system-config/launch/launch-node.py
Clark Boylan bc82cc3e90 Handle focal's insistence we don't use root in launch-node.py
It seems newer focal images force you to use the ubuntu user account. We
already have code that attempts to fallback to ubuntu, but we were not
properly catching the error when root fails which caused the whole
script to fail.

Address this by catching the exception, logging a message, then
continuing to the next possible option. If no possible options are found
we raise an exception already which handles the worst case situation.

Change-Id: Ie6013763daff01063840abce193050b33120a7a2
2021-04-21 16:51:46 -07:00

425 lines
16 KiB
Python
Executable File

#!/usr/bin/env python3
# Launch a new OpenStack project infrastructure node.
# Copyright (C) 2011-2012 OpenStack LLC.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
# implied.
#
# See the License for the specific language governing permissions and
# limitations under the License.
import argparse
import os
import shutil
import subprocess
import sys
import threading
import tempfile
import time
import traceback
import dns
import utils
import openstack
import paramiko
from sshclient import SSHException
SCRIPT_DIR = os.path.dirname(sys.argv[0])
try:
# This unactionable warning does not need to be printed over and over.
import requests.packages.urllib3
requests.packages.urllib3.disable_warnings()
except:
pass
class JobDir(object):
def __init__(self, keep=False):
self.keep = keep
self.root = tempfile.mkdtemp()
self.inventory_root = os.path.join(self.root, 'inventory')
os.makedirs(self.inventory_root)
self.hosts = os.path.join(self.inventory_root, 'hosts')
self.groups = os.path.join(self.inventory_root, 'groups')
self.key = os.path.join(self.root, 'id_rsa')
self.ansible_log = os.path.join(self.root, 'ansible_log.txt')
# XXX if we need more, we might like to setup an ansible.cfg
# file and use that rather than env vars. See
# zuul/launcher/ansiblelaunchserver.py as an example
self.env = os.environ.copy()
self.env['ANSIBLE_LOG_PATH'] = self.ansible_log
def __enter__(self):
return self
def __exit__(self, etype, value, tb):
if not self.keep:
shutil.rmtree(self.root)
def run(cmd, **args):
args['stdout'] = subprocess.PIPE
args['stderr'] = subprocess.STDOUT
print("Running: %s" % (cmd,))
proc = subprocess.Popen(cmd, **args)
out = ''
for line in iter(proc.stdout.readline, b''):
line = line.decode('utf-8')
sys.stdout.write(line)
sys.stdout.flush()
out += line
ret = proc.wait()
print("Return code: %s" % (ret,))
if ret != 0:
raise subprocess.CalledProcessError(ret, cmd, out)
return ret
def stream_syslog(ssh_client):
try:
ssh_client.ssh('tail -f /var/log/syslog')
except Exception:
print("Syslog stream terminated")
def bootstrap_server(server, key, name, volume_device, keep,
mount_path, fs_label, environment, timeout, ignore_ipv6):
ip = server.public_v4
ssh_kwargs = dict(pkey=key)
print("--- Running initial configuration on host %s ---" % ip)
ssh_client = None
for username in ['root', 'ubuntu', 'centos', 'admin']:
try:
ssh_client = utils.ssh_connect(ip, username, ssh_kwargs,
timeout=timeout)
except SSHException:
print("Username: " + username + " failed to ssh. "
"Trying next option.")
if ssh_client:
break
if not ssh_client:
raise Exception("Unable to log in via SSH")
# cloud-init puts the "please log in as user foo" message and
# subsequent exit() in root's authorized_keys -- overwrite it with
# a normal version to get root login working again.
if username != 'root':
ssh_client.ssh("sudo cp ~/.ssh/authorized_keys"
" ~root/.ssh/authorized_keys")
ssh_client.ssh("sudo chmod 644 ~root/.ssh/authorized_keys")
ssh_client.ssh("sudo chown root.root ~root/.ssh/authorized_keys")
ssh_client = utils.ssh_connect(ip, 'root', ssh_kwargs, timeout=timeout)
if not ignore_ipv6:
# Something up with RAX images that they have the ipv6 interface in
# /etc/network/interfaces but eth0 hasn't noticed yet; reload it
ssh_client.ssh('(ifdown eth0 && ifup eth0) || true')
if server.public_v6:
# The server may be waiting on Router Advertisements to configure
# this address. Wait for that to complete before pinging.
ssh_client.ssh("bash -c 'count=0 ; "
"while ! ip addr | grep -q %s && [\"$count\" -le 60 ] ; do "
" count=$((count + 1)) ; "
" echo \"Waiting for IPv6 address to configure\" ; "
" sleep 1 ; "
"done'" % server.public_v6)
ssh_client.ssh('ping6 -c5 -Q 0x10 review.openstack.org '
'|| ping6 -c5 -Q 0x10 wiki.openstack.org')
ssh_client.scp(os.path.join(SCRIPT_DIR, 'make_swap.sh'),
'make_swap.sh')
ssh_client.ssh('bash -x make_swap.sh')
if volume_device:
ssh_client.scp(os.path.join(SCRIPT_DIR, 'mount_volume.sh'),
'mount_volume.sh')
ssh_client.ssh('bash -x mount_volume.sh %s %s %s' %
(volume_device, mount_path, fs_label))
# Zero the ansible inventory cache so that next run finds the new server
inventory_cache_dir = '/var/cache/ansible/inventory'
for inventory_cache in os.listdir(inventory_cache_dir):
os.unlink(os.path.join(inventory_cache_dir, inventory_cache))
with JobDir(keep) as jobdir:
# Update the generated-groups file globally and incorporate it
# into our inventory
# Remove cloud and region from the environment to work
# around a bug in occ
expand_env = os.environ.copy()
for env_key in list(expand_env.keys()):
if env_key.startswith('OS_'):
expand_env.pop(env_key, None)
expand_env['ANSIBLE_LOG_PATH'] = jobdir.ansible_log
# Write out the private SSH key we generated
with open(jobdir.key, 'w') as key_file:
key.write_private_key(key_file)
os.chmod(jobdir.key, 0o600)
if ignore_ipv6:
host_ip = server.public_v4
else:
host_ip = server.interface_ip
# Write out inventory
with open(jobdir.hosts, 'w') as inventory_file:
inventory_file.write(
"{host} ansible_host={ip} ansible_user=root {python}".format(
host=name, ip=host_ip,
python='ansible_python_interpreter=/usr/bin/python3'))
t = threading.Thread(target=stream_syslog, args=(ssh_client,))
t.daemon = True
t.start()
inventory_list = (
'/home/zuul/src/opendev.org/opendev/system-config/inventory/base/hosts.yaml',
'/home/zuul/src/opendev.org/opendev/system-config/inventory/service/groups.yaml',
'/etc/ansible/hosts/emergency.yaml',
jobdir.inventory_root,
)
inventory_cmds = [v for e in inventory_list for v in ('-i', e)]
ansible_cmd = ['ansible-playbook', '--flush-cache' ] + \
inventory_cmds + \
['-l', name,
'--private-key={key}'.format(key=jobdir.key),
"--ssh-common-args='-o StrictHostKeyChecking=no'",
'-e', 'target={name}'.format(name=name)]
# Run the base playbook limited to just this server we just created
for playbook in [
'set-hostnames.yaml',
'base.yaml',
'apply-package-updates.yaml',
]:
run(ansible_cmd + [
os.path.join(SCRIPT_DIR, '..', 'playbooks', playbook)],
env=jobdir.env)
try:
ssh_client.ssh("reboot")
except Exception as e:
# Some init system kill the connection too fast after reboot.
# Deal with it by ignoring ssh errors when rebooting.
if e.rc == -1:
pass
else:
raise
def build_server(cloud, name, image, flavor,
volume, keep, network, boot_from_volume, config_drive,
mount_path, fs_label, availability_zone, environment,
volume_size, timeout, ignore_ipv6):
key = None
server = None
create_kwargs = dict(image=image, flavor=flavor, name=name,
reuse_ips=False, wait=True,
boot_from_volume=boot_from_volume,
volume_size=volume_size,
network=network,
config_drive=config_drive,
timeout=timeout)
if availability_zone:
create_kwargs['availability_zone'] = availability_zone
if volume:
create_kwargs['volumes'] = [volume]
key_name = 'launch-%i' % (time.time())
key = paramiko.RSAKey.generate(2048)
public_key = key.get_name() + ' ' + key.get_base64()
cloud.create_keypair(key_name, public_key)
create_kwargs['key_name'] = key_name
try:
server = cloud.create_server(**create_kwargs)
except Exception:
try:
cloud.delete_keypair(key_name)
except Exception:
print("Exception encountered deleting keypair:")
traceback.print_exc()
raise
try:
cloud.delete_keypair(key_name)
server = cloud.get_openstack_vars(server)
if volume:
volume = cloud.get_volume(volume)
volume_device = cloud.get_volume_attach_device(volume,
server['id'])
else:
volume_device = None
bootstrap_server(server, key, name, volume_device, keep,
mount_path, fs_label, environment, timeout, ignore_ipv6)
print('UUID=%s\nIPV4=%s\nIPV6=%s\n' % (
server.id, server.public_v4, server.public_v6))
except Exception:
print("****")
print("Server %s failed to build!" % (server.id))
try:
volumes = []
if boot_from_volume:
volumes = cloud.get_volumes(server)
if keep:
print("Keeping as requested")
# Write out the private SSH key we generated, as we
# may not have got far enough for ansible to run
with open('/tmp/%s.id_rsa' % server.id, 'w') as key_file:
key.write_private_key(key_file)
os.chmod(key_file.name, 0o600)
print("Private key saved in %s" % key_file.name)
print("Run to delete:")
print(" openstack server delete %s" % (server.id))
for attachment in volumes:
print(" openstack volume delete %s" % (attachment.id))
else:
cloud.delete_server(server.id, delete_ips=True)
for attachment in volumes:
cloud.delete_volume(attachment.id)
except Exception:
print("Exception encountered deleting server:")
traceback.print_exc()
print("The original exception follows:")
print("****")
# Raise the important exception that started this
raise
return server
def main():
parser = argparse.ArgumentParser()
parser.add_argument("name", help="server name")
parser.add_argument("--cloud", dest="cloud", required=True,
help="cloud name")
parser.add_argument("--region", dest="region",
help="cloud region")
parser.add_argument("--flavor", dest="flavor", default='1GB',
help="name (or substring) of flavor")
parser.add_argument("--image", dest="image",
default="Ubuntu 18.04 LTS (Bionic Beaver) (Cloud)",
help="image name")
parser.add_argument("--environment", dest="environment",
help="Puppet environment to use",
default=None)
parser.add_argument("--volume", dest="volume",
help="UUID of volume to attach to the new server.",
default=None)
parser.add_argument("--mount-path", dest="mount_path",
help="Path to mount cinder volume at.",
default=None)
parser.add_argument("--fs-label", dest="fs_label",
help="FS label to use when mounting cinder volume.",
default=None)
parser.add_argument("--boot-from-volume", dest="boot_from_volume",
help="Create a boot volume for the server and use it.",
action='store_true',
default=False)
parser.add_argument("--volume-size", dest="volume_size",
help="Size of volume (GB) for --boot-from-volume",
default="50")
parser.add_argument("--keep", dest="keep",
help="Don't clean up or delete the server on error.",
action='store_true',
default=False)
parser.add_argument("--verbose", dest="verbose", default=False,
action='store_true',
help="Be verbose about logging cloud actions")
parser.add_argument("--network", dest="network", default=None,
help="network label to attach instance to")
parser.add_argument("--config-drive", dest="config_drive",
help="Boot with config_drive attached.",
action='store_true',
default=False)
parser.add_argument("--timeout", dest="timeout",
help="Increase timeouts (default 600s)",
type=int, default=600)
parser.add_argument("--ignore_ipv6", dest="ignore_ipv6",
help="Ignore IPv6 addresses from API",
action='store_true', default=False)
parser.add_argument("--az", dest="availability_zone", default=None,
help="AZ to boot in.")
options = parser.parse_args()
openstack.enable_logging(debug=options.verbose)
cloud_kwargs = {}
if options.region:
cloud_kwargs['region_name'] = options.region
cloud = openstack.connect(cloud=options.cloud, **cloud_kwargs)
flavor = cloud.get_flavor(options.flavor)
if flavor:
print("Found flavor", flavor.name)
else:
print("Unable to find matching flavor; flavor list:")
for i in cloud.list_flavors():
print(i.name)
sys.exit(1)
image = cloud.get_image_exclude(options.image, 'deprecated')
if image:
print("Found image", image.name)
else:
print("Unable to find matching image; image list:")
for i in cloud.list_images():
print(i.name)
sys.exit(1)
# NOTE(ianw): 2019-06-26 for whatever reason OVH assigns an IPv6
# address that is seen in API queries, but it is not in the host
# metadata so the interface isn't autoconfigured. Auto skip it to
# avoid this bombing out.
if "-ovh" in options.cloud:
print("Ignoring IPv6 for OVH cloud instances")
options.ignore_ipv6 = True
server = build_server(cloud, options.name, image, flavor,
options.volume, options.keep,
options.network, options.boot_from_volume,
options.config_drive,
options.mount_path, options.fs_label,
options.availability_zone,
options.environment, options.volume_size,
options.timeout, options.ignore_ipv6)
dns.print_dns(cloud, server)
print("If this is a server that is expected to send email (ask, review,")
print("lists, etc) double check that the server's IPs are not listed on")
print("the spamhaus pbl.\n")
print("URLs to check:")
print("https://www.spamhaus.org/query/ip/%s" % server.public_v4)
print("https://www.spamhaus.org/query/ip/%s" % server.public_v6)
print()
print("When requesting an exception you can use the")
print("infra-root@openstack.org email address to verify the responsible")
print("party.")
if __name__ == '__main__':
main()