def set_up_to_succeed(app, caplog):
(bucket, source_prefix, dest_prefix) = get_s3_refs(app)
caplog.set_level(logging.INFO)
with capture_app_logs(app):
with mock_s3(app, bucket=bucket) as m3:
m3.Object(bucket,
f'{source_prefix}/2017/01/18/12345678_00012_1.pdf').put(
Body=b'a note attachment')
m3.Object(bucket,
f'{source_prefix}/2018/12/22/23456789_00003_1.png').put(
Body=b'another note attachment')
m3.Object(bucket,
f'{source_prefix}/2019/08/29/34567890_00014_2.xls').put(
Body=b'yet another note attachment')
m3.Object(bucket,
f'{dest_prefix}/12345678/12345678_00012_1.pdf').put(
Body=b'a note attachment')
m3.Object(bucket,
f'{dest_prefix}/23456789/23456789_00003_1.png').put(
Body=b'another note attachment')
m3.Object(bucket,
f'{dest_prefix}/34567890/34567890_00014_2.xls').put(
Body=b'yet another note attachment')
yield
assert 'No attachments missing on S3 when compared against the view.' in caplog.text
def test_run_with_all_param(self, app, caplog):
"""When 'all' is provided, copies all files."""
(bucket, source_prefix, dest_prefix) = get_s3_refs(app)
datestamp = 'all'
caplog.set_level(logging.INFO)
with capture_app_logs(app):
with mock_s3(app, bucket=bucket) as m3:
m3.Object(
bucket,
f'{source_prefix}/2019/08/28/12345678_00012_1.pdf').put(
Body=b'a note attachment')
m3.Object(
bucket,
f'{source_prefix}/2019/08/28/23456789_00003_1.png').put(
Body=b'another note attachment')
m3.Object(
bucket,
f'{source_prefix}/2019/08/29/34567890_00014_2.xls').put(
Body=b'ok to copy me')
response = MigrateSisAdvisingNoteAttachments().run(
datestamp=datestamp)
assert 'Will copy files from /sis-data/sis-sftp/incremental/advising-notes/attachment-files.' in caplog.text
assert 'Copied 3 attachments to the destination folder.' in caplog.text
assert response == (
'SIS advising note attachment migration complete for sis-data/sis-sftp/incremental/advising-notes/attachment-files.'
)
assert object_exists(
m3, bucket, f'{dest_prefix}/12345678/12345678_00012_1.pdf')
assert object_exists(
m3, bucket, f'{dest_prefix}/23456789/23456789_00003_1.png')
assert object_exists(
m3, bucket, f'{dest_prefix}/34567890/34567890_00014_2.xls')
def test_s3_nonexistent_object(self, app, caplog, bad_bucket):
"""Returns false on S3 checks for nonexistent objects."""
with capture_app_logs(app):
key = app.config[
'LOCH_S3_PREFIX_TESTEXT'] + '/00001/sonnet-xlv.html'
response = s3.object_exists(key)
assert response is False
def test_sync_canvas_snapshots(self, app, metadata_db, caplog):
"""Dispatches a complete sync job against fixtures."""
caplog.set_level(logging.INFO)
with capture_app_logs(app):
with mock_s3(app):
result = SyncCanvasSnapshots().run_wrapped()
assert 'Canvas snapshot sync job dispatched to workers' in result
assert_background_job_status('sync')
assert 'Dispatched S3 sync of snapshot quiz_dim-00000-0ab80c7c.gz' in caplog.text
assert 'Dispatched S3 sync of snapshot requests-00098-b14782f5.gz' in caplog.text
assert '311 successful dispatches, 0 failures' in caplog.text
schema = app.config['RDS_SCHEMA_METADATA']
count_results = rds.fetch(f'SELECT count(*) FROM {schema}.canvas_sync_job_status')
assert count_results[0]['count'] == 311
canvas_status_results = rds.fetch(f'SELECT DISTINCT status FROM {schema}.canvas_sync_job_status')
assert len(canvas_status_results) == 1
assert canvas_status_results[0]['status'] == 'created'
sync_results = rds.fetch(f'SELECT * FROM {schema}.canvas_sync_job_status LIMIT 1')
assert sync_results[0]['job_id'].startswith('sync_')
assert sync_results[0]['filename'] == 'account_dim-00000-5eb7ee9e.gz'
assert sync_results[0]['canvas_table'] == 'account_dim'
assert 'account_dim/part-00505-5c40f1f3-b611-4f64-a007-67b775e984fe.c000.txt.gz' in sync_results[0]['source_url']
assert sync_results[0]['destination_url'] is None
assert sync_results[0]['details'] is None
assert sync_results[0]['created_at']
assert sync_results[0]['updated_at']
def test_malformed_filenames(self, app, caplog):
(bucket, source_prefix, dest_prefix) = get_s3_refs(app)
datestamp = 'all'
caplog.set_level(logging.INFO)
with capture_app_logs(app):
with mock_s3(app, bucket=bucket) as m3:
m3.Object(
bucket,
f'{source_prefix}/2019/08/28/12345678_00012_1_May_7_2019_email.pdf'
).put(Body=b'extra chars in my name lol')
m3.Object(
bucket,
f'{source_prefix}/2019/08/28/23456789_00052_1.png.png'
).put(Body=b'somehow i got a redundant .ext')
m3.Object(
bucket,
f'{source_prefix}/2019/08/29/23456789_00053_1._DEGREE_COMPLETION_LETTER'
).put(Body=b'original file name mistaken for the .ext', )
m3.Object(
bucket,
f'{source_prefix}/2019/08/29/34567890_00014_2..7.19_(2)-edited_(1)-2_(1)_(1).xls'
).put(Body=b'is this a versioning scheme?', )
MigrateSisAdvisingNoteAttachments().run(datestamp=datestamp)
assert 'Copied 4 attachments to the destination folder.' in caplog.text
assert object_exists(
m3, bucket, f'{dest_prefix}/12345678/12345678_00012_1.pdf')
assert object_exists(
m3, bucket, f'{dest_prefix}/23456789/23456789_00052_1.png')
assert object_exists(
m3, bucket, f'{dest_prefix}/23456789/23456789_00053_1')
assert object_exists(
m3, bucket, f'{dest_prefix}/34567890/34567890_00014_2.xls')
def test_remove_obsolete_files(self, app, caplog, cleanup_s3):
"""Removes files from S3 following prefix and whitelist rules."""
caplog.set_level(logging.INFO)
with capture_app_logs(app):
prefix1 = app.config['LOCH_S3_PREFIX_TESTEXT'] + '/001'
prefix2 = app.config['LOCH_S3_PREFIX_TESTEXT'] + '/002'
assert s3.upload_from_url(
'http://shakespeare.mit.edu/Poetry/sonnet.XX.html',
prefix1 + '/xx/sonnet-xx.html')
assert s3.upload_from_url(
'http://shakespeare.mit.edu/Poetry/sonnet.XXI.html',
prefix1 + '/xxi/sonnet-xxi.html')
assert s3.upload_from_url(
'http://shakespeare.mit.edu/Poetry/sonnet.XXII.html',
prefix1 + '/xxii/sonnet-xxii.html')
assert s3.upload_from_url(
'http://shakespeare.mit.edu/Poetry/sonnet.XLV.html',
prefix2 + '/xlv/sonnet-xlv.html')
whitelist = ['sonnet-xxi.html', 'sonnet-xxii.html']
assert s3.delete_objects_with_prefix(prefix1, whitelist) is True
assert f'3 key(s) matching prefix "{prefix1}"' in caplog.text
assert '2 key(s) in whitelist' in caplog.text
assert 'will delete 1 object(s)' in caplog.text
assert s3.object_exists(prefix1 + '/xx/sonnet-xx.html') is False
assert s3.object_exists(prefix1 + '/xxi/sonnet-xxi.html') is True
assert s3.object_exists(prefix1 + '/xxii/sonnet-xxii.html') is True
assert s3.object_exists(prefix2 + '/xlv/sonnet-xlv.html') is True
def test_import_student_photos(self, app, metadata_db, student_tables,
caplog):
from nessie.jobs.import_student_photos import ImportStudentPhotos
caplog.set_level(logging.DEBUG)
with capture_app_logs(app):
with mock_s3(app):
result = ImportStudentPhotos().run_wrapped()
assert result == 'Student photo import completed: 1 succeeded, 9 had no photo available, 0 failed.'
response = s3.get_keys_with_prefix('cal1card-data/photos')
assert len(response) == 1
assert response[0] == 'cal1card-data/photos/61889.jpg'
success_rows = rds.fetch(
f"SELECT * FROM {app.config['RDS_SCHEMA_METADATA']}.photo_import_status WHERE status = 'success'"
)
assert len(success_rows) == 1
assert success_rows[0]['sid'] == '11667051'
failure_rows = rds.fetch(
f"SELECT * FROM {app.config['RDS_SCHEMA_METADATA']}.photo_import_status WHERE status = 'failure'"
)
assert len(failure_rows) == 0
not_found_rows = rds.fetch(
f"SELECT * FROM {app.config['RDS_SCHEMA_METADATA']}.photo_import_status WHERE status = 'photo_not_found'"
)
assert len(not_found_rows) == 9
def set_up_to_fail(app, caplog):
(bucket, source_prefix, dest_prefix) = get_s3_refs(app)
caplog.set_level(logging.INFO)
with capture_app_logs(app):
with mock_s3(app, bucket=bucket) as m3:
m3.Object(bucket,
f'{source_prefix}/2017/01/18/12345678_00012_1.pdf').put(
Body=b'a note attachment')
m3.Object(bucket,
f'{source_prefix}/2018/12/22/23456789_00003_1.png').put(
Body=b'another note attachment')
m3.Object(bucket,
f'{dest_prefix}/12345678/12345678_00012_1.pdf').put(
Body=b'a note attachment')
m3.Object(bucket,
f'{dest_prefix}/34567890/34567890_00014_2.xls').put(
Body=b'yet another note attachment')
m3.Object(
bucket, f'{dest_prefix}/45678901/45678901_00192_4.xls'
).put(
Body=b'bamboozled by a completely unexpected note attachment')
with pytest.raises(BackgroundJobError) as e:
yield
assert 'Attachments verification found missing attachments or sync failures:' in str(
e.value)
assert '\'attachment_sync_failure_count\': 1' in str(e.value)
assert '\'missing_s3_attachments_count\': 1' in str(e.value)
assert '\'attachment_sync_failures\': [\'sis-data/sis-sftp/incremental/advising-notes/attachment-files/2018/12/22/23456789_00003_1.png\']' in str(
e.value, )
assert '\'missing_s3_attachments\': [\'23456789_00003_1.png\']' in str(
e.value)
assert 'Attachments missing on S3 when compared against SIS notes views: 1' in caplog.text
def test_import_registrations(self, app, metadata_db, student_tables, caplog):
from nessie.jobs.import_registrations import ImportRegistrations
rows = redshift.fetch('SELECT * FROM student_test.student_term_gpas')
assert len(rows) == 0
rows = redshift.fetch('SELECT * FROM student_test.student_last_registrations')
assert len(rows) == 0
caplog.set_level(logging.DEBUG)
with capture_app_logs(app):
with mock_s3(app):
result = ImportRegistrations().run_wrapped()
assert result == 'Registrations import completed: 2 succeeded, 8 failed.'
rows = redshift.fetch('SELECT * FROM student_test.student_term_gpas ORDER BY sid')
assert len(rows) == 11
for row in rows[0:6]:
assert row['sid'] == '11667051'
for row in rows[7:10]:
assert row['sid'] == '1234567890'
row_2168 = next(r for r in rows if r['term_id'] == '2168')
assert row_2168['gpa'] == Decimal('3.000')
assert row_2168['units_taken_for_gpa'] == Decimal('8.0')
rows = redshift.fetch('SELECT * FROM student_test.student_last_registrations ORDER BY sid')
assert len(rows) == 2
assert rows[0]['sid'] == '11667051'
assert rows[1]['sid'] == '1234567890'
feed = json.loads(rows[1]['feed'], strict=False)
assert feed['term']['id'] == '2172'
assert feed['academicLevels'][0]['level']['description'] == 'Sophomore'
rows = redshift.fetch('SELECT * FROM student_test.student_api_demographics ORDER BY sid')
assert len(rows) == 2
assert rows[0]['sid'] == '11667051'
assert rows[1]['sid'] == '1234567890'
feed = json.loads(rows[1]['feed'], strict=False)
assert feed['gender']['genderOfRecord']['description'] == 'Female'
def test_server_error(self, app, caplog):
"""Logs unexpected server errors."""
with capture_app_logs(app):
canvas_error = MockResponse(429, {}, '{"message": "Easy, tiger."}')
with register_mock(canvas_data.get_snapshots, canvas_error):
response = canvas_data.get_snapshots()
assert '429 Client Error: Too Many Requests' in caplog.text
assert not response
def test_server_error(self, app, caplog):
"""Logs unexpected server errors."""
with capture_app_logs(app):
canvas_error = MockResponse(401, {},
'{"message": "Unauthorized."}')
with register_mock(dispatcher.dispatch, canvas_error):
response = dispatcher.dispatch('create_canvas_schema')
assert '401 Client Error: Unauthorized' in caplog.text
assert not response
def test_s3_upload_error_handling(self, app, caplog, bad_bucket):
"""Handles and logs connection errors on S3 upload."""
with capture_app_logs(app):
url = 'http://shakespeare.mit.edu/Poetry/sonnet.XLV.html'
key = app.config[
'LOCH_S3_PREFIX_TESTEXT'] + '/00001/sonnet-xlv.html'
with pytest.raises(ValueError):
s3.upload_from_url(url, key)
assert 'Error on S3 upload' in caplog.text
assert 'the bucket \'not-a-bucket-nohow\' does not exist, or is forbidden for access' in caplog.text
def test_run_with_no_param(self, mock_datetime, app, caplog, metadata_db,
prior_job_status):
"""When no parameter is provided, copies new files since the last succesful run."""
(bucket, source_prefix, dest_prefix) = get_s3_refs(app)
mock_datetime.utcnow.return_value = datetime(year=2019,
month=8,
day=29,
hour=5,
minute=21)
caplog.set_level(logging.INFO)
with capture_app_logs(app):
with mock_s3(app, bucket=bucket) as m3:
m3.Object(
bucket,
f'{source_prefix}/2019/08/25/45678912_00027_1.pdf').put(
Body=b'i\'ve already been copied')
m3.Object(
bucket,
f'{source_prefix}/2019/08/26/12345678_00012_1.pdf').put(
Body=b'a note attachment')
m3.Object(
bucket,
f'{source_prefix}/2019/08/28/23456789_00003_1.png').put(
Body=b'another note attachment')
m3.Object(
bucket,
f'{source_prefix}/2019/08/29/34567890_00014_2.xls').put(
Body=b'don\'t copy me')
response = MigrateSisAdvisingNoteAttachments().run()
assert 'Will copy files from /sis-data/sis-sftp/incremental/advising-notes/attachment-files/2019/08/25.' not in caplog.text
assert 'Will copy files from /sis-data/sis-sftp/incremental/advising-notes/attachment-files/2019/08/26.' in caplog.text
assert 'Will copy files from /sis-data/sis-sftp/incremental/advising-notes/attachment-files/2019/08/27.' in caplog.text
assert 'Will copy files from /sis-data/sis-sftp/incremental/advising-notes/attachment-files/2019/08/28.' in caplog.text
assert 'Will copy files from /sis-data/sis-sftp/incremental/advising-notes/attachment-files/2019/08/29.' not in caplog.text
assert 'Copied 1 attachments to the destination folder.' in caplog.text
assert 'Copied 0 attachments to the destination folder.' in caplog.text
assert response == (
'SIS advising note attachment migration complete for sis-data/sis-sftp/incremental/advising-notes/attachment-files/2019/08/26, \
sis-data/sis-sftp/incremental/advising-notes/attachment-files/2019/08/27, \
sis-data/sis-sftp/incremental/advising-notes/attachment-files/2019/08/28.')
assert not object_exists(
m3, bucket, f'{dest_prefix}/45678912/45678912_00027_1.xls')
assert object_exists(
m3, bucket, f'{dest_prefix}/12345678/12345678_00012_1.pdf')
assert object_exists(
m3, bucket, f'{dest_prefix}/23456789/23456789_00003_1.png')
assert not object_exists(
m3, bucket, f'{dest_prefix}/34567890/34567890_00014_2.xls')
def test_schema_creation_drop(self, app, caplog, ensure_drop_schema):
"""Can create and drop schemata on a real Redshift instance."""
schema_name = app.config['REDSHIFT_SCHEMA_BOAC']
schema = psycopg2.sql.Identifier(schema_name)
with capture_app_logs(app):
result = redshift.execute('CREATE SCHEMA {schema}', schema=schema)
assert result == 'CREATE SCHEMA'
result = redshift.execute('CREATE SCHEMA {schema}', schema=schema)
assert result is None
assert f'Schema "{schema_name}" already exists' in caplog.text
result = redshift.execute('DROP SCHEMA {schema}', schema=schema)
assert result == 'DROP SCHEMA'
def test_aborts_on_missing_term(self, app, caplog):
from nessie.jobs.create_sis_schema import CreateSisSchema
with mock_s3(app):
daily_path = get_s3_sis_daily_path()
historical_path = app.config[
'LOCH_S3_SIS_DATA_PATH'] + '/historical'
self._upload_data_to_s3(daily_path, historical_path)
s3.delete_objects(
[f'{daily_path}/enrollments/enrollments-2178.gz'])
with capture_app_logs(app):
with pytest.raises(BackgroundJobError) as e:
CreateSisSchema().update_manifests()
assert 'Expected filename enrollments-2178.gz not found in S3, aborting' in str(
e.value)
def test_import_term_gpas(self, app, metadata_db, student_tables, caplog):
from nessie.jobs.import_term_gpas import ImportTermGpas
caplog.set_level(logging.DEBUG)
with capture_app_logs(app):
with mock_s3(app):
result = ImportTermGpas().run_wrapped()
assert result == 'Term GPA import completed: 1 succeeded, 0 returned no registrations, 7 failed.'
rows = redshift.fetch(
'SELECT * FROM student_test.student_term_gpas')
assert len(rows) == 7
for row in rows:
assert row['sid'] == '11667051'
row_2178 = next(r for r in rows if r['term_id'] == '2178')
assert row_2178['gpa'] == Decimal('3.000')
assert row_2178['units_taken_for_gpa'] == Decimal('8.0')
def test_file_upload_and_skip(self, app, caplog, cleanup_s3):
"""Uploads files to real S3, skipping duplicates."""
url = 'http://shakespeare.mit.edu/Poetry/sonnet.XLV.html'
key = app.config['LOCH_S3_PREFIX_TESTEXT'] + '/00001/sonnet-xlv.html'
caplog.set_level(logging.DEBUG)
with capture_app_logs(app):
result = SyncFileToS3().run(url=url, key=key)
assert result is True
assert f'Key {key} does not exist, starting upload' in caplog.text
assert 'S3 upload complete' in caplog.text
result = SyncFileToS3().run(url=url, key=key)
assert result is False
assert f'Key {key} exists, skipping upload' in caplog.text
def test_affiliations_conflict(self, app, caplog):
feed = {
'academicStatuses': [
_active_grad_academic_status(),
],
'affiliations': [
_active_ucbx_affiliation(),
],
}
profile = {}
with capture_app_logs(app):
merge_sis_profile_academic_status(feed, profile)
assert profile['academicCareer'] == 'GRAD'
assert profile.get('academicCareerStatus') is None
assert profile['plans'][0]['description'] == 'On-Campus/Online Prfsnl MPH'
assert profile['plans'][0]['status'] == 'Active'
assert 'Conflict between affiliations and academicStatuses' in caplog.text
def test_source_url_error_handling(self, app, caplog):
"""Handles and logs connection errors to source URL."""
with capture_app_logs(app):
url = 'http://shakespeare.mit.edu/Poetry/sonnet.XLV.html'
key = app.config[
'LOCH_S3_PREFIX_TESTEXT'] + '/00001/sonnet-xlv.html'
responses.add(responses.GET,
url,
status=500,
body='{"message": "Internal server error."}')
with pytest.raises(ConnectionError):
s3.upload_from_url(url, key)
assert 'Received unexpected status code, aborting S3 upload' in caplog.text
assert 'status=500' in caplog.text
assert 'body={"message": "Internal server error."}' in caplog.text
assert f'url={url}' in caplog.text
assert f'key={key}' in caplog.text
def test_first_time_run_with_no_param(self, mock_datetime, app, caplog,
metadata_db):
"""When no parameter is provided and there is no prior successful run, copies all files."""
(bucket, source_prefix, dest_prefix) = get_s3_refs(app)
mock_datetime.utcnow.return_value = datetime(year=2019,
month=8,
day=29,
hour=5,
minute=21)
caplog.set_level(logging.INFO)
with capture_app_logs(app):
with mock_s3(app, bucket=bucket) as m3:
m3.Object(
bucket,
f'{source_prefix}/2019/08/28/12345678_00012_1.pdf').put(
Body=b'a note attachment')
m3.Object(
bucket,
f'{source_prefix}/2019/08/28/23456789_00003_1.png').put(
Body=b'another note attachment')
m3.Object(
bucket,
f'{source_prefix}/2019/08/29/34567890_00014_2.xls').put(
Body=b'ok to copy me')
response = MigrateSisAdvisingNoteAttachments().run()
assert 'Will copy files from /sis-data/sis-sftp/incremental/advising-notes/attachment-files/.' in caplog.text
assert 'Copied 3 attachments to the destination folder.' in caplog.text
assert response == (
'SIS advising note attachment migration complete for sis-data/sis-sftp/incremental/advising-notes/attachment-files/.'
)
assert object_exists(
m3, bucket, f'{dest_prefix}/12345678/12345678_00012_1.pdf')
assert object_exists(
m3, bucket, f'{dest_prefix}/23456789/23456789_00003_1.png')
assert object_exists(
m3, bucket, f'{dest_prefix}/34567890/34567890_00014_2.xls')
def test_run_with_invalid_param(self, app, caplog):
"""When invalid value is provided, job completes but copies zero files."""
(bucket, source_prefix, dest_prefix) = get_s3_refs(app)
datestamp = 'wrong!#$&'
caplog.set_level(logging.INFO)
with capture_app_logs(app):
with mock_s3(app, bucket=bucket) as m3:
m3.Object(
bucket,
f'{source_prefix}/2019/08/28/12345678_00012_1.pdf').put(
Body=b'a note attachment')
response = MigrateSisAdvisingNoteAttachments().run(
datestamp=datestamp)
assert 'Will copy files from /sis-data/sis-sftp/incremental/advising-notes/attachment-files/wrong!#$&.' in caplog.text
assert 'Copied 0 attachments to the destination folder.' in caplog.text
assert response == (
'SIS advising note attachment migration complete for sis-data/sis-sftp/incremental/advising-notes/attachment-files/wrong!#$&.'
)
assert not object_exists(
m3, bucket, f'{dest_prefix}/12345678/12345678_00012_1.pdf')
def test_metadata_tracked(self, app, metadata_db, student_tables, caplog):
from nessie.jobs.import_registrations import ImportRegistrations
rows = rds.fetch('SELECT * FROM nessie_metadata_test.registration_import_status')
assert len(rows) == 0
caplog.set_level(logging.DEBUG)
with capture_app_logs(app):
with mock_s3(app):
ImportRegistrations().run_wrapped()
rows = rds.fetch('SELECT * FROM nessie_metadata_test.registration_import_status')
assert len(rows) == 10
assert len([r for r in rows if r['status'] == 'failure']) == 8
assert next(r['status'] for r in rows if r['sid'] == '11667051') == 'success'
result = ImportRegistrations().run_wrapped()
assert result == 'Registrations import completed: 0 succeeded, 8 failed.'
result = ImportRegistrations().run_wrapped(load_mode='all')
assert result == 'Registrations import completed: 2 succeeded, 8 failed.'
rds.execute("DELETE FROM nessie_metadata_test.registration_import_status WHERE sid = '11667051'")
result = ImportRegistrations().run_wrapped()
assert result == 'Registrations import completed: 1 succeeded, 8 failed.'
assert next(r['status'] for r in rows if r['sid'] == '11667051') == 'success'
rds.execute("UPDATE nessie_metadata_test.registration_import_status SET status='failure' WHERE sid = '11667051'")
result = ImportRegistrations().run_wrapped()
assert result == 'Registrations import completed: 1 succeeded, 8 failed.'
assert next(r['status'] for r in rows if r['sid'] == '11667051') == 'success'
def test_course_not_found(self, app, caplog):
"""Logs 404 for unknown course."""
with capture_app_logs(app):
response = canvas_api.get_course_enrollments(9999999)
assert '404 Client Error' in caplog.text
assert not response
def test_resync_canvas_snapshots(self, app, metadata_db, caplog):
"""Dispatches a complete resync job against fixtures."""
caplog.set_level(logging.INFO)
snapshots = canvas_data.get_snapshots()['files']
def mock_metadata(job_id, snapshot, status, destination_size):
metadata.create_canvas_sync_status(job_id, snapshot['filename'],
snapshot['table'],
snapshot['url'])
key = '/'.join([
get_s3_canvas_daily_path(), snapshot['table'],
snapshot['filename']
])
metadata.update_canvas_sync_status(
job_id,
key,
status,
source_size=1048576,
destination_size=destination_size)
old_sync_job = 'sync_152550000'
latest_sync_job = 'sync_152560000'
# The older job should be ignored by the resync.
for snapshot in snapshots[0:5]:
mock_metadata(old_sync_job, snapshot, 'complete', 1048576)
for snapshot in snapshots[5:10]:
mock_metadata(old_sync_job, snapshot, 'error', None)
# The latest job synced five files successfully and ran into three problems.
for snapshot in snapshots[10:15]:
mock_metadata(latest_sync_job, snapshot, 'complete', 1048576)
stalled = snapshots[15]
errored = snapshots[16]
size_discrepancy = snapshots[17]
mock_metadata(latest_sync_job, stalled, 'streaming', None)
mock_metadata(latest_sync_job, errored, 'error', None)
mock_metadata(latest_sync_job, size_discrepancy, 'complete', 65536)
schema = app.config['RDS_SCHEMA_METADATA']
with capture_app_logs(app):
assert rds.fetch(
f'SELECT count(*) FROM {schema}.canvas_sync_job_status'
)[0]['count'] == 18
with mock_s3(app):
result = ResyncCanvasSnapshots().run_wrapped()
assert 'Canvas snapshot resync job dispatched to workers' in result
assert_background_job_status('resync')
assert f"Dispatched S3 resync of snapshot {stalled['filename']}" in caplog.text
assert f"Dispatched S3 resync of snapshot {errored['filename']}" in caplog.text
assert f"Dispatched S3 resync of snapshot {size_discrepancy['filename']}" in caplog.text
assert '3 successful dispatches, 0 failures' in caplog.text
assert rds.fetch(
f'SELECT count(*) FROM {schema}.canvas_sync_job_status'
)[0]['count'] == 21
resync_results = rds.fetch(
f"SELECT * FROM {schema}.canvas_sync_job_status WHERE job_id LIKE 'resync%'"
)
assert len(resync_results) == 3
urls = []
for r in resync_results:
assert r['job_id'].startswith('resync_')
assert r['filename']
assert r['canvas_table']
assert r['created_at']
assert r['updated_at']
urls.append(r['source_url'])
assert stalled['url'] in urls
assert errored['url'] in urls
assert size_discrepancy['url'] in urls
def test_connection_error_handling(self, app, caplog):
"""Handles and logs connection errors."""
with capture_app_logs(app):
with override_config(app, 'REDSHIFT_HOST', 'H.C. Earwicker'):
redshift.execute('SELECT 1')
assert 'could not translate host name "H.C. Earwicker" to address' in caplog.text
