openstack/zaqar
Code Issues Proposed changes

Report claim and delete latency separately in the benchmark tool

Browse Source 
The consumer worker actually performs two different operations. First
it claims a batch of messages, then it deletes each one before getting
the next batch.

Therefore, in order to better compare the latency of enqueuing a message
against other operations, this patch changes the reporting to provide
the individual latency of those other operations.

Change-Id: Iff7f6054904a03dc240d3f7a2a0529bd990f973b
This commit is contained in:
kgriffs 2014-08-21 11:11:34 -05:00 committed by Flavio Percoco
parent 04fa909282
commit 6c09339a2a
3 changed files with 51 additions and 33 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
README.rst
View File

@ -132,19 +132,20 @@ Verbose output looks similar to the following::
Consumer
========
duration_sec: 10.1
ms_per_req: 77.1
total_reqs: 160.0
successful_reqs: 160.0
reqs_per_sec: 15.8
duration_sec: 10.2
ms_per_claim: 37.6
ms_per_delete: 11.8
reqs_per_sec: 82.0
successful_reqs: 833.0
total_reqs: 833.0
Producer
========
duration_sec: 10.2
ms_per_req: 4.6
total_reqs: 8866.0
successful_reqs: 8866.0
reqs_per_sec: 870.5
ms_per_req: 3.8
reqs_per_sec: 1033.6
successful_reqs: 10523.0
total_reqs: 10523.0
.. _`OpenStack` : http://openstack.org/

9
zaqar/bench/conductor.py
View File

@ -40,11 +40,16 @@ def main():
if conf.verbose:
print()
for name, stat in stats.items():
for name, stats_group in stats.items():
print(name.capitalize())
print('=' * len(name))
print("\n".join("{}: {:.1f}".format(*it) for it in stat.items()))
values = sorted(stats_group.items(), key=lambda v: v[0])
formatted_vals = ["{}: {:.1f}".format(*v) for v in values]
print("\n".join(formatted_vals))
print('') # Blank line
else:
stats['params'] = {
'producer': {

56
zaqar/bench/consumer.py
View File

@ -40,7 +40,8 @@ def claim_delete(stats, test_duration, ttl, grace, limit):
cli = client.Client(conf.server_url)
queue = cli.queue(conf.queue_prefix + '1')
end = time.time() + test_duration
total_elapsed = 0
claim_total_elapsed = 0
delete_total_elapsed = 0
total_requests = 0
claim_total_requests = 0
delete_total_requests = 0
@ -54,7 +55,7 @@ def claim_delete(stats, test_duration, ttl, grace, limit):
sys.stderr.write("Could not claim messages : {0}\n".format(ex))
else:
total_elapsed += marktime.stop('claim_message').seconds
claim_total_elapsed += marktime.stop('claim_message').seconds
claim_total_requests += 1
try:
@ -62,10 +63,10 @@ def claim_delete(stats, test_duration, ttl, grace, limit):
for msg in claim:
# TODO(TheSriram): Simulate actual work before deletion
delete_total_requests += 1
msg.delete()
total_elapsed += marktime.stop('delete_message').seconds
delete_total_requests += 1
delete_total_elapsed += marktime.stop('delete_message').seconds
except TransportError as ex:
sys.stderr.write("Could not delete messages: {0}\n".format(ex))
@ -75,10 +76,13 @@ def claim_delete(stats, test_duration, ttl, grace, limit):
finally:
total_requests += 1
stats.put({'total_requests': total_requests,
'claim_total_requests': claim_total_requests,
'delete_total_requests': delete_total_requests,
'total_elapsed': total_elapsed})
stats.put({
'total_requests': total_requests,
'claim_total_requests': claim_total_requests,
'delete_total_requests': delete_total_requests,
'claim_total_elapsed': claim_total_elapsed,
'delete_total_elapsed': delete_total_elapsed,
})
def load_generator(stats, num_workers, test_duration, url, ttl, grace, limit):
@ -91,19 +95,21 @@ def load_generator(stats, num_workers, test_duration, url, ttl, grace, limit):
def crunch(stats):
total_requests = 0
total_latency = 0.0
claim_total_elapsed = 0.0
delete_total_elapsed = 0.0
claim_total_requests = 0
delete_total_requests = 0
while not stats.empty():
entry = stats.get_nowait()
total_requests += entry['total_requests']
total_latency += entry['total_elapsed']
claim_total_elapsed += entry['claim_total_elapsed']
delete_total_elapsed += entry['delete_total_elapsed']
claim_total_requests += entry['claim_total_requests']
delete_total_requests += entry['delete_total_requests']
return (total_requests, total_latency, claim_total_requests,
delete_total_requests)
return (total_requests, claim_total_elapsed, delete_total_elapsed,
claim_total_requests, delete_total_requests)
def run(upstream_queue):
@ -131,24 +137,30 @@ def run(upstream_queue):
for each_proc in procs:
each_proc.join()
(total_requests, total_latency, claim_total_requests,
delete_total_requests) = crunch(stats)
(total_requests, claim_total_elapsed, delete_total_elapsed,
claim_total_requests, delete_total_requests) = crunch(stats)
successful_requests = claim_total_requests + delete_total_requests
duration = time.time() - start
throughput = successful_requests / duration
latency = 1000 * total_latency / successful_requests
claim_latency = 1000 * claim_total_elapsed / claim_total_requests
delete_latency = 1000 * delete_total_elapsed / delete_total_requests
else:
duration = 0
total_requests = 0
successful_requests = 0
throughput = 0
latency = 0
claim_latency = 0
delete_latency = 0
upstream_queue.put({'consumer': {
'duration_sec': duration,
'total_reqs': total_requests,
'successful_reqs': successful_requests,
'reqs_per_sec': throughput,
'ms_per_req': latency}})
upstream_queue.put({
'consumer': {
'duration_sec': duration,
'total_reqs': total_requests,
'successful_reqs': successful_requests,
'reqs_per_sec': throughput,
'ms_per_claim': claim_latency,
'ms_per_delete': delete_latency,
}
})