Commit 0586cca5 by Michael Roytman

Refactor previous script for readability and commenting

parent 8e202d61
......@@ -3,66 +3,92 @@ import os
import pathlib2
import itertools
import sys
import argparse
class ContainerBalancer:
def __init__(self):
self.load_repo_path()
TRAVIS_BUILD_DIR = os.environ.get("TRAVIS_BUILD_DIR")
CONFIG_FILE_PATH = pathlib2.Path(TRAVIS_BUILD_DIR, "util", "parsefiles_config.yml")
def load_repo_path(self):
"""Loads the path for the configuration repository from TRAVIS_BUILD_DIR environment variable."""
def pack_containers(containers, num_shards):
"""
Determines an approximation of the optimal way to pack the containers into a given number of shards so as to
equalize the execution time amongst the shards.
if os.environ.get("TRAVIS_BUILD_DIR"):
self.repo_path = os.environ.get("TRAVIS_BUILD_DIR")
else:
raise EnvironmentError("TRAVIS_BUILD_DIR environment variable is not set.")
Input:
containers: A set of Docker containers
num_shards: A number of shards amongst which to distribute the Docker containers
"""
def pack_containers(self, containers):
# open config file containing container weights
config_file_path = pathlib2.Path(CONFIG_FILE_PATH)
num_shards = 3
with (config_file_path.open(mode='r')) as file:
try:
config = yaml.load(file)
except yaml.YAMLError, exc:
LOGGER.warning("error in configuration file: %s" % str(exc))
sys.exit(1)
config_file_path = pathlib2.Path(self.repo_path, "util", "parsefiles_config.yml")
# get container weights
weights = config.get("weights")
with config_file_path.open() as config_file:
config = yaml.load(config_file)
# convert all containers in config file to a list of tuples (<container>, <weight>)
weights_list = [x.items() for x in weights]
weights_list = list(itertools.chain.from_iterable(weights_list))
weights = config.get("weights")
# performs intersection between weighted containers and input containers
used_containers = [x for x in weights_list if x[0] in containers]
weights_list = [x.items() for x in weights]
weights_list = list(itertools.chain.from_iterable(weights_list))
# sorts used containers in descending order on the weight
sorted_containers = sorted(used_containers, key = lambda x: x[1], reverse=True)
used_containers = [x for x in weights_list if x[0] in containers]
shards = []
sorted_containers = sorted(used_containers, key = lambda x: x[1], reverse=True)
# for the number of shards
for i in range(0, num_shards):
# initialize initial dict
shards.append({"containers": [], "sum": 0})
shards = []
# for each container
for container in sorted_containers:
# find the shard with the current minimum execution time
shard = min(shards, key = lambda x: x["sum"])
for i in range(0, num_shards):
shards.append({"containers": [], "sum": 0})
# add the current container to the shard
shard["containers"].append(container)
for container in sorted_containers:
# shard with minimum execution time
shard = min(shards, key = lambda x: x["sum"])
# add the current container's weight to the shard's total expected execution time
shard["sum"] += container[1]
shard["containers"].append(container)
shard["sum"] += container[1]
return shards
return shards
def arg_parse():
parser = argparse.ArgumentParser(description = 'Given a list of containers as input and a number of shards, '
'finds an approximation of the optimal distribution of the containers over the shards, provided a set of hard-coded weights '
'in parsefiles_config.yml.')
parser.add_argument('num_shards', type = int, help = "the number of shards amongst which to distribute Docker builds")
return parser.parse_args()
if __name__ == '__main__':
balancer = ContainerBalancer()
args = arg_parse()
containers = []
# get containers from standard in
for line in sys.stdin:
# strip whitespace and brackets
line = line.strip()
line = line.strip("[]")
items = line.split()
containers.extend(items)
shards = balancer.pack_containers(containers)
# find optimal packing of the containers amongst shards
shards = pack_containers(containers, args.num_shards)
# print space separate list of containers for each shard
for shard in shards:
middle = " "
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment