-
Notifications
You must be signed in to change notification settings - Fork 14.5k
/
dataproc.py
2261 lines (2028 loc) · 97.2 KB
/
dataproc.py
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
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
#
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership. The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License.
#
"""This module contains Google Dataproc operators."""
import inspect
import ntpath
import os
import re
import time
import uuid
import warnings
from datetime import datetime, timedelta
from typing import TYPE_CHECKING, Dict, List, Optional, Sequence, Set, Tuple, Union
from google.api_core import operation # type: ignore
from google.api_core.exceptions import AlreadyExists, NotFound
from google.api_core.gapic_v1.method import DEFAULT, _MethodDefault
from google.api_core.retry import Retry, exponential_sleep_generator
from google.cloud.dataproc_v1 import Batch, Cluster
from google.protobuf.duration_pb2 import Duration
from google.protobuf.field_mask_pb2 import FieldMask
from airflow.exceptions import AirflowException
from airflow.models import BaseOperator
from airflow.providers.google.cloud.hooks.dataproc import DataprocHook, DataProcJobBuilder
from airflow.providers.google.cloud.hooks.gcs import GCSHook
from airflow.providers.google.cloud.links.dataproc import (
DATAPROC_BATCH_LINK,
DATAPROC_BATCHES_LINK,
DATAPROC_CLUSTER_LINK,
DATAPROC_JOB_LOG_LINK,
DATAPROC_WORKFLOW_LINK,
DATAPROC_WORKFLOW_TEMPLATE_LINK,
DataprocLink,
DataprocListLink,
)
from airflow.utils import timezone
if TYPE_CHECKING:
from airflow.utils.context import Context
class ClusterGenerator:
"""
Create a new Dataproc Cluster.
:param cluster_name: The name of the DataProc cluster to create. (templated)
:param project_id: The ID of the google cloud project in which
to create the cluster. (templated)
:param num_workers: The # of workers to spin up. If set to zero will
spin up cluster in a single node mode
:param storage_bucket: The storage bucket to use, setting to None lets dataproc
generate a custom one for you
:param init_actions_uris: List of GCS uri's containing
dataproc initialization scripts
:param init_action_timeout: Amount of time executable scripts in
init_actions_uris has to complete
:param metadata: dict of key-value google compute engine metadata entries
to add to all instances
:param image_version: the version of software inside the Dataproc cluster
:param custom_image: custom Dataproc image for more info see
https://cloud.google.com/dataproc/docs/guides/dataproc-images
:param custom_image_project_id: project id for the custom Dataproc image, for more info see
https://cloud.google.com/dataproc/docs/guides/dataproc-images
:param custom_image_family: family for the custom Dataproc image,
family name can be provide using --family flag while creating custom image, for more info see
https://cloud.google.com/dataproc/docs/guides/dataproc-images
:param autoscaling_policy: The autoscaling policy used by the cluster. Only resource names
including projectid and location (region) are valid. Example:
``projects/[projectId]/locations/[dataproc_region]/autoscalingPolicies/[policy_id]``
:param properties: dict of properties to set on
config files (e.g. spark-defaults.conf), see
https://cloud.google.com/dataproc/docs/reference/rest/v1/projects.regions.clusters#SoftwareConfig
:param optional_components: List of optional cluster components, for more info see
https://cloud.google.com/dataproc/docs/reference/rest/v1/ClusterConfig#Component
:param num_masters: The # of master nodes to spin up
:param master_machine_type: Compute engine machine type to use for the primary node
:param master_disk_type: Type of the boot disk for the primary node
(default is ``pd-standard``).
Valid values: ``pd-ssd`` (Persistent Disk Solid State Drive) or
``pd-standard`` (Persistent Disk Hard Disk Drive).
:param master_disk_size: Disk size for the primary node
:param worker_machine_type: Compute engine machine type to use for the worker nodes
:param worker_disk_type: Type of the boot disk for the worker node
(default is ``pd-standard``).
Valid values: ``pd-ssd`` (Persistent Disk Solid State Drive) or
``pd-standard`` (Persistent Disk Hard Disk Drive).
:param worker_disk_size: Disk size for the worker nodes
:param num_preemptible_workers: The # of preemptible worker nodes to spin up
:param labels: dict of labels to add to the cluster
:param zone: The zone where the cluster will be located. Set to None to auto-zone. (templated)
:param network_uri: The network uri to be used for machine communication, cannot be
specified with subnetwork_uri
:param subnetwork_uri: The subnetwork uri to be used for machine communication,
cannot be specified with network_uri
:param internal_ip_only: If true, all instances in the cluster will only
have internal IP addresses. This can only be enabled for subnetwork
enabled networks
:param tags: The GCE tags to add to all instances
:param region: The specified region where the dataproc cluster is created.
:param gcp_conn_id: The connection ID to use connecting to Google Cloud.
:param service_account: The service account of the dataproc instances.
:param service_account_scopes: The URIs of service account scopes to be included.
:param idle_delete_ttl: The longest duration that cluster would keep alive while
staying idle. Passing this threshold will cause cluster to be auto-deleted.
A duration in seconds.
:param auto_delete_time: The time when cluster will be auto-deleted.
:param auto_delete_ttl: The life duration of cluster, the cluster will be
auto-deleted at the end of this duration.
A duration in seconds. (If auto_delete_time is set this parameter will be ignored)
:param customer_managed_key: The customer-managed key used for disk encryption
``projects/[PROJECT_STORING_KEYS]/locations/[LOCATION]/keyRings/[KEY_RING_NAME]/cryptoKeys/[KEY_NAME]`` # noqa
:param enable_component_gateway: Provides access to the web interfaces of default and selected optional
components on the cluster.
"""
def __init__(
self,
project_id: str,
num_workers: Optional[int] = None,
zone: Optional[str] = None,
network_uri: Optional[str] = None,
subnetwork_uri: Optional[str] = None,
internal_ip_only: Optional[bool] = None,
tags: Optional[List[str]] = None,
storage_bucket: Optional[str] = None,
init_actions_uris: Optional[List[str]] = None,
init_action_timeout: str = "10m",
metadata: Optional[Dict] = None,
custom_image: Optional[str] = None,
custom_image_project_id: Optional[str] = None,
custom_image_family: Optional[str] = None,
image_version: Optional[str] = None,
autoscaling_policy: Optional[str] = None,
properties: Optional[Dict] = None,
optional_components: Optional[List[str]] = None,
num_masters: int = 1,
master_machine_type: str = 'n1-standard-4',
master_disk_type: str = 'pd-standard',
master_disk_size: int = 1024,
worker_machine_type: str = 'n1-standard-4',
worker_disk_type: str = 'pd-standard',
worker_disk_size: int = 1024,
num_preemptible_workers: int = 0,
service_account: Optional[str] = None,
service_account_scopes: Optional[List[str]] = None,
idle_delete_ttl: Optional[int] = None,
auto_delete_time: Optional[datetime] = None,
auto_delete_ttl: Optional[int] = None,
customer_managed_key: Optional[str] = None,
enable_component_gateway: Optional[bool] = False,
**kwargs,
) -> None:
self.project_id = project_id
self.num_masters = num_masters
self.num_workers = num_workers
self.num_preemptible_workers = num_preemptible_workers
self.storage_bucket = storage_bucket
self.init_actions_uris = init_actions_uris
self.init_action_timeout = init_action_timeout
self.metadata = metadata
self.custom_image = custom_image
self.custom_image_project_id = custom_image_project_id
self.custom_image_family = custom_image_family
self.image_version = image_version
self.properties = properties or {}
self.optional_components = optional_components
self.master_machine_type = master_machine_type
self.master_disk_type = master_disk_type
self.master_disk_size = master_disk_size
self.autoscaling_policy = autoscaling_policy
self.worker_machine_type = worker_machine_type
self.worker_disk_type = worker_disk_type
self.worker_disk_size = worker_disk_size
self.zone = zone
self.network_uri = network_uri
self.subnetwork_uri = subnetwork_uri
self.internal_ip_only = internal_ip_only
self.tags = tags
self.service_account = service_account
self.service_account_scopes = service_account_scopes
self.idle_delete_ttl = idle_delete_ttl
self.auto_delete_time = auto_delete_time
self.auto_delete_ttl = auto_delete_ttl
self.customer_managed_key = customer_managed_key
self.enable_component_gateway = enable_component_gateway
self.single_node = num_workers == 0
if self.custom_image and self.image_version:
raise ValueError("The custom_image and image_version can't be both set")
if self.custom_image_family and self.image_version:
raise ValueError("The image_version and custom_image_family can't be both set")
if self.custom_image_family and self.custom_image:
raise ValueError("The custom_image and custom_image_family can't be both set")
if self.single_node and self.num_preemptible_workers > 0:
raise ValueError("Single node cannot have preemptible workers.")
def _get_init_action_timeout(self) -> dict:
match = re.match(r"^(\d+)([sm])$", self.init_action_timeout)
if match:
val = float(match.group(1))
if match.group(2) == "s":
return {"seconds": int(val)}
elif match.group(2) == "m":
return {"seconds": int(timedelta(minutes=val).total_seconds())}
raise AirflowException(
"DataprocClusterCreateOperator init_action_timeout"
" should be expressed in minutes or seconds. i.e. 10m, 30s"
)
def _build_gce_cluster_config(self, cluster_data):
if self.zone:
zone_uri = f'https://www.googleapis.com/compute/v1/projects/{self.project_id}/zones/{self.zone}'
cluster_data['gce_cluster_config']['zone_uri'] = zone_uri
if self.metadata:
cluster_data['gce_cluster_config']['metadata'] = self.metadata
if self.network_uri:
cluster_data['gce_cluster_config']['network_uri'] = self.network_uri
if self.subnetwork_uri:
cluster_data['gce_cluster_config']['subnetwork_uri'] = self.subnetwork_uri
if self.internal_ip_only:
if not self.subnetwork_uri:
raise AirflowException("Set internal_ip_only to true only when you pass a subnetwork_uri.")
cluster_data['gce_cluster_config']['internal_ip_only'] = True
if self.tags:
cluster_data['gce_cluster_config']['tags'] = self.tags
if self.service_account:
cluster_data['gce_cluster_config']['service_account'] = self.service_account
if self.service_account_scopes:
cluster_data['gce_cluster_config']['service_account_scopes'] = self.service_account_scopes
return cluster_data
def _build_lifecycle_config(self, cluster_data):
if self.idle_delete_ttl:
cluster_data['lifecycle_config']['idle_delete_ttl'] = {"seconds": self.idle_delete_ttl}
if self.auto_delete_time:
utc_auto_delete_time = timezone.convert_to_utc(self.auto_delete_time)
cluster_data['lifecycle_config']['auto_delete_time'] = utc_auto_delete_time.strftime(
'%Y-%m-%dT%H:%M:%S.%fZ'
)
elif self.auto_delete_ttl:
cluster_data['lifecycle_config']['auto_delete_ttl'] = {"seconds": int(self.auto_delete_ttl)}
return cluster_data
def _build_cluster_data(self):
if self.zone:
master_type_uri = (
f"projects/{self.project_id}/zones/{self.zone}/machineTypes/{self.master_machine_type}"
)
worker_type_uri = (
f"projects/{self.project_id}/zones/{self.zone}/machineTypes/{self.worker_machine_type}"
)
else:
master_type_uri = self.master_machine_type
worker_type_uri = self.worker_machine_type
cluster_data = {
'gce_cluster_config': {},
'master_config': {
'num_instances': self.num_masters,
'machine_type_uri': master_type_uri,
'disk_config': {
'boot_disk_type': self.master_disk_type,
'boot_disk_size_gb': self.master_disk_size,
},
},
'worker_config': {
'num_instances': self.num_workers,
'machine_type_uri': worker_type_uri,
'disk_config': {
'boot_disk_type': self.worker_disk_type,
'boot_disk_size_gb': self.worker_disk_size,
},
},
'secondary_worker_config': {},
'software_config': {},
'lifecycle_config': {},
'encryption_config': {},
'autoscaling_config': {},
'endpoint_config': {},
}
if self.num_preemptible_workers > 0:
cluster_data['secondary_worker_config'] = {
'num_instances': self.num_preemptible_workers,
'machine_type_uri': worker_type_uri,
'disk_config': {
'boot_disk_type': self.worker_disk_type,
'boot_disk_size_gb': self.worker_disk_size,
},
'is_preemptible': True,
}
if self.storage_bucket:
cluster_data['config_bucket'] = self.storage_bucket
if self.image_version:
cluster_data['software_config']['image_version'] = self.image_version
elif self.custom_image:
project_id = self.custom_image_project_id or self.project_id
custom_image_url = (
f'https://www.googleapis.com/compute/beta/projects/{project_id}'
f'/global/images/{self.custom_image}'
)
cluster_data['master_config']['image_uri'] = custom_image_url
if not self.single_node:
cluster_data['worker_config']['image_uri'] = custom_image_url
elif self.custom_image_family:
project_id = self.custom_image_project_id or self.project_id
custom_image_url = (
'https://www.googleapis.com/compute/beta/projects/'
f'{project_id}/global/images/family/{self.custom_image_family}'
)
cluster_data['master_config']['image_uri'] = custom_image_url
if not self.single_node:
cluster_data['worker_config']['image_uri'] = custom_image_url
cluster_data = self._build_gce_cluster_config(cluster_data)
if self.single_node:
self.properties["dataproc:dataproc.allow.zero.workers"] = "true"
if self.properties:
cluster_data['software_config']['properties'] = self.properties
if self.optional_components:
cluster_data['software_config']['optional_components'] = self.optional_components
cluster_data = self._build_lifecycle_config(cluster_data)
if self.init_actions_uris:
init_actions_dict = [
{'executable_file': uri, 'execution_timeout': self._get_init_action_timeout()}
for uri in self.init_actions_uris
]
cluster_data['initialization_actions'] = init_actions_dict
if self.customer_managed_key:
cluster_data['encryption_config'] = {'gce_pd_kms_key_name': self.customer_managed_key}
if self.autoscaling_policy:
cluster_data['autoscaling_config'] = {'policy_uri': self.autoscaling_policy}
if self.enable_component_gateway:
cluster_data['endpoint_config'] = {'enable_http_port_access': self.enable_component_gateway}
return cluster_data
def make(self):
"""
Helper method for easier migration.
:return: Dict representing Dataproc cluster.
"""
return self._build_cluster_data()
class DataprocCreateClusterOperator(BaseOperator):
"""
Create a new cluster on Google Cloud Dataproc. The operator will wait until the
creation is successful or an error occurs in the creation process. If the cluster
already exists and ``use_if_exists`` is True then the operator will:
- if cluster state is ERROR then delete it if specified and raise error
- if cluster state is CREATING wait for it and then check for ERROR state
- if cluster state is DELETING wait for it and then create new cluster
Please refer to
https://cloud.google.com/dataproc/docs/reference/rest/v1/projects.regions.clusters
for a detailed explanation on the different parameters. Most of the configuration
parameters detailed in the link are available as a parameter to this operator.
.. seealso::
For more information on how to use this operator, take a look at the guide:
:ref:`howto/operator:DataprocCreateClusterOperator`
:param project_id: The ID of the google cloud project in which
to create the cluster. (templated)
:param cluster_name: Name of the cluster to create
:param labels: Labels that will be assigned to created cluster
:param cluster_config: Required. The cluster config to create.
If a dict is provided, it must be of the same form as the protobuf message
:class:`~google.cloud.dataproc_v1.types.ClusterConfig`
:param virtual_cluster_config: Optional. The virtual cluster config, used when creating a Dataproc
cluster that does not directly control the underlying compute resources, for example, when creating a
`Dataproc-on-GKE cluster
<https://cloud.google.com/dataproc/docs/concepts/jobs/dataproc-gke#create-a-dataproc-on-gke-cluster>`
:param region: The specified region where the dataproc cluster is created.
:param delete_on_error: If true the cluster will be deleted if created with ERROR state. Default
value is true.
:param use_if_exists: If true use existing cluster
:param request_id: Optional. A unique id used to identify the request. If the server receives two
``DeleteClusterRequest`` requests with the same id, then the second request will be ignored and the
first ``google.longrunning.Operation`` created and stored in the backend is returned.
:param retry: A retry object used to retry requests. If ``None`` is specified, requests will not be
retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete. Note that if
``retry`` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:param gcp_conn_id: The connection ID to use connecting to Google Cloud.
:param impersonation_chain: Optional service account to impersonate using short-term
credentials, or chained list of accounts required to get the access_token
of the last account in the list, which will be impersonated in the request.
If set as a string, the account must grant the originating account
the Service Account Token Creator IAM role.
If set as a sequence, the identities from the list must grant
Service Account Token Creator IAM role to the directly preceding identity, with first
account from the list granting this role to the originating account (templated).
"""
template_fields: Sequence[str] = (
'project_id',
'region',
'cluster_config',
'virtual_cluster_config',
'cluster_name',
'labels',
'impersonation_chain',
)
template_fields_renderers = {'cluster_config': 'json', 'virtual_cluster_config': 'json'}
operator_extra_links = (DataprocLink(),)
def __init__(
self,
*,
cluster_name: str,
region: str,
project_id: Optional[str] = None,
cluster_config: Optional[Union[Dict, Cluster]] = None,
virtual_cluster_config: Optional[Dict] = None,
labels: Optional[Dict] = None,
request_id: Optional[str] = None,
delete_on_error: bool = True,
use_if_exists: bool = True,
retry: Union[Retry, _MethodDefault] = DEFAULT,
timeout: float = 1 * 60 * 60,
metadata: Sequence[Tuple[str, str]] = (),
gcp_conn_id: str = "google_cloud_default",
impersonation_chain: Optional[Union[str, Sequence[str]]] = None,
**kwargs,
) -> None:
# TODO: remove one day
if cluster_config is None and virtual_cluster_config is None:
warnings.warn(
f"Passing cluster parameters by keywords to `{type(self).__name__}` will be deprecated. "
"Please provide cluster_config object using `cluster_config` parameter. "
"You can use `airflow.dataproc.ClusterGenerator.generate_cluster` "
"method to obtain cluster object.",
DeprecationWarning,
stacklevel=1,
)
# Remove result of apply defaults
if 'params' in kwargs:
del kwargs['params']
# Create cluster object from kwargs
if project_id is None:
raise AirflowException(
"project_id argument is required when building cluster from keywords parameters"
)
kwargs["project_id"] = project_id
cluster_config = ClusterGenerator(**kwargs).make()
# Remove from kwargs cluster params passed for backward compatibility
cluster_params = inspect.signature(ClusterGenerator.__init__).parameters
for arg in cluster_params:
if arg in kwargs:
del kwargs[arg]
super().__init__(**kwargs)
self.cluster_config = cluster_config
self.cluster_name = cluster_name
self.labels = labels
self.project_id = project_id
self.region = region
self.request_id = request_id
self.retry = retry
self.timeout = timeout
self.metadata = metadata
self.gcp_conn_id = gcp_conn_id
self.delete_on_error = delete_on_error
self.use_if_exists = use_if_exists
self.impersonation_chain = impersonation_chain
self.virtual_cluster_config = virtual_cluster_config
def _create_cluster(self, hook: DataprocHook):
operation = hook.create_cluster(
project_id=self.project_id,
region=self.region,
cluster_name=self.cluster_name,
labels=self.labels,
cluster_config=self.cluster_config,
virtual_cluster_config=self.virtual_cluster_config,
request_id=self.request_id,
retry=self.retry,
timeout=self.timeout,
metadata=self.metadata,
)
cluster = operation.result()
self.log.info("Cluster created.")
return cluster
def _delete_cluster(self, hook):
self.log.info("Deleting the cluster")
hook.delete_cluster(region=self.region, cluster_name=self.cluster_name, project_id=self.project_id)
def _get_cluster(self, hook: DataprocHook) -> Cluster:
return hook.get_cluster(
project_id=self.project_id,
region=self.region,
cluster_name=self.cluster_name,
retry=self.retry,
timeout=self.timeout,
metadata=self.metadata,
)
def _handle_error_state(self, hook: DataprocHook, cluster: Cluster) -> None:
if cluster.status.state != cluster.status.State.ERROR:
return
self.log.info("Cluster is in ERROR state")
gcs_uri = hook.diagnose_cluster(
region=self.region, cluster_name=self.cluster_name, project_id=self.project_id
)
self.log.info('Diagnostic information for cluster %s available at: %s', self.cluster_name, gcs_uri)
if self.delete_on_error:
self._delete_cluster(hook)
raise AirflowException("Cluster was created but was in ERROR state.")
raise AirflowException("Cluster was created but is in ERROR state")
def _wait_for_cluster_in_deleting_state(self, hook: DataprocHook) -> None:
time_left = self.timeout
for time_to_sleep in exponential_sleep_generator(initial=10, maximum=120):
if time_left < 0:
raise AirflowException(f"Cluster {self.cluster_name} is still DELETING state, aborting")
time.sleep(time_to_sleep)
time_left = time_left - time_to_sleep
try:
self._get_cluster(hook)
except NotFound:
break
def _wait_for_cluster_in_creating_state(self, hook: DataprocHook) -> Cluster:
time_left = self.timeout
cluster = self._get_cluster(hook)
for time_to_sleep in exponential_sleep_generator(initial=10, maximum=120):
if cluster.status.state != cluster.status.State.CREATING:
break
if time_left < 0:
raise AirflowException(f"Cluster {self.cluster_name} is still CREATING state, aborting")
time.sleep(time_to_sleep)
time_left = time_left - time_to_sleep
cluster = self._get_cluster(hook)
return cluster
def execute(self, context: 'Context') -> dict:
self.log.info('Creating cluster: %s', self.cluster_name)
hook = DataprocHook(gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain)
# Save data required to display extra link no matter what the cluster status will be
DataprocLink.persist(
context=context, task_instance=self, url=DATAPROC_CLUSTER_LINK, resource=self.cluster_name
)
try:
# First try to create a new cluster
cluster = self._create_cluster(hook)
except AlreadyExists:
if not self.use_if_exists:
raise
self.log.info("Cluster already exists.")
cluster = self._get_cluster(hook)
# Check if cluster is not in ERROR state
self._handle_error_state(hook, cluster)
if cluster.status.state == cluster.status.State.CREATING:
# Wait for cluster to be created
cluster = self._wait_for_cluster_in_creating_state(hook)
self._handle_error_state(hook, cluster)
elif cluster.status.state == cluster.status.State.DELETING:
# Wait for cluster to be deleted
self._wait_for_cluster_in_deleting_state(hook)
# Create new cluster
cluster = self._create_cluster(hook)
self._handle_error_state(hook, cluster)
return Cluster.to_dict(cluster)
class DataprocScaleClusterOperator(BaseOperator):
"""
Scale, up or down, a cluster on Google Cloud Dataproc.
The operator will wait until the cluster is re-scaled.
**Example**: ::
t1 = DataprocClusterScaleOperator(
task_id='dataproc_scale',
project_id='my-project',
cluster_name='cluster-1',
num_workers=10,
num_preemptible_workers=10,
graceful_decommission_timeout='1h',
dag=dag)
.. seealso::
For more detail on about scaling clusters have a look at the reference:
https://cloud.google.com/dataproc/docs/concepts/configuring-clusters/scaling-clusters
:param cluster_name: The name of the cluster to scale. (templated)
:param project_id: The ID of the google cloud project in which
the cluster runs. (templated)
:param region: The region for the dataproc cluster. (templated)
:param num_workers: The new number of workers
:param num_preemptible_workers: The new number of preemptible workers
:param graceful_decommission_timeout: Timeout for graceful YARN decommissioning.
Maximum value is 1d
:param gcp_conn_id: The connection ID to use connecting to Google Cloud.
:param impersonation_chain: Optional service account to impersonate using short-term
credentials, or chained list of accounts required to get the access_token
of the last account in the list, which will be impersonated in the request.
If set as a string, the account must grant the originating account
the Service Account Token Creator IAM role.
If set as a sequence, the identities from the list must grant
Service Account Token Creator IAM role to the directly preceding identity, with first
account from the list granting this role to the originating account (templated).
"""
template_fields: Sequence[str] = ('cluster_name', 'project_id', 'region', 'impersonation_chain')
operator_extra_links = (DataprocLink(),)
def __init__(
self,
*,
cluster_name: str,
project_id: Optional[str] = None,
region: str = 'global',
num_workers: int = 2,
num_preemptible_workers: int = 0,
graceful_decommission_timeout: Optional[str] = None,
gcp_conn_id: str = "google_cloud_default",
impersonation_chain: Optional[Union[str, Sequence[str]]] = None,
**kwargs,
) -> None:
super().__init__(**kwargs)
self.project_id = project_id
self.region = region
self.cluster_name = cluster_name
self.num_workers = num_workers
self.num_preemptible_workers = num_preemptible_workers
self.graceful_decommission_timeout = graceful_decommission_timeout
self.gcp_conn_id = gcp_conn_id
self.impersonation_chain = impersonation_chain
# TODO: Remove one day
warnings.warn(
f"The `{type(self).__name__}` operator is deprecated, "
"please use `DataprocUpdateClusterOperator` instead.",
DeprecationWarning,
stacklevel=1,
)
def _build_scale_cluster_data(self) -> dict:
scale_data = {
'config': {
'worker_config': {'num_instances': self.num_workers},
'secondary_worker_config': {'num_instances': self.num_preemptible_workers},
}
}
return scale_data
@property
def _graceful_decommission_timeout_object(self) -> Optional[Dict[str, int]]:
if not self.graceful_decommission_timeout:
return None
timeout = None
match = re.match(r"^(\d+)([smdh])$", self.graceful_decommission_timeout)
if match:
if match.group(2) == "s":
timeout = int(match.group(1))
elif match.group(2) == "m":
val = float(match.group(1))
timeout = int(timedelta(minutes=val).total_seconds())
elif match.group(2) == "h":
val = float(match.group(1))
timeout = int(timedelta(hours=val).total_seconds())
elif match.group(2) == "d":
val = float(match.group(1))
timeout = int(timedelta(days=val).total_seconds())
if not timeout:
raise AirflowException(
"DataprocClusterScaleOperator "
" should be expressed in day, hours, minutes or seconds. "
" i.e. 1d, 4h, 10m, 30s"
)
return {'seconds': timeout}
def execute(self, context: 'Context') -> None:
"""Scale, up or down, a cluster on Google Cloud Dataproc."""
self.log.info("Scaling cluster: %s", self.cluster_name)
scaling_cluster_data = self._build_scale_cluster_data()
update_mask = ["config.worker_config.num_instances", "config.secondary_worker_config.num_instances"]
hook = DataprocHook(gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain)
# Save data required to display extra link no matter what the cluster status will be
DataprocLink.persist(
context=context, task_instance=self, url=DATAPROC_CLUSTER_LINK, resource=self.cluster_name
)
operation = hook.update_cluster(
project_id=self.project_id,
region=self.region,
cluster_name=self.cluster_name,
cluster=scaling_cluster_data,
graceful_decommission_timeout=self._graceful_decommission_timeout_object,
update_mask={'paths': update_mask},
)
operation.result()
self.log.info("Cluster scaling finished")
class DataprocDeleteClusterOperator(BaseOperator):
"""
Deletes a cluster in a project.
:param region: Required. The Cloud Dataproc region in which to handle the request (templated).
:param cluster_name: Required. The cluster name (templated).
:param project_id: Optional. The ID of the Google Cloud project that the cluster belongs to (templated).
:param cluster_uuid: Optional. Specifying the ``cluster_uuid`` means the RPC should fail
if cluster with specified UUID does not exist.
:param request_id: Optional. A unique id used to identify the request. If the server receives two
``DeleteClusterRequest`` requests with the same id, then the second request will be ignored and the
first ``google.longrunning.Operation`` created and stored in the backend is returned.
:param retry: A retry object used to retry requests. If ``None`` is specified, requests will not be
retried.
:param timeout: The amount of time, in seconds, to wait for the request to complete. Note that if
``retry`` is specified, the timeout applies to each individual attempt.
:param metadata: Additional metadata that is provided to the method.
:param gcp_conn_id: The connection ID to use connecting to Google Cloud.
:param impersonation_chain: Optional service account to impersonate using short-term
credentials, or chained list of accounts required to get the access_token
of the last account in the list, which will be impersonated in the request.
If set as a string, the account must grant the originating account
the Service Account Token Creator IAM role.
If set as a sequence, the identities from the list must grant
Service Account Token Creator IAM role to the directly preceding identity, with first
account from the list granting this role to the originating account (templated).
"""
template_fields: Sequence[str] = ('project_id', 'region', 'cluster_name', 'impersonation_chain')
def __init__(
self,
*,
region: str,
cluster_name: str,
project_id: Optional[str] = None,
cluster_uuid: Optional[str] = None,
request_id: Optional[str] = None,
retry: Union[Retry, _MethodDefault] = DEFAULT,
timeout: Optional[float] = None,
metadata: Sequence[Tuple[str, str]] = (),
gcp_conn_id: str = "google_cloud_default",
impersonation_chain: Optional[Union[str, Sequence[str]]] = None,
**kwargs,
):
super().__init__(**kwargs)
self.project_id = project_id
self.region = region
self.cluster_name = cluster_name
self.cluster_uuid = cluster_uuid
self.request_id = request_id
self.retry = retry
self.timeout = timeout
self.metadata = metadata
self.gcp_conn_id = gcp_conn_id
self.impersonation_chain = impersonation_chain
def execute(self, context: 'Context') -> None:
hook = DataprocHook(gcp_conn_id=self.gcp_conn_id, impersonation_chain=self.impersonation_chain)
self.log.info("Deleting cluster: %s", self.cluster_name)
operation = hook.delete_cluster(
project_id=self.project_id,
region=self.region,
cluster_name=self.cluster_name,
cluster_uuid=self.cluster_uuid,
request_id=self.request_id,
retry=self.retry,
timeout=self.timeout,
metadata=self.metadata,
)
operation.result()
self.log.info("Cluster deleted.")
class DataprocJobBaseOperator(BaseOperator):
"""
The base class for operators that launch job on DataProc.
:param region: The specified region where the dataproc cluster is created.
:param job_name: The job name used in the DataProc cluster. This name by default
is the task_id appended with the execution data, but can be templated. The
name will always be appended with a random number to avoid name clashes.
:param cluster_name: The name of the DataProc cluster.
:param project_id: The ID of the Google Cloud project the cluster belongs to,
if not specified the project will be inferred from the provided GCP connection.
:param dataproc_properties: Map for the Hive properties. Ideal to put in
default arguments (templated)
:param dataproc_jars: HCFS URIs of jar files to add to the CLASSPATH of the Hive server and Hadoop
MapReduce (MR) tasks. Can contain Hive SerDes and UDFs. (templated)
:param gcp_conn_id: The connection ID to use connecting to Google Cloud.
:param delegate_to: The account to impersonate using domain-wide delegation of authority,
if any. For this to work, the service account making the request must have
domain-wide delegation enabled.
:param labels: The labels to associate with this job. Label keys must contain 1 to 63 characters,
and must conform to RFC 1035. Label values may be empty, but, if present, must contain 1 to 63
characters, and must conform to RFC 1035. No more than 32 labels can be associated with a job.
:param job_error_states: Job states that should be considered error states.
Any states in this set will result in an error being raised and failure of the
task. Eg, if the ``CANCELLED`` state should also be considered a task failure,
pass in ``{'ERROR', 'CANCELLED'}``. Possible values are currently only
``'ERROR'`` and ``'CANCELLED'``, but could change in the future. Defaults to
``{'ERROR'}``.
:param impersonation_chain: Optional service account to impersonate using short-term
credentials, or chained list of accounts required to get the access_token
of the last account in the list, which will be impersonated in the request.
If set as a string, the account must grant the originating account
the Service Account Token Creator IAM role.
If set as a sequence, the identities from the list must grant
Service Account Token Creator IAM role to the directly preceding identity, with first
account from the list granting this role to the originating account (templated).
:param asynchronous: Flag to return after submitting the job to the Dataproc API.
This is useful for submitting long running jobs and
waiting on them asynchronously using the DataprocJobSensor
:var dataproc_job_id: The actual "jobId" as submitted to the Dataproc API.
This is useful for identifying or linking to the job in the Google Cloud Console
Dataproc UI, as the actual "jobId" submitted to the Dataproc API is appended with
an 8 character random string.
:vartype dataproc_job_id: str
"""
job_type = ""
operator_extra_links = (DataprocLink(),)
def __init__(
self,
*,
region: str,
job_name: str = '{{task.task_id}}_{{ds_nodash}}',
cluster_name: str = "cluster-1",
project_id: Optional[str] = None,
dataproc_properties: Optional[Dict] = None,
dataproc_jars: Optional[List[str]] = None,
gcp_conn_id: str = 'google_cloud_default',
delegate_to: Optional[str] = None,
labels: Optional[Dict] = None,
job_error_states: Optional[Set[str]] = None,
impersonation_chain: Optional[Union[str, Sequence[str]]] = None,
asynchronous: bool = False,
**kwargs,
) -> None:
super().__init__(**kwargs)
self.gcp_conn_id = gcp_conn_id
self.delegate_to = delegate_to
self.labels = labels
self.job_name = job_name
self.cluster_name = cluster_name
self.dataproc_properties = dataproc_properties
self.dataproc_jars = dataproc_jars
self.region = region
self.job_error_states = job_error_states if job_error_states is not None else {'ERROR'}
self.impersonation_chain = impersonation_chain
self.hook = DataprocHook(gcp_conn_id=gcp_conn_id, impersonation_chain=impersonation_chain)
self.project_id = self.hook.project_id if project_id is None else project_id
self.job_template: Optional[DataProcJobBuilder] = None
self.job: Optional[dict] = None
self.dataproc_job_id = None
self.asynchronous = asynchronous
def create_job_template(self) -> DataProcJobBuilder:
"""Initialize `self.job_template` with default values"""
if self.project_id is None:
raise AirflowException(
"project id should either be set via project_id "
"parameter or retrieved from the connection,"
)
job_template = DataProcJobBuilder(
project_id=self.project_id,
task_id=self.task_id,
cluster_name=self.cluster_name,
job_type=self.job_type,
properties=self.dataproc_properties,
)
job_template.set_job_name(self.job_name)
job_template.add_jar_file_uris(self.dataproc_jars)
job_template.add_labels(self.labels)
self.job_template = job_template
return job_template
def _generate_job_template(self) -> str:
if self.job_template:
job = self.job_template.build()
return job['job']
raise Exception("Create a job template before")
def execute(self, context: 'Context'):
if self.job_template:
self.job = self.job_template.build()
if self.job is None:
raise Exception("The job should be set here.")
self.dataproc_job_id = self.job["job"]["reference"]["job_id"]
self.log.info('Submitting %s job %s', self.job_type, self.dataproc_job_id)
job_object = self.hook.submit_job(
project_id=self.project_id, job=self.job["job"], region=self.region
)
job_id = job_object.reference.job_id
self.log.info('Job %s submitted successfully.', job_id)
# Save data required for extra links no matter what the job status will be
DataprocLink.persist(
context=context, task_instance=self, url=DATAPROC_JOB_LOG_LINK, resource=job_id
)
if not self.asynchronous:
self.log.info('Waiting for job %s to complete', job_id)
self.hook.wait_for_job(job_id=job_id, region=self.region, project_id=self.project_id)
self.log.info('Job %s completed successfully.', job_id)
return job_id
else:
raise AirflowException("Create a job template before")
def on_kill(self) -> None:
"""
Callback called when the operator is killed.
Cancel any running job.
"""
if self.dataproc_job_id:
self.hook.cancel_job(project_id=self.project_id, job_id=self.dataproc_job_id, region=self.region)
class DataprocSubmitPigJobOperator(DataprocJobBaseOperator):
"""
Start a Pig query Job on a Cloud DataProc cluster. The parameters of the operation
will be passed to the cluster.
It's a good practice to define dataproc_* parameters in the default_args of the dag
like the cluster name and UDFs.
.. code-block:: python
default_args = {
"cluster_name": "cluster-1",
"dataproc_pig_jars": [
"gs://example/udf/jar/datafu/1.2.0/datafu.jar",
"gs://example/udf/jar/gpig/1.2/gpig.jar",
],
}
You can pass a pig script as string or file reference. Use variables to pass on
variables for the pig script to be resolved on the cluster or use the parameters to
be resolved in the script as template parameters.
**Example**: ::