partition(
input, id, npartitions=npartitions, column=column
)
except Exception:
raise RuntimeError(f"shuffle_transfer failed during shuffle {id}")
def shuffle_unpack(
id: ShuffleId, output_partition: int, barrier: object
) -> pd.DataFrame:
try:
return _get_worker_extension().get_output_partition(id, output_partition)
except Exception:
raise RuntimeError(f"shuffle_unpack failed during shuffle {id}")
def shuffle_barrier(id: ShuffleId, transfers: list[None]) -> None:
try:
return _get_worker_extension().barrier(id)
except Exception:
raise RuntimeError(f"shuffle_barrier failed during shuffle {id}")
def rearrange_by_column_p2p(
df: DataFrame,
column: str,
npartitions: int | None = None,
) -> DataFrame:
from dask.dataframe import DataFrame
npartitions = npartitions or df.npartitions
token = tokenize(df, column, npartitions)
empty = df._meta.copy()
for c, dt in empty.dtypes.items():
if dt == object:
empty[c] = empty[c].astype(
"string"
) # TODO: we fail at non-string object dtypes
empty[column] = empty[column].astype("int64") # TODO: this shouldn't be necesssary
name = f"shuffle-p2p-{token}"
layer = P2PShuffleLayer(
name,
column,
npartitions,
npartitions_input=df.npartitions,
ignore_index=True,
name_input=df._name,
meta_input=empty,
)
return DataFrame(
HighLevelGraph.from_collections(name, layer, [df]),
name,
empty,
[None] * (npartitions + 1),
)
class P2PShuffleLayer(SimpleShuffleLayer):
def __init__(
self,
name: str,
column: str,
npartitions: int,
npartitions_input: int,
ignore_index: bool,
name_input: str,
meta_input: pd.DataFrame,
parts_out: list | None = None,
annotations: dict | None = None,
):
annotations = annotations or {}
annotations.update({"shuffle": lambda key: key[1]})
super().__init__(
name,
column,
npartitions,
npartitions_input,
ignore_index,
name_input,
meta_input,
parts_out,
annotations=annotations,
)
def get_split_keys(self) -> list:
# TODO: This is doing some funky stuff to set priorities but we don't need this
return []
def __repr__(self) -> str:
return (
f"{type(self).__name__}<name='{self.name}', npartitions={self.npartitions}>"
)
def _cull(self, parts_out: list) -> P2PShuffleLayer:
return P2PShuffleLayer(
self.name,
self.column,
self.npartitions,
self.npartitions_input,
self.ignore_index,
self.name_input,
self.meta_input,
parts_out=parts_out,
)
def _construct_graph(self, deserializing: Any = None) -> dict[tuple | str, tuple]:
token = tokenize(self.name_input, self.column, self.npartitions, self.parts_out)
dsk: dict[tuple | str, tuple] = {}
_barrier_key = barrier_key(ShuffleId(token))
name = "shuffle-transfer-" + token
transfer_keys = list()
for i in range(self.npartitions_input):
transfer_keys.append((name, i))
dsk[(name, i)] = (
shuffle_transfer,
(self.name_input, i),
token,
self.npartitions,
self.column,
)
dsk[_barrier_key] = (shuffle_barrier, token, transfer_keys)
name = self.name
for part_out in self.parts_out:
dsk[(name, part_out)] = (shuffle_unpack, token, part_out, _barrier_key)
return dsk
_BARRIER_PREFIX = "shuffle-barrier-"
def barrier_key(shuffle_id: ShuffleId) -> str:
return _BARRIER_PREFIX + shuffle_id
def id_from_key(key: str) -> ShuffleId:
assert key.startswith(_BARRIER_PREFIX)
return ShuffleId(key.replace(_BARRIER_PREFIX, ""))
|
