def hypercube_shuffle():
"""
Only hypercube shuffle
"""
json_query_files = [
"q1_hc_1.json", "q1_hc_2.json", "q1_hc_3.json", "q1_hc_4.json",
"q1_hc_5.json", "q1_hc_6.json", "q1_hc_7.json", "q1_hc_8.json"]
for query_file_path in json_query_files:
with open(query_file_path) as jf:
query = json.load(jf)
print "executing query {}".format(query_file_path)
client.execute_json(query)
def hypercube_shuffle():
"""
Only hypercube shuffle
"""
json_query_files = [
"q1_hc_1.json", "q1_hc_2.json", "q1_hc_3.json", "q1_hc_4.json",
"q1_hc_5.json", "q1_hc_6.json", "q1_hc_7.json", "q1_hc_8.json"
]
for query_file_path in json_query_files:
with open(query_file_path) as jf:
query = json.load(jf)
print "executing query {}".format(query_file_path)
client.execute_json(query)
def execute_original_query():
with open("q8_lfj_local.json") as f:
json_query = json.load(f)
new_order = (0, 1, 2, 3, 4, 5)
print "executing {}".format(new_order)
new_query, a, b = attrorder.change_order(json_query, new_order)
# new_query["profilingMode"] = 'QUERY'
result, status = client.execute_json(new_query)
def execute_original_query():
with open("q8_lfj_local.json") as f:
json_query = json.load(f)
new_order = (0, 1, 2, 3, 4, 5)
print "executing {}".format(new_order)
new_query, a, b = attrorder.change_order(json_query, new_order)
# new_query["profilingMode"] = 'QUERY'
result, status = client.execute_json(new_query)
def exp_q8():
with open("q8_lfj_local.json") as f:
json_query = json.load(f)
for order in sampled_orders[18:]:
# print "executing {}".format(order)
new_query, a, b = attrorder.change_order(json_query, order)
result, status = client.execute_json(new_query)
if result == 'success':
output = (order, attrorder.q8(order), float(status["elapsedNanos"]) / client.NANO_IN_ONE_SEC)
print ",".join([str(i) for i in output])
else:
raise ValueError("Error when executing {}".format(order))
def exp_q8():
with open("q8_lfj_local.json") as f:
json_query = json.load(f)
for order in sampled_orders[18:]:
# print "executing {}".format(order)
new_query, a, b = attrorder.change_order(json_query, order)
result, status = client.execute_json(new_query)
if result == 'success':
output = (order, attrorder.q8(order),
float(status["elapsedNanos"]) / client.NANO_IN_ONE_SEC)
print ",".join([str(i) for i in output])
else:
raise ValueError("Error when executing {}".format(order))
def random_order_lfj(query_file, log_file_name, orders):
with open(query_file) as f:
json_query = json.load(f)
with open(log_file_name, "wb") as f:
csvwriter = csv.writer(f)
csvwriter.writerow(["query_id", "order", "time", "success or fail"])
for order in orders:
json_query, a, b = attrorder.change_order(json_query, order)
print "execute leapfrog join with order {}".format(order)
result, status = client.execute_json(json_query)
# log experiment result
if result == 'success':
print "success"
time = float(status["elapsedNanos"]) / client.NANO_IN_ONE_SEC
csvwriter.writerow([
status["queryId"], str(order), time, "success"])
else:
print "error"
csvwriter.writerow(["N.A.", str(order), "N.A.", "fail"])
def random_order_lfj(query_file, log_file_name, orders):
with open(query_file) as f:
json_query = json.load(f)
with open(log_file_name, "wb") as f:
csvwriter = csv.writer(f)
csvwriter.writerow(["query_id", "order", "time", "success or fail"])
for order in orders:
json_query, a, b = attrorder.change_order(json_query, order)
print "execute leapfrog join with order {}".format(order)
result, status = client.execute_json(json_query)
# log experiment result
if result == 'success':
print "success"
time = float(status["elapsedNanos"]) / client.NANO_IN_ONE_SEC
csvwriter.writerow(
[status["queryId"],
str(order), time, "success"])
else:
print "error"
csvwriter.writerow(["N.A.", str(order), "N.A.", "fail"])
def scl_exp(log_file_name):
queries = [
"triangle_worker_32.json",
"triangle_worker_16.json",
"triangle_worker_2.json",
"triangle_worker_4.json",
"triangle_worker_8.json"
]
with open(log_file_name, "wb") as f:
csvwriter = csv.writer(f)
csvwriter.writerow(["query_id", "time", "success or fail"])
for query in queries:
with open(query) as qf:
json_query = json.load(qf)
if query == "triangle_worker_2.json":
workers = [1, 2]
elif query == "triangle_worker_4.json":
workers = list(range(1, 5))
elif query == "triangle_worker_8.json":
workers = list(range(1, 9))
elif query == "triangle_worker_16.json":
workers = list(range(1, 17))
elif query == "triangle_worker_32.json":
workers = list(range(1, 33))
for fragment in json_query["plan"]["fragments"]:
fragment["workers"] = workers
result, status = client.execute_json(json_query)
# log experiment result
if result == 'success':
print "success"
time = float(status["elapsedNanos"]) / client.NANO_IN_ONE_SEC
csvwriter.writerow([
status["queryId"], time, "success"])
else:
print "error"
csvwriter.writerow(["N.A.", "N.A.", "fail"])