def main():
# Parse up the command line arguments.
parser = argparse.ArgumentParser(description='Update haproxy to use all instances running in a security group.')
parser.add_argument('--access-key', required=True)
parser.add_argument('--secret-key', required=True)
parser.add_argument('--eip',
help='The Elastic IP to bind to when VIP seems unhealthy.')
parser.add_argument('--health-check-url',
help='The URL to check. Assigns EIP to self if health check fails.')
args = parser.parse_args()
# Do a health check on the url if specified.
try:
if args.health_check_url and args.eip:
logging.info('Performing health check.')
try:
logging.info('Checking %s' % args.health_check_url)
response = urllib2.urlopen(args.health_check_url)
logging.info('Response: %s' % response.read())
except:
# Assign the EIP to self.
logging.warn('Health check failed. Assigning %s to self.' % args.eip)
steal_elastic_ip(access_key=args.access_key,
secret_key=args.secret_key,
ip=args.eip )
except:
pass
def main():
# Parse up the command line arguments.
parser = argparse.ArgumentParser(
description=
'Update haproxy to use all instances running in a security group.')
parser.add_argument('--access-key', required=True)
parser.add_argument('--secret-key', required=True)
parser.add_argument(
'--eip', help='The Elastic IP to bind to when VIP seems unhealthy.')
parser.add_argument(
'--health-check-url',
help='The URL to check. Assigns EIP to self if health check fails.')
args = parser.parse_args()
# Do a health check on the url if specified.
try:
if args.health_check_url and args.eip:
logging.info('Performing health check.')
try:
logging.info('Checking %s' % args.health_check_url)
response = urllib2.urlopen(args.health_check_url)
logging.info('Response: %s' % response.read())
except:
# Assign the EIP to self.
logging.warn('Health check failed. Assigning %s to self.' %
args.eip)
steal_elastic_ip(access_key=args.access_key,
secret_key=args.secret_key,
ip=args.eip)
except:
pass
def main():
# Parse up the command line arguments.
parser = argparse.ArgumentParser(description='Update haproxy to use all instances running in a security group.')
parser.add_argument('--security-group', required=False, nargs='+', type=str)
parser.add_argument('--access-key', required=True)
parser.add_argument('--secret-key', required=True)
parser.add_argument('--output', default='haproxy.cfg',
help='Defaults to haproxy.cfg if not specified.')
parser.add_argument('--template', default='templates/haproxy.tpl')
parser.add_argument('--haproxy', default='./haproxy',
help='The haproxy binary to call. Defaults to haproxy if not specified.')
parser.add_argument('--pid', default='/var/run/haproxy.pid',
help='The pid file for haproxy. Defaults to /var/run/haproxy.pid.')
parser.add_argument('--eip',
help='The Elastic IP to bind to when VIP seems unhealthy.')
parser.add_argument('--health-check-url',
help='The URL to check. Assigns EIP to self if health check fails.')
parser.add_argument('--autoscaling-group',
help='The AutoScaling Group that health instance located in.')
parser.add_argument('--region',
help='The region that the SecurityGroup located in.')
args = parser.parse_args()
# Fetch a list of all the instances in these security groups.
instances = {}
if args.security_group:
for security_group in args.security_group:
logging.info('Getting instances for %s.' % security_group)
instances[security_group] = get_running_instances_in_security_group(access_key=args.access_key,
secret_key=args.secret_key,
security_group=security_group,
region=args.region)
if args.autoscaling_group:
autoscaling_group=args.autoscaling_group
logging.info('Getting instances for %s.' % autoscaling_group)
instances[autoscaling_group] = get_running_instances_in_autoscaling_group(access_key=args.access_key,
secret_key=args.secret_key,
autoscaling_group=autoscaling_group,
region=args.region)
print(instances)
# Generate the new config from the template.
logging.info('Generating configuration for haproxy.')
new_configuration = generate_haproxy_config(template=args.template,
instances=instances)
# See if this new config is different. If it is then restart using it.
# Otherwise just delete the temporary file and do nothing.
logging.info('Comparing to existing configuration.')
old_configuration = file_contents(filename=args.output)
if new_configuration != old_configuration:
logging.info('Existing configuration is outdated.')
# Overwite the existing config file.
logging.info('Writing new configuration.')
file_contents(filename=args.output,
content=generate_haproxy_config(template=args.template,
instances=instances ))
# Get PID if haproxy is already running.
logging.info('Fetching PID from %s.' % args.pid)
pid = file_contents(filename=args.pid)
# Restart haproxy.
logging.info('Restarting haproxy.')
command = '''%s -p %s -f %s -sf %s''' % (args.haproxy, args.pid, args.output, pid or '')
logging.info('Executing: %s' % command)
subprocess.call(command, shell=True)
else:
logging.info('Configuration unchanged. Skipping restart.')
# Do a health check on the url if specified.
try:
if args.health_check_url and args.eip:
logging.info('Performing health check.')
try:
logging.info('Checking %s' % args.health_check_url)
response = urllib2.urlopen(args.health_check_url)
logging.info('Response: %s' % response.read())
except:
# Assign the EIP to self.
logging.warn('Health check failed. Assigning %s to self.' % args.eip)
steal_elastic_ip(access_key=args.access_key,
secret_key=args.secret_key,
ip=args.eip,
region=args.region )
except:
pass
def main():
# Parse up the command line arguments.
parser = argparse.ArgumentParser(description='Update haproxy to use all instances running in a security group.')
parser.add_argument('--security-group', required=True, nargs='+', type=str)
parser.add_argument('--access-key', required=True)
parser.add_argument('--secret-key', required=True)
parser.add_argument('--output', default='haproxy.cfg',
help='Defaults to haproxy.cfg if not specified.')
parser.add_argument('--template', default='templates/haproxy.tpl')
parser.add_argument('--haproxy', default='./haproxy',
help='The haproxy binary to call. Defaults to haproxy if not specified.')
parser.add_argument('--pid', default='/var/run/haproxy.pid',
help='The pid file for haproxy. Defaults to /var/run/haproxy.pid.')
parser.add_argument('--eip',
help='The Elastic IP to bind to when VIP seems unhealthy.')
parser.add_argument('--health-check-url',
help='The URL to check. Assigns EIP to self if health check fails.')
args = parser.parse_args()
# Fetch a list of all the instances in these security groups.
instances = {}
for security_group in args.security_group:
logging.info('Getting instances for %s.' % security_group)
instances[security_group] = get_running_instances(access_key=args.access_key,
secret_key=args.secret_key,
security_group=security_group)
# Generate the new config from the template.
logging.info('Generating configuration for haproxy.')
new_configuration = generate_haproxy_config(template=args.template,
instances=instances)
# See if this new config is different. If it is then restart using it.
# Otherwise just delete the temporary file and do nothing.
logging.info('Comparing to existing configuration.')
old_configuration = file_contents(filename=args.output)
if new_configuration != old_configuration:
logging.info('Existing configuration is outdated.')
# Overwite the existing config file.
logging.info('Writing new configuration.')
file_contents(filename=args.output,
content=generate_haproxy_config(template=args.template,
instances=instances ))
# Get PID if haproxy is already running.
logging.info('Fetching PID from %s.' % args.pid)
pid = file_contents(filename=args.pid)
# Restart haproxy.
logging.info('Restarting haproxy.')
command = '''%s -p %s -f %s -sf %s''' % (args.haproxy, args.pid, args.output, pid or '')
logging.info('Executing: %s' % command)
subprocess.call(command, shell=True)
else:
logging.info('Configuration unchanged. Skipping restart.')
# Do a health check on the url if specified.
try:
if args.health_check_url and args.eip:
logging.info('Performing health check.')
try:
logging.info('Checking %s' % args.health_check_url)
response = urllib2.urlopen(args.health_check_url)
logging.info('Response: %s' % response.read())
except:
# Assign the EIP to self.
logging.warn('Health check failed. Assigning %s to self.' % args.eip)
steal_elastic_ip(access_key=args.access_key,
secret_key=args.secret_key,
ip=args.eip )
except:
pass
def main():
# Parse up the command line arguments.
parser = argparse.ArgumentParser(
description="Update haproxy to use all instances running with an aws:autoscaling:groupName tag of the given value."
)
parser.add_argument("--tag", required=True, nargs="+", type=str)
parser.add_argument("--access-key", required=True)
parser.add_argument("--secret-key", required=True)
parser.add_argument("--output", default="haproxy.cfg", help="Defaults to haproxy.cfg if not specified.")
parser.add_argument("--template", default="templates/haproxy.tpl")
parser.add_argument(
"--haproxy", default="./haproxy", help="The haproxy binary to call. Defaults to haproxy if not specified."
)
parser.add_argument(
"--pid", default="/var/run/haproxy.pid", help="The pid file for haproxy. Defaults to /var/run/haproxy.pid."
)
parser.add_argument("--eip", help="The Elastic IP to bind to when VIP seems unhealthy.")
parser.add_argument("--health-check-url", help="The URL to check. Assigns EIP to self if health check fails.")
args = parser.parse_args()
FORMAT = "%(asctime)-15s %(message)s"
logging.basicConfig(format=FORMAT)
# Fetch a list of all the instances in these security groups.
instances = {}
for tag in args.tag:
# logging.info('Getting instances for %s.' % tag)
instances[tag] = get_running_instances_by_tag(
access_key=args.access_key, secret_key=args.secret_key, tag_key="aws:autoscaling:groupName", tag_value=tag
)
# Generate the new config from the template.
# logging.info('Generating configuration for haproxy.')
new_configuration = generate_haproxy_config(template=args.template, instances=instances)
# See if this new config is different. If it is then restart using it.
# Otherwise just delete the temporary file and do nothing.
# logging.info('Comparing to existing configuration.')
old_configuration = file_contents(filename=args.output)
if new_configuration != old_configuration:
logging.info("Existing configuration is outdated.")
# Overwite the existing config file.
logging.info("Writing new configuration.")
file_contents(
filename=args.output, content=generate_haproxy_config(template=args.template, instances=instances)
)
# Get PID if haproxy is already running.
# logging.info('Fetching PID from %s.' % args.pid)
# pid = file_contents(filename=args.pid)
# Letting Monit handle the actual restarting
# logging.info('Restarting haproxy.')
# command = '''%s -p %s -f %s -sf %s''' % (args.haproxy, args.pid, args.output, pid or '')
# logging.info('Executing: %s' % command)
# subprocess.call(command, shell=True)
else:
logging.info("Configuration unchanged. ")
# Do a health check on the url if specified.
try:
if args.health_check_url and args.eip:
logging.info("Performing health check.")
try:
logging.info("Checking %s" % args.health_check_url)
response = urllib2.urlopen(args.health_check_url)
logging.info("Response: %s" % response.read())
except:
# Assign the EIP to self.
logging.warn("Health check failed. Assigning %s to self." % args.eip)
steal_elastic_ip(access_key=args.access_key, secret_key=args.secret_key, ip=args.eip)
except:
pass
def main():
# Parse up the command line arguments.
parser = argparse.ArgumentParser(description='Update haproxy to use all instances running in a security group.')
parser.add_argument('--config', default="/etc/haproxy-autoscale/haproxy-autoscale.conf",
help='Path to the config file.')
parser.add_argument('--autoscaling-group', required=True, nargs='+', type=str)
parser.add_argument('--region', default="us-east-1", help="AWS region name")
parser.add_argument('--output', default='/etc/haproxy/haproxy.cfg',
help='Defaults to haproxy.cfg if not specified.')
parser.add_argument('--template', default='/etc/haproxy-autoscale/haproxy.tpl')
parser.add_argument('--haproxy', default='/usr/sbin/haproxy',
help='The haproxy binary to call. Defaults to haproxy if not specified.')
parser.add_argument('--pid', default='/var/run/haproxy.pid',
help='The pid file for haproxy. Defaults to /var/run/haproxy.pid.')
parser.add_argument('--eip',
help='The Elastic IP to bind to when VIP seems unhealthy.')
parser.add_argument('--health-check-url',
help='The URL to check. Assigns EIP to self if health check fails.')
args = parser.parse_args()
# haproxy-autoscale config file
config_fh = open(args.config, "ro")
config = configparser.RawConfigParser()
config.read_file(config_fh)
config_fh.close()
access_key = config.get("credential", "access_key")
secret_key = config.get("credential", "secret_key")
# Fetch a list of all the instances in these security groups.
instances = {}
for autoscaling_group in args.autoscaling_group:
logging.info('Getting instances for %s.' % autoscaling_group)
instances[autoscaling_group] = list_as_instances(access_key, secret_key, args.region,
autoscaling_group)
# Generate the new config from the template.
logging.info('Generating configuration for haproxy.')
new_configuration = generate_haproxy_config(args.template, instances)
# See if this new config is different. If it is then restart using it.
# Otherwise just delete the temporary file and do nothing.
logging.info('Comparing to existing configuration.')
old_configuration = file_contents(args.output)
if new_configuration != old_configuration:
logging.info('Existing configuration is outdated.')
# Overwite the existing config file.
logging.info('Writing new configuration.')
file_contents(args.output, generate_haproxy_config(args.template, instances))
# Get PID if haproxy is already running.
logging.info('Fetching PID from %s.' % args.pid)
pid = file_contents(args.pid)
# Restart haproxy.
logging.info('Restarting haproxy.')
command = '''%s -p %s -f %s -sf %s''' % (args.haproxy, args.pid, args.output, pid or '')
logging.info('Executing: %s' % command)
subprocess.call(command, shell=True)
else:
logging.info('Configuration unchanged. Skipping restart.')
# Do a health check on the url if specified.
try:
if args.health_check_url and args.eip:
logging.info('Performing health check.')
try:
logging.info('Checking %s' % args.health_check_url)
response = urllib2.urlopen(args.health_check_url)
logging.info('Response: %s' % response.read())
except:
# Assign the EIP to self.
logging.warn('Health check failed. Assigning %s to self.' % args.eip)
steal_elastic_ip(access_key, secret_key, args.eip)
except:
pass
