def check_dump_integrity(hostname,
port,
dbname,
username,
password,
authdb,
dump_dir,
check_type=COUNT_COLLECTIONS):
# connect to the db
mongo = pymongo.MongoClient(hostname, int(port))
if username and password:
mongo[authdb].authenticate(username, password)
db = mongo[dbname]
# Get a list of relevant collections from the database
db_collections = db.collection_names()
db_relevant_collections = [
match.group(1) for coll in db_collections
for match in [DB_RELEVANT_COLLECTIONS_REGEX.search(coll)] if match
]
db_relevant_collections.sort()
db_relevant_collections.sort(key=len)
#Get a list of relevant collections from the dump
dump_exists = True
dump_path = dump_dir + dbname if dump_dir[
-1:] == '/' else dump_dir + '/' + dbname
dump_collections = []
try:
dump_collections = [file for file in os.listdir(dump_path)]
except OSError as e:
if e.errno == 2:
dump_exists = False
else:
logging.error(e)
print(e)
dump_relevant_collections = [
match.group(1) for coll in dump_collections
for match in [DUMP_RELEVANT_COLLECTIONS_REGEX.search(coll)] if match
]
dump_relevant_collections.sort()
dump_relevant_collections.sort(key=len)
#Find out format of dump (.json or .bson)
dump_format = 'bson'
if dump_relevant_collections[0].split('.', 1)[1] == 'json':
dump_format = 'json'
#CHECK NUMBER OF COLLECTIONS
if check_type == COUNT_COLLECTIONS:
num_collections_in_db = len(db_relevant_collections)
num_collections_dumped = len(dump_relevant_collections)
#Print integrity of number of collections in dump
log_output = 'DUMP FOR {} '.format(dbname)
if not dump_exists:
log_output += 'DOES NOT EXIST. '
elif num_collections_dumped < num_collections_in_db:
log_output += 'IS MISSING COLLECTIONS. '
elif num_collections_dumped > num_collections_in_db:
log_output += 'HAS TOO MANY COLLECTIONS. '
else:
log_output += 'IS OK ON COLLECTIONS. '
log_output += 'Number of collections in database: {}, Number of collections dumped: {}'.format(
num_collections_in_db, num_collections_dumped)
logging.info('\n' + log_output)
print('\n' + log_output)
#Print list of any collections missing from dump
if dump_exists and (num_collections_dumped < num_collections_in_db):
dump_relevant_collections_split = [
dump_coll.split('.' + dump_format, 1)[0]
for dump_coll in dump_relevant_collections
]
missing_collections = [
coll for coll in db_relevant_collections
if coll not in dump_relevant_collections_split
]
missing_collections.sort()
missing_collections.sort(key=len)
logging.info('\n' +
'Missing Collections: {}'.format(missing_collections))
print('\n' + 'Missing Collections: {}'.format(missing_collections))
#CHECK NUMBER OF DOCUMENTS
elif check_type == COUNT_DOCUMENTS:
logger.info(
'\n' +
'Counting number of documents in {} database'.format(dbname))
print('\n' +
'Counting number of documents in {} database'.format(dbname))
total_documents_in_db = 0
db_collection_doc_counts = {}
#Sum total documents in db
for coll in db_relevant_collections:
num_docs_in_coll = db[coll].count()
total_documents_in_db += num_docs_in_coll
#Save document count for db collection
db_collection_doc_counts[coll] = num_docs_in_coll
logging.info("Database {} {} document count: {}".format(
dbname, coll, num_docs_in_coll))
print("Database {} {} document count: {}".format(
dbname, coll, num_docs_in_coll))
logger.info('\n' +
'Counting number of documents in {} dump'.format(dbname))
print('\n' + 'Counting number of documents in {} dump'.format(dbname))
total_documents_dumped = 0
dump_collection_doc_counts = {}
#Sum up total number of documents in dump
for coll in dump_relevant_collections:
collection = SmappCollection(dump_format, dump_path + '/' + coll)
num_docs_in_coll = collection.count_tweets()
total_documents_dumped += num_docs_in_coll
#Save document count for dump collection
dump_collection_doc_counts[coll.split('.' + dump_format,
1)[0]] = num_docs_in_coll
#Calculate # and % missing documents for collection
num_docs_in_db_coll = db_collection_doc_counts[coll.split(
'.' + dump_format, 1)[0]]
num_docs_missing = num_docs_in_db_coll - num_docs_in_coll
percentage_docs_missing = 0
if num_docs_in_db_coll != 0:
percentage_docs_missing = (num_docs_missing /
num_docs_in_db_coll) * 100
logging.info("Dump {} {} document count: {}".format(
dbname, coll, num_docs_in_coll))
print("Dump {0} {1} document count: {2} (Missing {3}, {4:.2f}%)".
format(dbname, coll, num_docs_in_coll, num_docs_missing,
percentage_docs_missing))
# print("".format())
#Calculate # and % missing documents overall
total_docs_missing = total_documents_in_db - total_documents_dumped
percentage_total_docs_missing = 0
if total_documents_in_db != 0:
percentage_total_docs_missing = (total_docs_missing /
total_documents_in_db) * 100
#Print integrity of number of documents in dump
log_output = 'DUMP FOR {} '.format(dbname)
if not dump_exists:
log_output += 'DOES NOT EXIST. '
elif total_documents_dumped < total_documents_in_db:
log_output += 'IS MISSING DOCUMENTS. '
elif total_documents_dumped > total_documents_in_db:
log_output += 'HAS TOO MANY DOCUMENTS. '
else:
log_output += 'IS OK ON DOCUMENTS. '
log_output += 'Total documents in database: {0}, Total documents dumped: {1} (Missing {2}, {3:.2f}%)'.format(
total_documents_in_db, total_documents_dumped, total_docs_missing,
percentage_total_docs_missing)
logging.info('\n' + log_output)
print('\n' + log_output)
#Print list of any collections from dump missing documents
if dump_exists and (total_documents_dumped < total_documents_in_db):
collections_missing_docs = [
coll for coll, count in db_collection_doc_counts.items()
if (coll not in dump_collection_doc_counts
or dump_collection_doc_counts[coll] != count)
]
collections_missing_docs.sort()
collections_missing_docs.sort(key=len)
logging.info('\n' + 'Collections Missing Documents: {}'.format(
collections_missing_docs))
print('\n' + 'Collections Missing Documents: {}'.format(
collections_missing_docs))
def test_count_tweets(self):
file_path = '{}/{}'.format(os.path.dirname(os.path.realpath(__file__)),
config['bson']['valid'])
collection = SmappCollection('bson', file_path)
count = collection.count_tweets()
self.assertEqual(1187, count)
def check_dump_integrity(hostname, port, dbname, username, password, authdb, dump_dir, check_type = COUNT_COLLECTIONS):
# connect to the db
mongo = pymongo.MongoClient(hostname, int(port))
if username and password:
mongo[authdb].authenticate(username, password)
db = mongo[dbname]
# Get a list of relevant collections from the database
db_collections = db.collection_names()
db_relevant_collections = [match.group(1) for coll in db_collections for match in [DB_RELEVANT_COLLECTIONS_REGEX.search(coll)] if match]
db_relevant_collections.sort()
db_relevant_collections.sort(key=len)
#Get a list of relevant collections from the dump
dump_exists = True
dump_path = dump_dir + dbname if dump_dir[-1:] == '/' else dump_dir + '/' + dbname
dump_collections = []
try:
dump_collections = [file for file in os.listdir(dump_path)]
except OSError as e:
if e.errno == 2:
dump_exists = False
else:
logging.error(e)
print(e)
dump_relevant_collections = [match.group(1) for coll in dump_collections for match in [DUMP_RELEVANT_COLLECTIONS_REGEX.search(coll)] if match]
dump_relevant_collections.sort()
dump_relevant_collections.sort(key=len)
#Find out format of dump (.json or .bson)
dump_format = 'bson'
if dump_relevant_collections[0].split('.', 1)[1] == 'json':
dump_format = 'json'
#CHECK NUMBER OF COLLECTIONS
if check_type == COUNT_COLLECTIONS:
num_collections_in_db = len(db_relevant_collections)
num_collections_dumped = len(dump_relevant_collections)
#Print integrity of number of collections in dump
log_output = 'DUMP FOR {} '.format(dbname)
if not dump_exists:
log_output += 'DOES NOT EXIST. '
elif num_collections_dumped < num_collections_in_db:
log_output += 'IS MISSING COLLECTIONS. '
elif num_collections_dumped > num_collections_in_db:
log_output += 'HAS TOO MANY COLLECTIONS. '
else:
log_output += 'IS OK ON COLLECTIONS. '
log_output += 'Number of collections in database: {}, Number of collections dumped: {}'.format(num_collections_in_db, num_collections_dumped)
logging.info('\n' + log_output)
print('\n' + log_output)
#Print list of any collections missing from dump
if dump_exists and (num_collections_dumped < num_collections_in_db):
dump_relevant_collections_split = [dump_coll.split('.' + dump_format, 1)[0] for dump_coll in dump_relevant_collections]
missing_collections = [coll for coll in db_relevant_collections if coll not in dump_relevant_collections_split]
missing_collections.sort()
missing_collections.sort(key=len)
logging.info('\n' + 'Missing Collections: {}'.format(missing_collections))
print('\n' + 'Missing Collections: {}'.format(missing_collections))
#CHECK NUMBER OF DOCUMENTS
elif check_type == COUNT_DOCUMENTS:
logger.info('\n' + 'Counting number of documents in {} database'.format(dbname))
print('\n' + 'Counting number of documents in {} database'.format(dbname))
total_documents_in_db = 0
db_collection_doc_counts = {}
#Sum total documents in db
for coll in db_relevant_collections:
num_docs_in_coll = db[coll].count()
total_documents_in_db += num_docs_in_coll
#Save document count for db collection
db_collection_doc_counts[coll] = num_docs_in_coll
logging.info("Database {} {} document count: {}".format(dbname, coll, num_docs_in_coll))
print("Database {} {} document count: {}".format(dbname, coll, num_docs_in_coll))
logger.info('\n' + 'Counting number of documents in {} dump'.format(dbname))
print('\n' + 'Counting number of documents in {} dump'.format(dbname))
total_documents_dumped = 0
dump_collection_doc_counts = {}
#Sum up total number of documents in dump
for coll in dump_relevant_collections:
collection = SmappCollection(dump_format, dump_path + '/' + coll)
num_docs_in_coll = collection.count_tweets()
total_documents_dumped += num_docs_in_coll
#Save document count for dump collection
dump_collection_doc_counts[coll.split('.' + dump_format, 1)[0]] = num_docs_in_coll
#Calculate # and % missing documents for collection
num_docs_in_db_coll = db_collection_doc_counts[coll.split('.' + dump_format, 1)[0]]
num_docs_missing = num_docs_in_db_coll - num_docs_in_coll
percentage_docs_missing = 0
if num_docs_in_db_coll != 0:
percentage_docs_missing = (num_docs_missing/num_docs_in_db_coll) * 100
logging.info("Dump {} {} document count: {}".format(dbname, coll, num_docs_in_coll))
print("Dump {0} {1} document count: {2} (Missing {3}, {4:.2f}%)".format(dbname, coll, num_docs_in_coll, num_docs_missing, percentage_docs_missing))
# print("".format())
#Calculate # and % missing documents overall
total_docs_missing = total_documents_in_db - total_documents_dumped
percentage_total_docs_missing = 0
if total_documents_in_db != 0:
percentage_total_docs_missing = (total_docs_missing/total_documents_in_db) * 100
#Print integrity of number of documents in dump
log_output = 'DUMP FOR {} '.format(dbname)
if not dump_exists:
log_output += 'DOES NOT EXIST. '
elif total_documents_dumped < total_documents_in_db:
log_output += 'IS MISSING DOCUMENTS. '
elif total_documents_dumped > total_documents_in_db:
log_output += 'HAS TOO MANY DOCUMENTS. '
else:
log_output += 'IS OK ON DOCUMENTS. '
log_output += 'Total documents in database: {0}, Total documents dumped: {1} (Missing {2}, {3:.2f}%)'.format(total_documents_in_db, total_documents_dumped, total_docs_missing, percentage_total_docs_missing)
logging.info('\n' + log_output)
print('\n' + log_output)
#Print list of any collections from dump missing documents
if dump_exists and (total_documents_dumped < total_documents_in_db):
collections_missing_docs = [coll for coll, count in db_collection_doc_counts.items() if (coll not in dump_collection_doc_counts or dump_collection_doc_counts[coll] != count)]
collections_missing_docs.sort()
collections_missing_docs.sort(key=len)
logging.info('\n' + 'Collections Missing Documents: {}'.format(collections_missing_docs))
print('\n' + 'Collections Missing Documents: {}'.format(collections_missing_docs))
