def main():
# collect arguments
parser = argparse.ArgumentParser(description="run script", add_help=False)
parser.add_argument(
'-v', '--verbose', action='store_const', dest='loglevel', const=logging.DEBUG,
default=logging.INFO, help='enables verbose wrapper debugging info')
parser.add_argument(
'-t', '--tool', help='Provide tool name', required=True)
parser.add_argument(
'--run-id', help='Run ID to unify benchmark results in ES',
nargs='?', default="NA"
)
index_args, unknown = parser.parse_known_args()
index_args.index_results = False
index_args.prefix = "snafu-%s" % index_args.tool
setup_loggers("snafu", index_args.loglevel)
log_level_str = 'DEBUG' if index_args.loglevel == logging.DEBUG else 'INFO'
logger.info("logging level is %s" % log_level_str)
# set up a standard format for time
FMT = '%Y-%m-%dT%H:%M:%SGMT'
# instantiate elasticsearch instance and check connection
es_settings = {}
es_settings["server"] = os.getenv("es")
es_settings["verify_cert"] = os.getenv("es_verify_cert", "true").lower()
if es_settings["server"] and ":443" in es_settings["server"]:
es_settings["verify_cert"] = "false"
if es_settings["server"]:
index_args.prefix = os.getenv("es_index", "")
logger.info("Using elasticsearch server with host: %s" % es_settings['server'])
logger.info("Using index prefix for ES: %s" % index_args.prefix)
index_args.index_results = True
try:
if es_settings["verify_cert"] == "false":
logger.info("Turning off TLS certificate verification")
urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)
ssl_ctx = ssl.create_default_context()
ssl_ctx.check_hostname = False
ssl_ctx.verify_mode = ssl.CERT_NONE
es = elasticsearch.Elasticsearch([es_settings["server"]], send_get_body_as='POST',
ssl_context=ssl_ctx, use_ssl=True)
else:
es = elasticsearch.Elasticsearch([es_settings["server"]], send_get_body_as='POST')
logger.info("Connected to the elasticsearch cluster with info as follows:")
logger.info(json.dumps(es.info(), indent=4))
except Exception as e:
logger.warn("Elasticsearch connection caused an exception: %s" % e)
index_args.index_results = False
index_args.document_size_capacity_bytes = 0
# call py es bulk using a process generator to feed it ES documents
if index_args.index_results:
parallel_setting = strtobool(os.environ.get('parallel', "false"))
res_beg, res_end, res_suc, res_dup, res_fail, res_retry = streaming_bulk(es,
process_generator(
index_args,
parser),
parallel_setting)
logger.info(
"Indexed results - %s success, %s duplicates, %s failures, with %s retries." % (
res_suc,
res_dup,
res_fail,
res_retry))
start_t = time.strftime('%Y-%m-%dT%H:%M:%SGMT', time.gmtime(res_beg))
end_t = time.strftime('%Y-%m-%dT%H:%M:%SGMT', time.gmtime(res_end))
else:
start_t = time.strftime('%Y-%m-%dT%H:%M:%SGMT', time.gmtime())
# need to loop through generator and pass on all yields
# this will execute all jobs without elasticsearch
for i in process_generator(index_args, parser):
pass
end_t = time.strftime('%Y-%m-%dT%H:%M:%SGMT', time.gmtime())
start_t = datetime.datetime.strptime(start_t, FMT)
end_t = datetime.datetime.strptime(end_t, FMT)
# get time delta for indexing run
tdelta = end_t - start_t
total_capacity_bytes = index_args.document_size_capacity_bytes
logger.info("Duration of execution - %s, with total size of %s bytes" % (tdelta, total_capacity_bytes))
def main():
# collect arguments
parser = configargparse.get_argument_parser(
description="Run benchmark-wrapper and export results.",
add_config_file_help=True,
add_env_var_help=True,
default_config_files=["./snafu.yml"],
ignore_unknown_config_file_keys=False,
add_help=False,
)
parser.add_argument(
"-v",
"--verbose",
action="store_const",
dest="loglevel",
const=logging.DEBUG,
default=logging.INFO,
help="enables verbose wrapper debugging info",
)
parser.add_argument("--config",
help="Config file to load",
is_config_file=True)
parser.add_argument("-t",
"--tool",
help="Provide tool name",
required=True)
parser.add_argument("--run-id",
help="Run ID to unify benchmark results in ES",
nargs="?",
default="NA")
parser.add_argument("--archive-file",
help="Archive file that will be indexed into ES")
parser.add_argument(
"--create-archive",
action="store_const",
dest="createarchive",
const=True,
default=False,
help="enables creation of archive file",
)
index_args, unknown = parser.parse_known_args()
index_args.index_results = False
index_args.prefix = "snafu-%s" % index_args.tool
setup_loggers("snafu", index_args.loglevel)
log_level_str = "DEBUG" if index_args.loglevel == logging.DEBUG else "INFO"
logger.info("logging level is %s" % log_level_str)
# Log loaded benchmarks
show_db_tb = index_args.loglevel == logging.DEBUG
benchmarks.DETECTED_BENCHMARKS.log(logger=logger,
level=logging.INFO,
show_tb=show_db_tb)
# set up a standard format for time
FMT = "%Y-%m-%dT%H:%M:%SGMT"
# instantiate elasticsearch instance and check connection
es_settings = {}
es_settings["server"] = os.getenv("es")
es_settings["verify_cert"] = os.getenv("es_verify_cert", "true").lower()
if es_settings["server"] and ":443" in es_settings["server"]:
es_settings["verify_cert"] = "false"
if es_settings["server"]:
index_args.prefix = os.getenv("es_index", index_args.prefix)
logger.info("Using elasticsearch server with host: %s" %
es_settings["server"])
logger.info("Using index prefix for ES: %s" % index_args.prefix)
index_args.index_results = True
try:
if es_settings["verify_cert"] == "false":
logger.info("Turning off TLS certificate verification")
urllib3.disable_warnings(
urllib3.exceptions.InsecureRequestWarning)
ssl_ctx = ssl.create_default_context()
ssl_ctx.check_hostname = False
ssl_ctx.verify_mode = ssl.CERT_NONE
es = elasticsearch.Elasticsearch([es_settings["server"]],
send_get_body_as="POST",
ssl_context=ssl_ctx,
use_ssl=False)
else:
es = elasticsearch.Elasticsearch([es_settings["server"]],
send_get_body_as="POST")
logger.info(
"Connected to the elasticsearch cluster with info as follows:")
logger.info(json.dumps(es.info(), indent=4))
except Exception as e:
logger.warn("Elasticsearch connection caused an exception: %s" % e)
index_args.index_results = False
index_args.document_size_capacity_bytes = 0
# call py es bulk using a process generator to feed it ES documents
if index_args.index_results:
parallel_setting = strtobool(os.environ.get("parallel", "false"))
if "archive" in index_args.tool:
if index_args.archive_file:
# if processing a archive file use the process archive file function
res_beg, res_end, res_suc, res_dup, res_fail, res_retry = streaming_bulk(
es, process_archive_file(index_args), parallel_setting)
else:
logger.error(
"Attempted to index archive without specifying a file, use --archive-file=<file>"
)
exit(1)
else:
# else run a test and process new result documents
res_beg, res_end, res_suc, res_dup, res_fail, res_retry = streaming_bulk(
es, process_generator(index_args, parser), parallel_setting)
logger.info(
"Indexed results - %s success, %s duplicates, %s failures, with %s retries."
% (res_suc, res_dup, res_fail, res_retry))
start_t = time.strftime("%Y-%m-%dT%H:%M:%SGMT", time.gmtime(res_beg))
end_t = time.strftime("%Y-%m-%dT%H:%M:%SGMT", time.gmtime(res_end))
else:
logger.info("Not connected to Elasticsearch")
start_t = time.strftime("%Y-%m-%dT%H:%M:%SGMT", time.gmtime())
# need to loop through generator and pass on all yields
# this will execute all jobs without elasticsearch
if "archive" in index_args.tool:
if index_args.archive_file:
logger.info(
"Processing archive file, but not indexing results...")
for es_friendly_doc in process_archive_file(index_args):
pass
else:
logger.error(
"Attempted to index archive without specifying a file, use --archive-file=<file>"
)
exit(1)
else:
for i in process_generator(index_args, parser):
pass
end_t = time.strftime("%Y-%m-%dT%H:%M:%SGMT", time.gmtime())
start_t = datetime.datetime.strptime(start_t, FMT)
end_t = datetime.datetime.strptime(end_t, FMT)
# get time delta for indexing run
tdelta = end_t - start_t
total_capacity_bytes = index_args.document_size_capacity_bytes
logger.info(
"Duration of execution - {}, with total size of {} bytes".format(
tdelta, total_capacity_bytes))