def response_scroller(api_client, logger, query, result_size):
"""
Use the api_client instance of GreyNoise SDK to fetch the query results and traverse through the results if the result set is too large.
"""
# This will keep the count of how many events are remaining to be sent to Splunk
remaining_chunk_size = result_size
completion_flag = False
scroll = None
while not completion_flag:
event_count = 0
size = None
# Avoid the extra call if expected number of events are already retrieved
if remaining_chunk_size == 0:
logger.debug("No GreyNoise query results remaining to be sent, completing the search...")
break
# Do not fetch bunch of results if user does not request so many results
# Fetch only required numbers of events to keep away if the requested size is less than 10,000
if remaining_chunk_size < 10000:
size = remaining_chunk_size
logger.debug("Size for the GNQL query is configured to {}".format(size))
api_response = api_client.query(query=query, size=size, scroll=scroll)
if api_response.get('count', None):
# If this is the last page of API response, the scroll will not be present
scroll = api_response.get('scroll', None)
api_data = api_response.get('data', [])
for ip_data in api_data:
if event_count == 0 and remaining_chunk_size == result_size:
yield event_generator.make_valid_event('query', ip_data, True)
else:
yield event_generator.make_valid_event('query', ip_data, False)
event_count = event_count + 1
if event_count == remaining_chunk_size:
completion_flag = True
break
remaining_chunk_size = remaining_chunk_size - event_count
logger.debug("Statistics: Remaining chunk size: {} : Events written:{}".format(remaining_chunk_size, event_count))
else:
message = api_response.get('message', '')
query = api_response.get('query', '')
logger.info("No results returned for GreyNoise query: {}, message: {}".format(str(query), str(message)))
event = {
'message': message,
'query': query
}
yield event_generator.make_invalid_event('query', event, True)
exit(1)
# If we are on the last page of the results, scroll will not be present.
if scroll is None:
logger.debug("Last page of the GreyNoise query results detected, completing the search...")
completion_flag = True
def event_filter(chunk_index, result, records_dict, ip_field, noise_events,
method):
api_results = result['response']
error_flag = True
# Before yielding events, make the ip lookup dict which will have the following format:
# {<ip-address>: <API response for that IP address>}
ip_lookup = {}
if result['message'] == 'ok':
error_flag = False
for event in api_results:
ip_lookup[event['ip']] = event
for record in records_dict[0]:
if error_flag:
# Exception has occured while fetching the noise statuses from API
if ip_field in record and record[ip_field] != '':
# These calls have been failed due to API failure, as this event have IP address value, considering them as noise
if noise_events:
event = {'ip': record[ip_field], 'error': api_results}
yield event_generator.make_invalid_event(
method, event, True, record)
else:
# Either the record is not having IP field or the value of the IP field is ''
# send the record as it is as it doesn't have any IP address, after appending all fields
# Considering this event as non-noisy
if not noise_events:
yield event_generator.make_invalid_event(
method, {}, True, record)
else:
# Successful execution of the API call
if ip_field in record and record[ip_field] != '':
# Check if the IP field is not an iterable to avoid any error while referencing ip in ip_lookup
if isinstance(
record[ip_field],
six.string_types) and record[ip_field] in ip_lookup:
if ip_lookup[record[ip_field]]['noise'] == noise_events:
yield event_generator.make_valid_event(
method, ip_lookup[record[ip_field]], True, record)
else:
# Meaning ip is either invalid or not returned by the API, which is case of `multi` method only
# Invalid IPs are considered as non-noise
if not noise_events:
event = {
'ip':
record[ip_field],
'error':
'IP address doesn\'t match the valid IP format'
}
yield event_generator.make_invalid_event(
method, event, True, record)
else:
if not noise_events:
# Either the record is not having IP field or the value of the IP field is ''
# send the record as it is as it doesn't have any IP address, after appending all fields
# Considering this event as non-noisy
yield event_generator.make_invalid_event(
method, {}, True, record)
def do_generate(self, api_key, logger):
ip_address = self.ip
try:
# Strip the spaces from the parameter value if given
if ip_address:
ip_address = ip_address.strip()
logger.info("Initiating to fetch RIOT status for ip: {}".format(
str(ip_address)))
# Opting default timout 60 seconds for the request
api_client = GreyNoise(api_key=api_key,
timeout=60,
integration_name="Splunk")
context_info = api_client.riot(ip_address)
logger.info(
"Successfully retrieved the RIOT status for ip={}".format(
str(ip_address)))
# Process the API response and send the context information of IP with extractions in the Splunk
results = event_generator.make_valid_event('ip', context_info,
True)
yield results
except ValueError:
logger.error(
"IP address: {} doesn\'t match the valid IP format".format(
str(ip_address)))
self.write_error("IP address doesn\'t match the valid IP format")
def do_generate(self, api_key, logger):
"""
Method to fetch the api response and process and send the response with extractions in the Splunk.
:param api_key: GreyNoise API Key.
:logger: logger object.
"""
ip_address = self.ip
try:
# Strip the spaces from the parameter value if given
if ip_address:
ip_address = ip_address.strip()
logger.info(
"Initiating to fetch context information for ip: {}".format(
str(ip_address)))
# Opting default timout 60 seconds for the request
api_client = GreyNoise(api_key=api_key,
timeout=60,
integration_name=INTEGRATION_NAME)
session_key = self._metadata.searchinfo.session_key
context_info = get_response_for_generating(session_key, api_client,
ip_address, 'ip',
logger)
logger.info(
"Successfully retrieved the context information for ip={}".
format(str(ip_address)))
# Process the API response and send the context information of IP with extractions in the Splunk
results = event_generator.make_valid_event('ip', context_info,
True)
yield results
except ValueError as e:
error_msg = str(e).split(":")
logger.error(e)
self.write_error(error_msg[0])
def transform(self, records):
"""Method that processes and yield event records to the Splunk events pipeline."""
ip_addresses = self.ip
ip_field = self.ip_field
api_key = ""
EVENTS_PER_CHUNK = 5000
THREADS = 3
USE_CACHE = False
logger = utility.setup_logger(
session_key=self._metadata.searchinfo.session_key,
log_context=self._metadata.searchinfo.command)
if ip_addresses and ip_field:
logger.error(
"Please use parameter ip to work gnquick as generating command or "
"use parameter ip_field to work gnquick as transforming command."
)
self.write_error(
"Please use parameter ip to work gnquick as generating command or "
"use parameter ip_field to work gnquick as transforming command"
)
exit(1)
try:
message = ''
api_key = utility.get_api_key(
self._metadata.searchinfo.session_key, logger=logger)
except APIKeyNotFoundError as e:
message = str(e)
except HTTPError as e:
message = str(e)
if message:
self.write_error(message)
logger.error(
"Error occured while retrieving API key, Error: {}".format(
message))
exit(1)
if ip_addresses and not ip_field:
# This peice of code will work as generating command and will not use the Splunk events.
# Splitting the ip_addresses by commas and stripping spaces from both the sides for each IP address
ip_addresses = [ip.strip() for ip in ip_addresses.split(',')]
logger.info("Started retrieving results")
try:
logger.debug(
"Initiating to fetch noise and RIOT status for IP address(es): {}"
.format(str(ip_addresses)))
api_client = GreyNoise(api_key=api_key,
timeout=120,
integration_name=INTEGRATION_NAME)
# CACHING START
cache_enabled, cache_client = utility.get_caching(
self._metadata.searchinfo.session_key, 'multi', logger)
if int(cache_enabled) == 1 and cache_client is not None:
cache_start = time.time()
ips_not_in_cache, ips_in_cache = utility.get_ips_not_in_cache(
cache_client, ip_addresses, logger)
try:
response = []
if len(ips_in_cache) >= 1:
response = cache_client.query_kv_store(
ips_in_cache)
if response is None:
logger.debug(
"KVStore is not ready. Skipping caching mechanism."
)
noise_status = api_client.quick(ip_addresses)
elif response == []:
noise_status = utility.fetch_response_from_api(
api_client.quick, cache_client, ip_addresses,
logger)
else:
noise_status = utility.fetch_response_from_api(
api_client.quick, cache_client,
ips_not_in_cache, logger)
noise_status.extend(response)
except Exception:
logger.debug(
"An exception occurred while fetching response from cache.\n{}"
.format(traceback.format_exc()))
logger.debug(
"Generating command with caching took {} seconds.".
format(time.time() - cache_start))
else:
# Opting timout 120 seconds for the requests
noise_status = api_client.quick(ip_addresses)
logger.info("Retrieved results successfully")
# CACHING END
# Process the API response and send the noise and RIOT status information of IP with extractions
# to the Splunk, Using this flag to handle the field extraction issue in custom commands
# Only the fields extracted from the first event of generated by custom command
# will be extracted from all events
first_record_flag = True
# Flag to indicate whether erroneous IPs are present
erroneous_ip_present = False
for ip in ip_addresses:
for sample in noise_status:
if ip == sample['ip']:
yield event_generator.make_valid_event(
'quick', sample, first_record_flag)
if first_record_flag:
first_record_flag = False
logger.debug(
"Fetched noise and RIOT status for ip={} from GreyNoise API"
.format(str(ip)))
break
else:
erroneous_ip_present = True
try:
validate_ip(ip, strict=True)
except ValueError as e:
error_msg = str(e).split(":")
logger.debug(
"Generating noise and RIOT status for ip={} manually"
.format(str(ip)))
event = {'ip': ip, 'error': error_msg[0]}
yield event_generator.make_invalid_event(
'quick', event, first_record_flag)
if first_record_flag:
first_record_flag = False
if erroneous_ip_present:
logger.warn(
"Value of one or more IP address(es) is either invalid or non-routable"
)
self.write_warning(
"Value of one or more IP address(es) passed to {command_name} "
"is either invalid or non-routable".format(
command_name=str(
self._metadata.searchinfo.command)))
except RateLimitError:
logger.error(
"Rate limit error occured while fetching the context information for ips={}"
.format(str(ip_addresses)))
self.write_error(
"The Rate Limit has been exceeded. Please contact the Administrator"
)
except RequestFailure as e:
response_code, response_message = e.args
if response_code == 401:
msg = "Unauthorized. Please check your API key."
else:
# Need to handle this, as splunklib is unable to handle the exception with
# (400, {'error': 'error_reason'}) format
msg = (
"The API call to the GreyNoise platform have been failed "
"with status_code: {} and error: {}").format(
response_code,
response_message['error'] if isinstance(
response_message, dict) else response_message)
logger.error("{}".format(str(msg)))
self.write_error(msg)
except ConnectionError:
logger.error(
"Error while connecting to the Server. Please check your connection and try again."
)
self.write_error(
"Error while connecting to the Server. Please check your connection and try again."
)
except RequestException:
logger.error(
"There was an ambiguous exception that occurred while handling your Request. Please try again."
)
self.write_error(
"There was an ambiguous exception that occurred while handling your Request. Please try again."
)
except Exception:
logger.error("Exception: {} ".format(
str(traceback.format_exc())))
self.write_error(
"Exception occured while fetching the noise and RIOT status of the IP address(es). "
"See greynoise_main.log for more details.")
elif ip_field:
# Enter the mechanism only when the Search is complete and all the events are available
if self.search_results_info and not self.metadata.preview:
try:
# Strip the spaces from the parameter value if given
ip_field = ip_field.strip()
# Validating the given parameter
try:
ip_field = validator.Fieldname(
option_name='ip_field').validate(ip_field)
except ValueError as e:
# Validator will throw ValueError with error message when the parameters are not proper
logger.error(str(e))
self.write_error(str(e))
exit(1)
# API key validation
if not self.api_validation_flag:
api_key_validation, message = utility.validate_api_key(
api_key, logger)
logger.debug(
"API validation status: {}, message: {}".format(
api_key_validation, str(message)))
self.api_validation_flag = True
if not api_key_validation:
logger.info(message)
self.write_error(message)
exit(1)
# This piece of code will work as transforming command and will use
# the Splunk ingested events and field which is specified in ip_field.
chunk_dict = event_generator.batch(records, ip_field,
EVENTS_PER_CHUNK,
logger)
# This means there are only 1000 or below IPs to call in the entire bunch of records
# Use one thread with single thread with caching mechanism enabled for the chunk
if len(chunk_dict) == 1:
logger.info(
"Less then 1000 distinct IPs are present, "
"optimizing the IP requests call to GreyNoise API..."
)
THREADS = 1
USE_CACHE = True
api_client = GreyNoise(api_key=api_key,
timeout=120,
use_cache=USE_CACHE,
integration_name=INTEGRATION_NAME)
# When no records found, batch will return {0:([],[])}
tot_time_start = time.time()
if len(list(chunk_dict.values())[0][0]) >= 1:
for event in event_generator.get_all_events(
self._metadata.searchinfo.session_key,
api_client,
'multi',
ip_field,
chunk_dict,
logger,
threads=THREADS):
yield event
else:
logger.info(
"No events found, please increase the search timespan to have more search results."
)
tot_time_end = time.time()
logger.debug(
"Total execution time => {}".format(tot_time_end -
tot_time_start))
except Exception:
logger.info(
"Exception occured while adding the noise and RIOT status to the events, Error: {}"
.format(traceback.format_exc()))
self.write_error(
"Exception occured while adding the noise and RIOT status of "
"the IP addresses to events. See greynoise_main.log for more details."
)
else:
logger.error(
"Please specify exactly one parameter from ip and ip_field with some value."
)
self.write_error(
"Please specify exactly one parameter from ip and ip_field with some value."
)
def transform(self, records):
"""Method that processes and yield event records to the Splunk events pipeline."""
ip_address = self.ip
ip_field = self.ip_field
api_key = ""
EVENTS_PER_CHUNK = 1
THREADS = 3
USE_CACHE = False
logger = utility.setup_logger(
session_key=self._metadata.searchinfo.session_key, log_context=self._metadata.searchinfo.command)
if ip_address and ip_field:
logger.error("Please use parameter ip to work gnriot as generating command or "
"use parameter ip_field to work gnriot as transforming command.")
self.write_error("Please use parameter ip to work gnriot as generating command or "
"use parameter ip_field to work gnriot as transforming command")
exit(1)
try:
message = ''
api_key = utility.get_api_key(self._metadata.searchinfo.session_key, logger=logger)
except APIKeyNotFoundError as e:
message = str(e)
except HTTPError as e:
message = str(e)
if message:
self.write_error(message)
logger.error("Error occured while retrieving API key, Error: {}".format(message))
exit(1)
if ip_address and not ip_field:
# This peice of code will work as generating command and will not use the Splunk events.
# Strip the spaces from the parameter value if given
ip_address = ip_address.strip()
logger.info("Started retrieving results")
try:
logger.debug("Initiating to fetch RIOT information for IP address: {}".format(str(ip_address)))
api_client = GreyNoise(api_key=api_key, timeout=120, integration_name=INTEGRATION_NAME)
# Opting timout 120 seconds for the requests
session_key = self._metadata.searchinfo.session_key
riot_information = utility.get_response_for_generating(
session_key, api_client, ip_address, 'greynoise_riot', logger)
logger.info("Retrieved results successfully")
# Process the API response and send the riot information of IP with extractions to the Splunk
yield event_generator.make_valid_event('riot', riot_information, True)
logger.debug("Fetched RIOT information for ip={} from GreyNoise API".format(str(ip_address)))
except ValueError as e:
error_msg = str(e).split(":")
logger.debug("Generating RIOT information for ip={} manually".format(str(ip_address)))
event = {
'ip': ip_address,
'error': error_msg[0]
}
yield event_generator.make_invalid_event('riot', event, True)
logger.warn(error_msg)
self.write_warning(
"Value of IP address passed to {command_name} is either invalid or non-routable".format(
command_name=str(self._metadata.searchinfo.command)))
except RateLimitError:
logger.error("Rate limit error occured while fetching the context information for ip={}".format(
str(ip_address)))
self.write_error("The Rate Limit has been exceeded. Please contact the Administrator")
except RequestFailure as e:
response_code, response_message = e.args
if response_code == 401:
msg = "Unauthorized. Please check your API key."
else:
# Need to handle this, as splunklib is unable to handle the exception with
# (400, {'error': 'error_reason'}) format
msg = ("The API call to the GreyNoise platform have been failed "
"with status_code: {} and error: {}").format(
response_code, response_message['error'] if isinstance(response_message, dict)
else response_message)
logger.error("{}".format(str(msg)))
self.write_error(msg)
except ConnectionError:
logger.error("Error while connecting to the Server. Please check your connection and try again.")
self.write_error("Error while connecting to the Server. Please check your connection and try again.")
except RequestException:
logger.error(
"There was an ambiguous exception that occurred while handling your Request. Please try again.")
self.write_error(
"There was an ambiguous exception that occurred while handling your Request. Please try again.")
except Exception:
logger.error("Exception: {} ".format(str(traceback.format_exc())))
self.write_error("Exception occured while fetching the RIOT information of the IP address. "
"See greynoise_main.log for more details.")
elif ip_field:
logger.info("Started retrieving RIOT information for the IP addresses present in field: {}".format(
str(ip_field)))
# Enter the mechanism only when the Search is complete and all the events are available
if self.search_results_info and not self.metadata.preview:
try:
# Strip the spaces from the parameter value if given
ip_field = ip_field.strip()
# Validating the given parameter
try:
ip_field = validator.Fieldname(option_name='ip_field').validate(ip_field)
except ValueError as e:
# Validator will throw ValueError with error message when the parameters are not proper
logger.error(str(e))
self.write_error(str(e))
exit(1)
# API key validation
if not self.api_validation_flag:
api_key_validation, message = utility.validate_api_key(api_key, logger)
logger.debug("API validation status: {}, message: {}".format(api_key_validation, str(message)))
self.api_validation_flag = True
if not api_key_validation:
logger.info(message)
self.write_error(message)
exit(1)
# This piece of code will work as transforming command and will use
# the Splunk ingested events and field which is specified in ip_field.
# divide the records in the form of dict of tuples having chunk_index as key
# {<index>: (<records>, <All the ips in records>)}
chunk_dict = event_generator.batch(
records, ip_field, EVENTS_PER_CHUNK, logger, optimize_requests=False)
logger.debug("Successfully divided events into chunks")
# This means there are only 1000 or below IPs to call in the entire bunch of records
# Use one thread with single thread with caching mechanism enabled for the chunk
if len(chunk_dict) == 1:
logger.debug("Less then 1000 distinct IPs are present, "
"optimizing the IP requests call to GreyNoise API...")
THREADS = 1
USE_CACHE = True
api_client = GreyNoise(
api_key=api_key, timeout=120, use_cache=USE_CACHE, integration_name=INTEGRATION_NAME)
# When no records found, batch will return {0:([],[])}
if len(chunk_dict) > 0:
for event in event_generator.get_all_events(
self._metadata.searchinfo.session_key, api_client, 'greynoise_riot', ip_field,
chunk_dict, logger, threads=THREADS):
yield event
logger.info("Successfully sent all the results to the Splunk")
else:
logger.info("No events found, please increase the search timespan to have more search results.")
except Exception:
logger.info(
"Exception occured while adding the RIOT information to the events, Error: {}".format(
traceback.format_exc()))
self.write_error("Exception occured while adding the RIOT information of the IP addresses "
"to events. See greynoise_main.log for more details.")
else:
logger.error("Please specify exactly one parameter from ip and ip_field with some value.")
self.write_error("Please specify exactly one parameter from ip and ip_field with some value.")