def show(run_id: str) -> None:
"""Describe a provided ingestion run to datahub"""
payload_obj = {"runId": run_id, "dryRun": True, "hardDelete": True}
(
structured_rows,
entities_affected,
aspects_modified,
aspects_affected,
unsafe_entity_count,
unsafe_entities,
) = post_rollback_endpoint(payload_obj, "/runs?action=rollback")
if aspects_modified >= ELASTIC_MAX_PAGE_SIZE:
click.echo(
f"this run created at least {entities_affected} new entities and updated at least {aspects_modified} aspects"
)
else:
click.echo(
f"this run created {entities_affected} new entities and updated {aspects_modified} aspects"
)
click.echo(
"rolling back will delete the entities created and revert the updated aspects"
)
click.echo()
click.echo(
f"showing first {len(structured_rows)} of {aspects_modified} aspects touched by this run"
)
click.echo(tabulate(structured_rows, RUN_TABLE_COLUMNS, tablefmt="grid"))
def rollback(run_id: str, dry_run: bool) -> None:
"""Rollback a provided ingestion run to datahub"""
cli_utils.test_connectivity_complain_exit("ingest")
if not dry_run:
click.confirm(
"This will permanently delete data from DataHub. Do you want to continue?",
abort=True,
)
payload_obj = {"runId": run_id, "dryRun": dry_run}
structured_rows, entities_affected, aspects_affected = post_rollback_endpoint(
payload_obj, "/runs?action=rollback"
)
click.echo(
"Rolling back deletes the entities created by a run and reverts the updated aspects"
)
click.echo(
f"This rollback {'will' if dry_run else ''} {'delete' if dry_run else 'deleted'} {entities_affected} entities and {'will roll' if dry_run else 'rolled'} back {aspects_affected} aspects"
)
click.echo(
f"showing first {len(structured_rows)} of {aspects_affected} aspects {'that will be' if dry_run else ''} reverted by this run"
)
click.echo(tabulate(structured_rows, RUN_TABLE_COLUMNS, tablefmt="grid"))
def rollback(run_id: str) -> None:
"""Rollback a provided ingestion run to datahub"""
click.confirm(
"This will permanently delete data from DataHub. Do you want to continue?",
abort=True,
)
payload_obj = {"runId": run_id, "dryRun": False}
structured_rows, entities_affected, aspects_affected = post_rollback_endpoint(
payload_obj, "/runs?action=rollback")
click.echo(
"rolling back deletes the entities created by a run and reverts the updated aspects"
)
click.echo(
f"this rollback deleted {entities_affected} entities and rolled back {aspects_affected} aspects"
)
click.echo(
f"showing first {len(structured_rows)} of {aspects_affected} aspects reverted by this run"
)
click.echo(tabulate(structured_rows, RUN_TABLE_COLUMNS, tablefmt="grid"))
def delete_for_registry(
registry_id: str,
soft: bool,
dry_run: bool,
) -> DeletionResult:
deletion_result = DeletionResult()
deletion_result.num_entities = 1
deletion_result.num_records = UNKNOWN_NUM_RECORDS # Default is unknown
registry_delete = {"registryId": registry_id, "dryRun": dry_run, "soft": soft}
(
structured_rows,
entities_affected,
aspects_affected,
unsafe_aspects,
unsafe_entity_count,
unsafe_entities,
) = cli_utils.post_rollback_endpoint(registry_delete, "/entities?action=deleteAll")
deletion_result.num_entities = entities_affected
deletion_result.num_records = aspects_affected
deletion_result.sample_records = structured_rows
deletion_result.end()
return deletion_result
def rollback(
run_id: str, force: bool, dry_run: bool, safe: bool, report_dir: str
) -> None:
"""Rollback a provided ingestion run to datahub"""
cli_utils.test_connectivity_complain_exit("ingest")
if not force and not dry_run:
click.confirm(
"This will permanently delete data from DataHub. Do you want to continue?",
abort=True,
)
payload_obj = {"runId": run_id, "dryRun": dry_run, "safe": safe}
(
structured_rows,
entities_affected,
aspects_reverted,
aspects_affected,
unsafe_entity_count,
unsafe_entities,
) = post_rollback_endpoint(payload_obj, "/runs?action=rollback")
click.echo(
"Rolling back deletes the entities created by a run and reverts the updated aspects"
)
click.echo(
f"This rollback {'will' if dry_run else ''} {'delete' if dry_run else 'deleted'} {entities_affected} entities and {'will roll' if dry_run else 'rolled'} back {aspects_reverted} aspects"
)
click.echo(
f"showing first {len(structured_rows)} of {aspects_reverted} aspects {'that will be ' if dry_run else ''}reverted by this run"
)
click.echo(tabulate(structured_rows, RUN_TABLE_COLUMNS, tablefmt="grid"))
if aspects_affected > 0:
if safe:
click.echo(
f"WARNING: This rollback {'will hide' if dry_run else 'has hidden'} {aspects_affected} aspects related to {unsafe_entity_count} entities being rolled back that are not part ingestion run id."
)
else:
click.echo(
f"WARNING: This rollback {'will delete' if dry_run else 'has deleted'} {aspects_affected} aspects related to {unsafe_entity_count} entities being rolled back that are not part ingestion run id."
)
if unsafe_entity_count > 0:
now = datetime.now()
current_time = now.strftime("%Y-%m-%d %H:%M:%S")
try:
folder_name = report_dir + "/" + current_time
ingestion_config_file_name = folder_name + "/config.json"
os.makedirs(os.path.dirname(ingestion_config_file_name), exist_ok=True)
with open(ingestion_config_file_name, "w") as file_handle:
json.dump({"run_id": run_id}, file_handle)
csv_file_name = folder_name + "/unsafe_entities.csv"
with open(csv_file_name, "w") as file_handle:
writer = csv.writer(file_handle)
writer.writerow(["urn"])
for row in unsafe_entities:
writer.writerow([row.get("urn")])
except IOError as e:
print(e)
sys.exit("Unable to write reports to " + report_dir)
