acryl-datahub-cloud 0.3.12rc3__py3-none-any.whl → 0.3.12rc4__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Potentially problematic release.
This version of acryl-datahub-cloud might be problematic. Click here for more details.
- acryl_datahub_cloud/_codegen_config.json +1 -1
- acryl_datahub_cloud/datahub_forms_notifications/forms_notifications_source.py +37 -2
- acryl_datahub_cloud/metadata/schema.avsc +9 -0
- acryl_datahub_cloud/metadata/schemas/AssertionAnalyticsRunEvent.avsc +9 -0
- acryl_datahub_cloud/sdk/__init__.py +5 -1
- acryl_datahub_cloud/sdk/assertion.py +313 -162
- acryl_datahub_cloud/sdk/assertion_input/__init__.py +0 -0
- acryl_datahub_cloud/sdk/{assertion_input.py → assertion_input/assertion_input.py} +430 -147
- acryl_datahub_cloud/sdk/assertion_input/freshness_assertion_input.py +261 -0
- acryl_datahub_cloud/sdk/assertion_input/smart_column_metric_assertion_input.py +947 -0
- acryl_datahub_cloud/sdk/assertions_client.py +493 -7
- acryl_datahub_cloud/sdk/entities/assertion.py +4 -1
- {acryl_datahub_cloud-0.3.12rc3.dist-info → acryl_datahub_cloud-0.3.12rc4.dist-info}/METADATA +38 -38
- {acryl_datahub_cloud-0.3.12rc3.dist-info → acryl_datahub_cloud-0.3.12rc4.dist-info}/RECORD +17 -14
- {acryl_datahub_cloud-0.3.12rc3.dist-info → acryl_datahub_cloud-0.3.12rc4.dist-info}/WHEEL +0 -0
- {acryl_datahub_cloud-0.3.12rc3.dist-info → acryl_datahub_cloud-0.3.12rc4.dist-info}/entry_points.txt +0 -0
- {acryl_datahub_cloud-0.3.12rc3.dist-info → acryl_datahub_cloud-0.3.12rc4.dist-info}/top_level.txt +0 -0
|
@@ -6,19 +6,24 @@ from typing import TYPE_CHECKING, Any, Optional, Union
|
|
|
6
6
|
|
|
7
7
|
from acryl_datahub_cloud.sdk.assertion import (
|
|
8
8
|
AssertionMode,
|
|
9
|
+
FreshnessAssertion,
|
|
9
10
|
SmartFreshnessAssertion,
|
|
10
11
|
SmartVolumeAssertion,
|
|
11
12
|
_AssertionPublic,
|
|
12
13
|
)
|
|
13
|
-
from acryl_datahub_cloud.sdk.assertion_input import (
|
|
14
|
+
from acryl_datahub_cloud.sdk.assertion_input.assertion_input import (
|
|
14
15
|
AssertionIncidentBehavior,
|
|
15
16
|
DetectionMechanismInputTypes,
|
|
16
17
|
ExclusionWindowInputTypes,
|
|
17
18
|
InferenceSensitivity,
|
|
19
|
+
TimeWindowSizeInputTypes,
|
|
18
20
|
_AssertionInput,
|
|
19
21
|
_SmartFreshnessAssertionInput,
|
|
20
22
|
_SmartVolumeAssertionInput,
|
|
21
23
|
)
|
|
24
|
+
from acryl_datahub_cloud.sdk.assertion_input.freshness_assertion_input import (
|
|
25
|
+
_FreshnessAssertionInput,
|
|
26
|
+
)
|
|
22
27
|
from acryl_datahub_cloud.sdk.entities.assertion import Assertion, TagsInputType
|
|
23
28
|
from acryl_datahub_cloud.sdk.entities.monitor import Monitor
|
|
24
29
|
from acryl_datahub_cloud.sdk.errors import SDKUsageError
|
|
@@ -171,7 +176,7 @@ class AssertionsClient:
|
|
|
171
176
|
# 3. Merge the assertion input with the existing assertion and monitor entities or create a new assertion
|
|
172
177
|
# if the assertion does not exist:
|
|
173
178
|
merged_assertion_input_or_created_assertion = (
|
|
174
|
-
self.
|
|
179
|
+
self._retrieve_and_merge_smart_freshness_assertion_and_monitor(
|
|
175
180
|
assertion_input=assertion_input,
|
|
176
181
|
dataset_urn=dataset_urn,
|
|
177
182
|
urn=urn,
|
|
@@ -212,7 +217,7 @@ class AssertionsClient:
|
|
|
212
217
|
|
|
213
218
|
return SmartFreshnessAssertion._from_entities(assertion_entity, monitor_entity)
|
|
214
219
|
|
|
215
|
-
def
|
|
220
|
+
def _retrieve_and_merge_smart_freshness_assertion_and_monitor(
|
|
216
221
|
self,
|
|
217
222
|
assertion_input: _SmartFreshnessAssertionInput,
|
|
218
223
|
dataset_urn: Union[str, DatasetUrn],
|
|
@@ -277,7 +282,7 @@ class AssertionsClient:
|
|
|
277
282
|
)
|
|
278
283
|
|
|
279
284
|
# 4. Merge the existing assertion with the validated input:
|
|
280
|
-
merged_assertion_input = self.
|
|
285
|
+
merged_assertion_input = self._merge_smart_freshness_input(
|
|
281
286
|
dataset_urn=dataset_urn,
|
|
282
287
|
urn=urn,
|
|
283
288
|
display_name=display_name,
|
|
@@ -350,6 +355,7 @@ class AssertionsClient:
|
|
|
350
355
|
incident_behavior=incident_behavior,
|
|
351
356
|
tags=tags,
|
|
352
357
|
created_by=updated_by,
|
|
358
|
+
schedule=schedule,
|
|
353
359
|
)
|
|
354
360
|
|
|
355
361
|
# 3. Check for any issues e.g. different dataset urns
|
|
@@ -363,7 +369,7 @@ class AssertionsClient:
|
|
|
363
369
|
)
|
|
364
370
|
|
|
365
371
|
# 4. Merge the existing assertion with the validated input:
|
|
366
|
-
merged_assertion_input = self.
|
|
372
|
+
merged_assertion_input = self._merge_smart_volume_input(
|
|
367
373
|
dataset_urn=dataset_urn,
|
|
368
374
|
urn=urn,
|
|
369
375
|
display_name=display_name,
|
|
@@ -384,6 +390,93 @@ class AssertionsClient:
|
|
|
384
390
|
|
|
385
391
|
return merged_assertion_input
|
|
386
392
|
|
|
393
|
+
def _retrieve_and_merge_freshness_assertion_and_monitor(
|
|
394
|
+
self,
|
|
395
|
+
assertion_input: _FreshnessAssertionInput,
|
|
396
|
+
dataset_urn: Union[str, DatasetUrn],
|
|
397
|
+
urn: Union[str, AssertionUrn],
|
|
398
|
+
display_name: Optional[str],
|
|
399
|
+
enabled: Optional[bool],
|
|
400
|
+
detection_mechanism: DetectionMechanismInputTypes,
|
|
401
|
+
incident_behavior: Optional[
|
|
402
|
+
Union[AssertionIncidentBehavior, list[AssertionIncidentBehavior]]
|
|
403
|
+
],
|
|
404
|
+
tags: Optional[TagsInputType],
|
|
405
|
+
updated_by: Optional[Union[str, CorpUserUrn]],
|
|
406
|
+
now_utc: datetime,
|
|
407
|
+
schedule: Optional[Union[str, models.CronScheduleClass]],
|
|
408
|
+
freshness_schedule_check_type: Optional[
|
|
409
|
+
Union[str, models.FreshnessAssertionScheduleTypeClass]
|
|
410
|
+
] = None,
|
|
411
|
+
lookback_window: Optional[TimeWindowSizeInputTypes] = None,
|
|
412
|
+
) -> Union[FreshnessAssertion, _FreshnessAssertionInput]:
|
|
413
|
+
# 1. Retrieve any existing assertion and monitor entities:
|
|
414
|
+
maybe_assertion_entity, monitor_urn, maybe_monitor_entity = (
|
|
415
|
+
self._retrieve_assertion_and_monitor(assertion_input)
|
|
416
|
+
)
|
|
417
|
+
|
|
418
|
+
# 2.1 If the assertion and monitor entities exist, create an assertion object from them:
|
|
419
|
+
if maybe_assertion_entity and maybe_monitor_entity:
|
|
420
|
+
existing_assertion = FreshnessAssertion._from_entities(
|
|
421
|
+
maybe_assertion_entity, maybe_monitor_entity
|
|
422
|
+
)
|
|
423
|
+
# 2.2 If the assertion exists but the monitor does not, create a placeholder monitor entity to be able to create the assertion:
|
|
424
|
+
elif maybe_assertion_entity and not maybe_monitor_entity:
|
|
425
|
+
monitor_mode = (
|
|
426
|
+
"ACTIVE" if enabled else "INACTIVE" if enabled is not None else "ACTIVE"
|
|
427
|
+
)
|
|
428
|
+
existing_assertion = FreshnessAssertion._from_entities(
|
|
429
|
+
maybe_assertion_entity,
|
|
430
|
+
Monitor(id=monitor_urn, info=("ASSERTION", monitor_mode)),
|
|
431
|
+
)
|
|
432
|
+
# 2.3 If the assertion does not exist, create a new assertion with a generated urn and return the assertion input:
|
|
433
|
+
elif not maybe_assertion_entity:
|
|
434
|
+
logger.info(
|
|
435
|
+
f"No existing assertion entity found for assertion urn {urn}, creating a new assertion with a generated urn"
|
|
436
|
+
)
|
|
437
|
+
return self._create_freshness_assertion(
|
|
438
|
+
dataset_urn=dataset_urn,
|
|
439
|
+
display_name=display_name,
|
|
440
|
+
detection_mechanism=detection_mechanism,
|
|
441
|
+
incident_behavior=incident_behavior,
|
|
442
|
+
tags=tags,
|
|
443
|
+
created_by=updated_by,
|
|
444
|
+
schedule=schedule,
|
|
445
|
+
freshness_schedule_check_type=freshness_schedule_check_type,
|
|
446
|
+
lookback_window=lookback_window,
|
|
447
|
+
)
|
|
448
|
+
|
|
449
|
+
# 3. Check for any issues e.g. different dataset urns
|
|
450
|
+
if (
|
|
451
|
+
existing_assertion
|
|
452
|
+
and hasattr(existing_assertion, "dataset_urn")
|
|
453
|
+
and existing_assertion.dataset_urn != assertion_input.dataset_urn
|
|
454
|
+
):
|
|
455
|
+
raise SDKUsageError(
|
|
456
|
+
f"Dataset URN mismatch, existing assertion: {existing_assertion.dataset_urn} != new assertion: {dataset_urn}"
|
|
457
|
+
)
|
|
458
|
+
|
|
459
|
+
# 4. Merge the existing assertion with the validated input:
|
|
460
|
+
merged_assertion_input = self._merge_freshness_input(
|
|
461
|
+
dataset_urn=dataset_urn,
|
|
462
|
+
urn=urn,
|
|
463
|
+
display_name=display_name,
|
|
464
|
+
enabled=enabled,
|
|
465
|
+
detection_mechanism=detection_mechanism,
|
|
466
|
+
incident_behavior=incident_behavior,
|
|
467
|
+
tags=tags,
|
|
468
|
+
now_utc=now_utc,
|
|
469
|
+
assertion_input=assertion_input,
|
|
470
|
+
maybe_assertion_entity=maybe_assertion_entity,
|
|
471
|
+
maybe_monitor_entity=maybe_monitor_entity,
|
|
472
|
+
existing_assertion=existing_assertion,
|
|
473
|
+
schedule=schedule,
|
|
474
|
+
freshness_schedule_check_type=freshness_schedule_check_type,
|
|
475
|
+
lookback_window=lookback_window,
|
|
476
|
+
)
|
|
477
|
+
|
|
478
|
+
return merged_assertion_input
|
|
479
|
+
|
|
387
480
|
def _retrieve_assertion_and_monitor(
|
|
388
481
|
self,
|
|
389
482
|
assertion_input: _AssertionInput,
|
|
@@ -423,7 +516,7 @@ class AssertionsClient:
|
|
|
423
516
|
|
|
424
517
|
return maybe_assertion_entity, monitor_urn, maybe_monitor_entity
|
|
425
518
|
|
|
426
|
-
def
|
|
519
|
+
def _merge_smart_freshness_input(
|
|
427
520
|
self,
|
|
428
521
|
dataset_urn: Union[str, DatasetUrn],
|
|
429
522
|
urn: Union[str, AssertionUrn],
|
|
@@ -554,7 +647,129 @@ class AssertionsClient:
|
|
|
554
647
|
|
|
555
648
|
return merged_assertion_input
|
|
556
649
|
|
|
557
|
-
def
|
|
650
|
+
def _merge_freshness_input(
|
|
651
|
+
self,
|
|
652
|
+
dataset_urn: Union[str, DatasetUrn],
|
|
653
|
+
urn: Union[str, AssertionUrn],
|
|
654
|
+
display_name: Optional[str],
|
|
655
|
+
enabled: Optional[bool],
|
|
656
|
+
detection_mechanism: DetectionMechanismInputTypes,
|
|
657
|
+
incident_behavior: Optional[
|
|
658
|
+
Union[AssertionIncidentBehavior, list[AssertionIncidentBehavior]]
|
|
659
|
+
],
|
|
660
|
+
tags: Optional[TagsInputType],
|
|
661
|
+
now_utc: datetime,
|
|
662
|
+
assertion_input: _FreshnessAssertionInput,
|
|
663
|
+
maybe_assertion_entity: Optional[Assertion],
|
|
664
|
+
maybe_monitor_entity: Optional[Monitor],
|
|
665
|
+
existing_assertion: FreshnessAssertion,
|
|
666
|
+
schedule: Optional[Union[str, models.CronScheduleClass]],
|
|
667
|
+
freshness_schedule_check_type: Optional[
|
|
668
|
+
Union[str, models.FreshnessAssertionScheduleTypeClass]
|
|
669
|
+
] = None,
|
|
670
|
+
lookback_window: Optional[TimeWindowSizeInputTypes] = None,
|
|
671
|
+
) -> _FreshnessAssertionInput:
|
|
672
|
+
"""Merge the input with the existing assertion and monitor entities.
|
|
673
|
+
|
|
674
|
+
Args:
|
|
675
|
+
dataset_urn: The urn of the dataset to be monitored.
|
|
676
|
+
urn: The urn of the assertion.
|
|
677
|
+
display_name: The display name of the assertion.
|
|
678
|
+
enabled: Whether the assertion is enabled.
|
|
679
|
+
incident_behavior: The incident behavior to be applied to the assertion.
|
|
680
|
+
tags: The tags to be applied to the assertion.
|
|
681
|
+
now_utc: The current UTC time from when the function is called.
|
|
682
|
+
assertion_input: The validated input to the function.
|
|
683
|
+
maybe_assertion_entity: The existing assertion entity from the DataHub instance.
|
|
684
|
+
maybe_monitor_entity: The existing monitor entity from the DataHub instance.
|
|
685
|
+
existing_assertion: The existing assertion from the DataHub instance.
|
|
686
|
+
schedule: The schedule to be applied to the assertion.
|
|
687
|
+
freshness_schedule_check_type: The freshness schedule check type to be applied to the assertion.
|
|
688
|
+
lookback_window: The lookback window to be applied to the assertion.
|
|
689
|
+
|
|
690
|
+
Returns:
|
|
691
|
+
The merged assertion input.
|
|
692
|
+
"""
|
|
693
|
+
merged_assertion_input = _FreshnessAssertionInput(
|
|
694
|
+
urn=urn,
|
|
695
|
+
entity_client=self.client.entities,
|
|
696
|
+
dataset_urn=dataset_urn,
|
|
697
|
+
display_name=_merge_field(
|
|
698
|
+
display_name,
|
|
699
|
+
"display_name",
|
|
700
|
+
assertion_input,
|
|
701
|
+
existing_assertion,
|
|
702
|
+
maybe_assertion_entity.description if maybe_assertion_entity else None,
|
|
703
|
+
),
|
|
704
|
+
enabled=_merge_field(
|
|
705
|
+
enabled,
|
|
706
|
+
"enabled",
|
|
707
|
+
assertion_input,
|
|
708
|
+
existing_assertion,
|
|
709
|
+
existing_assertion.mode == AssertionMode.ACTIVE
|
|
710
|
+
if existing_assertion
|
|
711
|
+
else None,
|
|
712
|
+
),
|
|
713
|
+
schedule=_merge_field(
|
|
714
|
+
schedule,
|
|
715
|
+
"schedule",
|
|
716
|
+
assertion_input,
|
|
717
|
+
existing_assertion,
|
|
718
|
+
existing_assertion.schedule if existing_assertion else None,
|
|
719
|
+
),
|
|
720
|
+
freshness_schedule_check_type=_merge_field(
|
|
721
|
+
freshness_schedule_check_type,
|
|
722
|
+
"freshness_schedule_check_type",
|
|
723
|
+
assertion_input,
|
|
724
|
+
existing_assertion,
|
|
725
|
+
existing_assertion._freshness_schedule_check_type
|
|
726
|
+
if existing_assertion
|
|
727
|
+
else None,
|
|
728
|
+
),
|
|
729
|
+
lookback_window=_merge_field(
|
|
730
|
+
lookback_window,
|
|
731
|
+
"lookback_window",
|
|
732
|
+
assertion_input,
|
|
733
|
+
existing_assertion,
|
|
734
|
+
existing_assertion.lookback_window if existing_assertion else None,
|
|
735
|
+
),
|
|
736
|
+
detection_mechanism=_merge_field(
|
|
737
|
+
detection_mechanism,
|
|
738
|
+
"detection_mechanism",
|
|
739
|
+
assertion_input,
|
|
740
|
+
existing_assertion,
|
|
741
|
+
FreshnessAssertion._get_detection_mechanism(
|
|
742
|
+
maybe_assertion_entity, maybe_monitor_entity, default=None
|
|
743
|
+
)
|
|
744
|
+
if maybe_assertion_entity and maybe_monitor_entity
|
|
745
|
+
else None,
|
|
746
|
+
),
|
|
747
|
+
incident_behavior=_merge_field(
|
|
748
|
+
incident_behavior,
|
|
749
|
+
"incident_behavior",
|
|
750
|
+
assertion_input,
|
|
751
|
+
existing_assertion,
|
|
752
|
+
FreshnessAssertion._get_incident_behavior(maybe_assertion_entity)
|
|
753
|
+
if maybe_assertion_entity
|
|
754
|
+
else None,
|
|
755
|
+
),
|
|
756
|
+
tags=_merge_field(
|
|
757
|
+
tags,
|
|
758
|
+
"tags",
|
|
759
|
+
assertion_input,
|
|
760
|
+
existing_assertion,
|
|
761
|
+
maybe_assertion_entity.tags if maybe_assertion_entity else None,
|
|
762
|
+
),
|
|
763
|
+
created_by=existing_assertion.created_by
|
|
764
|
+
or DEFAULT_CREATED_BY, # Override with the existing assertion's created_by or the default created_by if not set
|
|
765
|
+
created_at=existing_assertion.created_at
|
|
766
|
+
or now_utc, # Override with the existing assertion's created_at or now if not set
|
|
767
|
+
updated_by=assertion_input.updated_by, # Override with the input's updated_by
|
|
768
|
+
updated_at=assertion_input.updated_at, # Override with the input's updated_at (now)
|
|
769
|
+
)
|
|
770
|
+
return merged_assertion_input
|
|
771
|
+
|
|
772
|
+
def _merge_smart_volume_input(
|
|
558
773
|
self,
|
|
559
774
|
dataset_urn: Union[str, DatasetUrn],
|
|
560
775
|
urn: Union[str, AssertionUrn],
|
|
@@ -919,6 +1134,112 @@ class AssertionsClient:
|
|
|
919
1134
|
# raise e
|
|
920
1135
|
return SmartVolumeAssertion._from_entities(assertion_entity, monitor_entity)
|
|
921
1136
|
|
|
1137
|
+
def _create_freshness_assertion(
|
|
1138
|
+
self,
|
|
1139
|
+
*,
|
|
1140
|
+
dataset_urn: Union[str, DatasetUrn],
|
|
1141
|
+
display_name: Optional[str] = None,
|
|
1142
|
+
enabled: bool = True,
|
|
1143
|
+
freshness_schedule_check_type: Optional[
|
|
1144
|
+
Union[str, models.FreshnessAssertionScheduleTypeClass]
|
|
1145
|
+
] = None,
|
|
1146
|
+
lookback_window: Optional[TimeWindowSizeInputTypes] = None,
|
|
1147
|
+
detection_mechanism: DetectionMechanismInputTypes = None,
|
|
1148
|
+
incident_behavior: Optional[
|
|
1149
|
+
Union[AssertionIncidentBehavior, list[AssertionIncidentBehavior]]
|
|
1150
|
+
] = None,
|
|
1151
|
+
tags: Optional[TagsInputType] = None,
|
|
1152
|
+
created_by: Optional[Union[str, CorpUserUrn]] = None,
|
|
1153
|
+
schedule: Optional[Union[str, models.CronScheduleClass]] = None,
|
|
1154
|
+
) -> FreshnessAssertion:
|
|
1155
|
+
"""Create a freshness assertion.
|
|
1156
|
+
|
|
1157
|
+
Note: keyword arguments are required.
|
|
1158
|
+
|
|
1159
|
+
The created assertion will use the default daily schedule ("0 0 * * *").
|
|
1160
|
+
|
|
1161
|
+
Args:
|
|
1162
|
+
dataset_urn: The urn of the dataset to be monitored.
|
|
1163
|
+
display_name: The display name of the assertion. If not provided, a random display
|
|
1164
|
+
name will be generated.
|
|
1165
|
+
enabled: Whether the assertion is enabled. Defaults to True.
|
|
1166
|
+
detection_mechanism: The detection mechanism to be used for the assertion. Information
|
|
1167
|
+
schema is recommended. Valid values are:
|
|
1168
|
+
- "information_schema" or DetectionMechanism.INFORMATION_SCHEMA
|
|
1169
|
+
- "audit_log" or DetectionMechanism.AUDIT_LOG
|
|
1170
|
+
- {
|
|
1171
|
+
"type": "last_modified_column",
|
|
1172
|
+
"column_name": "last_modified",
|
|
1173
|
+
"additional_filter": "last_modified > '2021-01-01'",
|
|
1174
|
+
} or DetectionMechanism.LAST_MODIFIED_COLUMN(column_name='last_modified',
|
|
1175
|
+
additional_filter='last_modified > 2021-01-01')
|
|
1176
|
+
- "datahub_operation" or DetectionMechanism.DATAHUB_OPERATION
|
|
1177
|
+
freshness_schedule_check_type: The freshness schedule check type to be applied to the assertion. Valid values are:
|
|
1178
|
+
- "since_the_last_check" or models.FreshnessAssertionScheduleTypeClass.SINCE_THE_LAST_CHECK
|
|
1179
|
+
- "cron" or models.FreshnessAssertionScheduleTypeClass.CRON
|
|
1180
|
+
lookback_window: The lookback window to be applied to the assertion. Valid values are:
|
|
1181
|
+
- from models.TimeWindowSize objects: models.TimeWindowSizeClass(
|
|
1182
|
+
unit=models.CalendarIntervalClass.DAY,
|
|
1183
|
+
multiple=1)
|
|
1184
|
+
- from TimeWindowSize objects: TimeWindowSize(unit='DAY', multiple=1)
|
|
1185
|
+
incident_behavior: The incident behavior to be applied to the assertion. Valid values are:
|
|
1186
|
+
- "raise_on_fail" or AssertionIncidentBehavior.RAISE_ON_FAIL
|
|
1187
|
+
- "resolve_on_pass" or AssertionIncidentBehavior.RESOLVE_ON_PASS
|
|
1188
|
+
tags: The tags to be applied to the assertion. Valid values are:
|
|
1189
|
+
- a list of strings (strings will be converted to TagUrn objects)
|
|
1190
|
+
- a list of TagUrn objects
|
|
1191
|
+
- a list of TagAssociationClass objects
|
|
1192
|
+
created_by: Optional urn of the user who created the assertion. The format is
|
|
1193
|
+
"urn:li:corpuser:<username>", which you can find on the Users & Groups page.
|
|
1194
|
+
The default is the datahub system user.
|
|
1195
|
+
TODO: Retrieve the SDK user as the default instead of the datahub system user.
|
|
1196
|
+
schedule: Optional cron formatted schedule for the assertion. If not provided, a default
|
|
1197
|
+
schedule will be used. The schedule determines when the assertion will be evaluated.
|
|
1198
|
+
The format is a cron expression, e.g. "0 * * * *" for every hour using UTC timezone.
|
|
1199
|
+
Alternatively, a models.CronScheduleClass object can be provided with string parameters
|
|
1200
|
+
cron and timezone. Use `from datahub.metadata import schema_classes as models` to import the class.
|
|
1201
|
+
|
|
1202
|
+
Returns:
|
|
1203
|
+
FreshnessAssertion: The created assertion.
|
|
1204
|
+
"""
|
|
1205
|
+
_print_experimental_warning()
|
|
1206
|
+
now_utc = datetime.now(timezone.utc)
|
|
1207
|
+
if created_by is None:
|
|
1208
|
+
logger.warning(
|
|
1209
|
+
f"Created by is not set, using {DEFAULT_CREATED_BY} as a placeholder"
|
|
1210
|
+
)
|
|
1211
|
+
created_by = DEFAULT_CREATED_BY
|
|
1212
|
+
assertion_input = _FreshnessAssertionInput(
|
|
1213
|
+
urn=None,
|
|
1214
|
+
entity_client=self.client.entities,
|
|
1215
|
+
dataset_urn=dataset_urn,
|
|
1216
|
+
display_name=display_name,
|
|
1217
|
+
enabled=enabled,
|
|
1218
|
+
detection_mechanism=detection_mechanism,
|
|
1219
|
+
freshness_schedule_check_type=freshness_schedule_check_type,
|
|
1220
|
+
lookback_window=lookback_window,
|
|
1221
|
+
incident_behavior=incident_behavior,
|
|
1222
|
+
tags=tags,
|
|
1223
|
+
created_by=created_by,
|
|
1224
|
+
created_at=now_utc,
|
|
1225
|
+
updated_by=created_by,
|
|
1226
|
+
updated_at=now_utc,
|
|
1227
|
+
schedule=schedule,
|
|
1228
|
+
)
|
|
1229
|
+
assertion_entity, monitor_entity = (
|
|
1230
|
+
assertion_input.to_assertion_and_monitor_entities()
|
|
1231
|
+
)
|
|
1232
|
+
# If assertion creation fails, we won't try to create the monitor
|
|
1233
|
+
self.client.entities.create(assertion_entity)
|
|
1234
|
+
# TODO: Wrap monitor creation in a try-except and delete the assertion if monitor creation fails (once delete is implemented https://linear.app/acryl-data/issue/OBS-1350/add-delete-method-to-entity-clientpy)
|
|
1235
|
+
# try:
|
|
1236
|
+
self.client.entities.create(monitor_entity)
|
|
1237
|
+
# except Exception as e:
|
|
1238
|
+
# logger.error(f"Error creating monitor: {e}")
|
|
1239
|
+
# self.client.entities.delete(assertion_entity)
|
|
1240
|
+
# raise e
|
|
1241
|
+
return FreshnessAssertion._from_entities(assertion_entity, monitor_entity)
|
|
1242
|
+
|
|
922
1243
|
def sync_smart_volume_assertion(
|
|
923
1244
|
self,
|
|
924
1245
|
*,
|
|
@@ -1106,6 +1427,171 @@ class AssertionsClient:
|
|
|
1106
1427
|
|
|
1107
1428
|
return SmartVolumeAssertion._from_entities(assertion_entity, monitor_entity)
|
|
1108
1429
|
|
|
1430
|
+
def sync_freshness_assertion(
|
|
1431
|
+
self,
|
|
1432
|
+
*,
|
|
1433
|
+
dataset_urn: Union[str, DatasetUrn],
|
|
1434
|
+
urn: Optional[Union[str, AssertionUrn]] = None,
|
|
1435
|
+
display_name: Optional[str] = None,
|
|
1436
|
+
enabled: Optional[bool] = None,
|
|
1437
|
+
detection_mechanism: DetectionMechanismInputTypes = None,
|
|
1438
|
+
incident_behavior: Optional[
|
|
1439
|
+
Union[AssertionIncidentBehavior, list[AssertionIncidentBehavior]]
|
|
1440
|
+
] = None,
|
|
1441
|
+
tags: Optional[TagsInputType] = None,
|
|
1442
|
+
updated_by: Optional[Union[str, CorpUserUrn]] = None,
|
|
1443
|
+
freshness_schedule_check_type: Optional[
|
|
1444
|
+
Union[str, models.FreshnessAssertionScheduleTypeClass]
|
|
1445
|
+
] = None,
|
|
1446
|
+
schedule: Optional[Union[str, models.CronScheduleClass]] = None,
|
|
1447
|
+
lookback_window: Optional[TimeWindowSizeInputTypes] = None,
|
|
1448
|
+
) -> FreshnessAssertion:
|
|
1449
|
+
"""Upsert and merge a freshness assertion.
|
|
1450
|
+
|
|
1451
|
+
Note: keyword arguments are required.
|
|
1452
|
+
|
|
1453
|
+
Upsert and merge is a combination of create and update. If the assertion does not exist,
|
|
1454
|
+
it will be created. If it does exist, it will be updated. Existing assertion fields will
|
|
1455
|
+
be updated if the input value is not None. If the input value is None, the existing value
|
|
1456
|
+
will be preserved. If the input value can be un-set e.g. by passing an empty list or
|
|
1457
|
+
empty string.
|
|
1458
|
+
|
|
1459
|
+
Schedule behavior:
|
|
1460
|
+
- Create case: Uses default daily schedule (\"0 0 * * *\") or provided schedule
|
|
1461
|
+
- Update case: Uses existing schedule or provided schedule.
|
|
1462
|
+
|
|
1463
|
+
Args:
|
|
1464
|
+
dataset_urn: The urn of the dataset to be monitored.
|
|
1465
|
+
urn: The urn of the assertion. If not provided, a urn will be generated and the assertion
|
|
1466
|
+
will be _created_ in the DataHub instance.
|
|
1467
|
+
display_name: The display name of the assertion. If not provided, a random display name
|
|
1468
|
+
will be generated.
|
|
1469
|
+
enabled: Whether the assertion is enabled. If not provided, the existing value
|
|
1470
|
+
will be preserved.
|
|
1471
|
+
detection_mechanism: The detection mechanism to be used for the assertion. Information
|
|
1472
|
+
schema is recommended. Valid values are:
|
|
1473
|
+
- "information_schema" or DetectionMechanism.INFORMATION_SCHEMA
|
|
1474
|
+
- "audit_log" or DetectionMechanism.AUDIT_LOG
|
|
1475
|
+
- {
|
|
1476
|
+
"type": "last_modified_column",
|
|
1477
|
+
"column_name": "last_modified",
|
|
1478
|
+
"additional_filter": "last_modified > '2021-01-01'",
|
|
1479
|
+
} or DetectionMechanism.LAST_MODIFIED_COLUMN(column_name='last_modified',
|
|
1480
|
+
additional_filter='last_modified > 2021-01-01')
|
|
1481
|
+
- {
|
|
1482
|
+
"type": "high_watermark_column",
|
|
1483
|
+
"column_name": "id",
|
|
1484
|
+
"additional_filter": "id > 1000",
|
|
1485
|
+
} or DetectionMechanism.HIGH_WATERMARK_COLUMN(column_name='id',
|
|
1486
|
+
additional_filter='id > 1000')
|
|
1487
|
+
- "datahub_operation" or DetectionMechanism.DATAHUB_OPERATION
|
|
1488
|
+
incident_behavior: The incident behavior to be applied to the assertion. Valid values are:
|
|
1489
|
+
- "raise_on_fail" or AssertionIncidentBehavior.RAISE_ON_FAIL
|
|
1490
|
+
- "resolve_on_pass" or AssertionIncidentBehavior.RESOLVE_ON_PASS
|
|
1491
|
+
tags: The tags to be applied to the assertion. Valid values are:
|
|
1492
|
+
- a list of strings (strings will be converted to TagUrn objects)
|
|
1493
|
+
- a list of TagUrn objects
|
|
1494
|
+
- a list of TagAssociationClass objects
|
|
1495
|
+
updated_by: Optional urn of the user who updated the assertion. The format is
|
|
1496
|
+
"urn:li:corpuser:<username>", which you can find on the Users & Groups page.
|
|
1497
|
+
The default is the datahub system user.
|
|
1498
|
+
TODO: Retrieve the SDK user as the default instead of the datahub system user.
|
|
1499
|
+
schedule: Optional cron formatted schedule for the assertion. If not provided, a default
|
|
1500
|
+
schedule will be used. The schedule determines when the assertion will be evaluated.
|
|
1501
|
+
The format is a cron expression, e.g. "0 * * * *" for every hour using UTC timezone.
|
|
1502
|
+
Alternatively, a models.CronScheduleClass object can be provided with string parameters
|
|
1503
|
+
cron and timezone. Use `from datahub.metadata import schema_classes as models` to import the class.
|
|
1504
|
+
|
|
1505
|
+
Returns:
|
|
1506
|
+
FreshnessAssertion: The created or updated assertion.
|
|
1507
|
+
"""
|
|
1508
|
+
_print_experimental_warning()
|
|
1509
|
+
now_utc = datetime.now(timezone.utc)
|
|
1510
|
+
|
|
1511
|
+
if updated_by is None:
|
|
1512
|
+
logger.warning(
|
|
1513
|
+
f"updated_by is not set, using {DEFAULT_CREATED_BY} as a placeholder"
|
|
1514
|
+
)
|
|
1515
|
+
updated_by = DEFAULT_CREATED_BY
|
|
1516
|
+
|
|
1517
|
+
# 1. If urn is not set, create a new assertion
|
|
1518
|
+
if urn is None:
|
|
1519
|
+
logger.info("URN is not set, creating a new assertion")
|
|
1520
|
+
return self._create_freshness_assertion(
|
|
1521
|
+
dataset_urn=dataset_urn,
|
|
1522
|
+
display_name=display_name,
|
|
1523
|
+
enabled=enabled if enabled is not None else True,
|
|
1524
|
+
detection_mechanism=detection_mechanism,
|
|
1525
|
+
incident_behavior=incident_behavior,
|
|
1526
|
+
tags=tags,
|
|
1527
|
+
created_by=updated_by,
|
|
1528
|
+
schedule=schedule,
|
|
1529
|
+
freshness_schedule_check_type=freshness_schedule_check_type,
|
|
1530
|
+
lookback_window=lookback_window,
|
|
1531
|
+
)
|
|
1532
|
+
|
|
1533
|
+
# 2. If urn is set, first validate the input:
|
|
1534
|
+
assertion_input = _FreshnessAssertionInput(
|
|
1535
|
+
urn=urn,
|
|
1536
|
+
entity_client=self.client.entities,
|
|
1537
|
+
dataset_urn=dataset_urn,
|
|
1538
|
+
display_name=display_name,
|
|
1539
|
+
detection_mechanism=detection_mechanism,
|
|
1540
|
+
incident_behavior=incident_behavior,
|
|
1541
|
+
tags=tags,
|
|
1542
|
+
created_by=updated_by, # This will be overridden by the actual created_by
|
|
1543
|
+
created_at=now_utc, # This will be overridden by the actual created_at
|
|
1544
|
+
updated_by=updated_by,
|
|
1545
|
+
updated_at=now_utc,
|
|
1546
|
+
schedule=schedule,
|
|
1547
|
+
freshness_schedule_check_type=freshness_schedule_check_type,
|
|
1548
|
+
lookback_window=lookback_window,
|
|
1549
|
+
)
|
|
1550
|
+
|
|
1551
|
+
# 3. Merge the assertion input with the existing assertion and monitor entities or create a new assertion
|
|
1552
|
+
# if the assertion does not exist:
|
|
1553
|
+
merged_assertion_input_or_created_assertion = (
|
|
1554
|
+
self._retrieve_and_merge_freshness_assertion_and_monitor(
|
|
1555
|
+
assertion_input=assertion_input,
|
|
1556
|
+
dataset_urn=dataset_urn,
|
|
1557
|
+
urn=urn,
|
|
1558
|
+
display_name=display_name,
|
|
1559
|
+
enabled=enabled,
|
|
1560
|
+
detection_mechanism=detection_mechanism,
|
|
1561
|
+
incident_behavior=incident_behavior,
|
|
1562
|
+
tags=tags,
|
|
1563
|
+
updated_by=updated_by,
|
|
1564
|
+
now_utc=now_utc,
|
|
1565
|
+
schedule=schedule,
|
|
1566
|
+
freshness_schedule_check_type=freshness_schedule_check_type,
|
|
1567
|
+
lookback_window=lookback_window,
|
|
1568
|
+
)
|
|
1569
|
+
)
|
|
1570
|
+
|
|
1571
|
+
# Return early if we created a new assertion in the merge:
|
|
1572
|
+
if isinstance(merged_assertion_input_or_created_assertion, _AssertionPublic):
|
|
1573
|
+
# We know this is the correct type because we passed the assertion_class parameter
|
|
1574
|
+
assert isinstance(
|
|
1575
|
+
merged_assertion_input_or_created_assertion, FreshnessAssertion
|
|
1576
|
+
)
|
|
1577
|
+
return merged_assertion_input_or_created_assertion
|
|
1578
|
+
|
|
1579
|
+
# 4. Upsert the assertion and monitor entities:
|
|
1580
|
+
assertion_entity, monitor_entity = (
|
|
1581
|
+
merged_assertion_input_or_created_assertion.to_assertion_and_monitor_entities()
|
|
1582
|
+
)
|
|
1583
|
+
# If assertion upsert fails, we won't try to upsert the monitor
|
|
1584
|
+
self.client.entities.upsert(assertion_entity)
|
|
1585
|
+
# TODO: Wrap monitor upsert in a try-except and delete the assertion if monitor upsert fails (once delete is implemented https://linear.app/acryl-data/issue/OBS-1350/add-delete-method-to-entity-clientpy)
|
|
1586
|
+
# try:
|
|
1587
|
+
self.client.entities.upsert(monitor_entity)
|
|
1588
|
+
# except Exception as e:
|
|
1589
|
+
# logger.error(f"Error upserting monitor: {e}")
|
|
1590
|
+
# self.client.entities.delete(assertion_entity)
|
|
1591
|
+
# raise e
|
|
1592
|
+
|
|
1593
|
+
return FreshnessAssertion._from_entities(assertion_entity, monitor_entity)
|
|
1594
|
+
|
|
1109
1595
|
|
|
1110
1596
|
def _merge_field(
|
|
1111
1597
|
input_field_value: Any,
|
|
@@ -34,7 +34,7 @@ AssertionInfoInputType: TypeAlias = Union[
|
|
|
34
34
|
models.FreshnessAssertionInfoClass,
|
|
35
35
|
models.VolumeAssertionInfoClass,
|
|
36
36
|
models.SqlAssertionInfoClass,
|
|
37
|
-
|
|
37
|
+
models.FieldAssertionInfoClass,
|
|
38
38
|
# TODO: models.SchemaAssertionInfoClass,
|
|
39
39
|
# TODO: models.CustomAssertionInfoClass,
|
|
40
40
|
]
|
|
@@ -189,6 +189,9 @@ class Assertion(HasPlatformInstance, HasTags, Entity):
|
|
|
189
189
|
elif isinstance(assertion, models.SqlAssertionInfoClass):
|
|
190
190
|
info.sqlAssertion = assertion
|
|
191
191
|
info.type = models.AssertionTypeClass.SQL
|
|
192
|
+
elif isinstance(assertion, models.FieldAssertionInfoClass):
|
|
193
|
+
info.fieldAssertion = assertion
|
|
194
|
+
info.type = models.AssertionTypeClass.FIELD
|
|
192
195
|
else:
|
|
193
196
|
assert_never(assertion)
|
|
194
197
|
|