# Copyright (c) 2011 X.commerce, a business unit of eBay Inc. # Copyright 2010 United States Government as represented by the # Administrator of the National Aeronautics and Space Administration. # All Rights Reserved. # # 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. """Implements vlans, bridges, and iptables rules using linux utilities.""" import calendar import inspect import os import re import time import netaddr import netifaces from oslo_concurrency import processutils from oslo_config import cfg from oslo_log import log as logging from oslo_serialization import jsonutils from oslo_utils import excutils from oslo_utils import fileutils from oslo_utils import importutils from oslo_utils import timeutils import six from nova import exception from nova.i18n import _, _LE, _LW from nova import objects from nova import paths from nova.pci import utils as pci_utils from nova import utils LOG = logging.getLogger(__name__) linux_net_opts = [ cfg.MultiStrOpt('dhcpbridge_flagfile', default=['/etc/nova/nova-dhcpbridge.conf'], help='Location of flagfiles for dhcpbridge'), cfg.StrOpt('networks_path', default=paths.state_path_def('networks'), help='Location to keep network config files'), cfg.StrOpt('public_interface', default='eth0', help='Interface for public IP addresses'), cfg.StrOpt('dhcpbridge', default=paths.bindir_def('nova-dhcpbridge'), help='Location of nova-dhcpbridge'), cfg.StrOpt('routing_source_ip', default='$my_ip', help='Public IP of network host'), cfg.IntOpt('dhcp_lease_time', default=86400, help='Lifetime of a DHCP lease in seconds'), cfg.MultiStrOpt('dns_server', default=[], help='If set, uses specific DNS server for dnsmasq. Can' ' be specified multiple times.'), cfg.BoolOpt('use_network_dns_servers', default=False, help='If set, uses the dns1 and dns2 from the network ref.' ' as dns servers.'), cfg.ListOpt('dmz_cidr', default=[], help='A list of dmz ranges that should be accepted'), cfg.MultiStrOpt('force_snat_range', default=[], help='Traffic to this range will always be snatted to the ' 'fallback ip, even if it would normally be bridged out ' 'of the node. Can be specified multiple times.'), cfg.StrOpt('dnsmasq_config_file', default='', help='Override the default dnsmasq settings with this file'), cfg.StrOpt('linuxnet_interface_driver', default='nova.network.linux_net.LinuxBridgeInterfaceDriver', help='Driver used to create ethernet devices.'), cfg.StrOpt('linuxnet_ovs_integration_bridge', default='br-int', help='Name of Open vSwitch bridge used with linuxnet'), cfg.BoolOpt('send_arp_for_ha', default=False, help='Send gratuitous ARPs for HA setup'), cfg.IntOpt('send_arp_for_ha_count', default=3, help='Send this many gratuitous ARPs for HA setup'), cfg.BoolOpt('use_single_default_gateway', default=False, help='Use single default gateway. Only first nic of vm will ' 'get default gateway from dhcp server'), cfg.MultiStrOpt('forward_bridge_interface', default=['all'], help='An interface that bridges can forward to. If this ' 'is set to all then all traffic will be forwarded. ' 'Can be specified multiple times.'), cfg.StrOpt('metadata_host', default='$my_ip', help='The IP address for the metadata API server'), cfg.IntOpt('metadata_port', default=8775, min=1, max=65535, help='The port for the metadata API port'), cfg.StrOpt('iptables_top_regex', default='', help='Regular expression to match the iptables rule that ' 'should always be on the top.'), cfg.StrOpt('iptables_bottom_regex', default='', help='Regular expression to match the iptables rule that ' 'should always be on the bottom.'), cfg.StrOpt('iptables_drop_action', default='DROP', help='The table that iptables to jump to when a packet is ' 'to be dropped.'), cfg.IntOpt('ovs_vsctl_timeout', default=120, help='Amount of time, in seconds, that ovs_vsctl should wait ' 'for a response from the database. 0 is to wait forever.'), cfg.BoolOpt('fake_network', default=False, help='If passed, use fake network devices and addresses'), cfg.IntOpt('ebtables_exec_attempts', default=3, help='Number of times to retry ebtables commands on failure.'), cfg.FloatOpt('ebtables_retry_interval', default=1.0, help='Number of seconds to wait between ebtables retries.'), ] CONF = cfg.CONF CONF.register_opts(linux_net_opts) CONF.import_opt('host', 'nova.netconf') CONF.import_opt('use_ipv6', 'nova.netconf') CONF.import_opt('my_ip', 'nova.netconf') CONF.import_opt('network_device_mtu', 'nova.objects.network') # NOTE(vish): Iptables supports chain names of up to 28 characters, and we # add up to 12 characters to binary_name which is used as a prefix, # so we limit it to 16 characters. # (max_chain_name_length - len('-POSTROUTING') == 16) def get_binary_name(): """Grab the name of the binary we're running in.""" return os.path.basename(inspect.stack()[-1][1])[:16] binary_name = get_binary_name() class IptablesRule(object): """An iptables rule. You shouldn't need to use this class directly, it's only used by IptablesManager. """ def __init__(self, chain, rule, wrap=True, top=False): self.chain = chain self.rule = rule self.wrap = wrap self.top = top def __eq__(self, other): return ((self.chain == other.chain) and (self.rule == other.rule) and (self.top == other.top) and (self.wrap == other.wrap)) def __ne__(self, other): return not self == other def __repr__(self): if self.wrap: chain = '%s-%s' % (binary_name, self.chain) else: chain = self.chain # new rules should have a zero [packet: byte] count return '[0:0] -A %s %s' % (chain, self.rule) class IptablesTable(object): """An iptables table.""" def __init__(self): self.rules = [] self.remove_rules = [] self.chains = set() self.unwrapped_chains = set() self.remove_chains = set() self.dirty = True def has_chain(self, name, wrap=True): if wrap: return name in self.chains else: return name in self.unwrapped_chains def add_chain(self, name, wrap=True): """Adds a named chain to the table. The chain name is wrapped to be unique for the component creating it, so different components of Nova can safely create identically named chains without interfering with one another. At the moment, its wrapped name is -, so if nova-compute creates a chain named 'OUTPUT', it'll actually end up named 'nova-compute-OUTPUT'. """ if wrap: self.chains.add(name) else: self.unwrapped_chains.add(name) self.dirty = True def remove_chain(self, name, wrap=True): """Remove named chain. This removal "cascades". All rule in the chain are removed, as are all rules in other chains that jump to it. If the chain is not found, this is merely logged. """ if wrap: chain_set = self.chains else: chain_set = self.unwrapped_chains if name not in chain_set: LOG.warning(_LW('Attempted to remove chain %s which does not ' 'exist'), name) return self.dirty = True # non-wrapped chains and rules need to be dealt with specially, # so we keep a list of them to be iterated over in apply() if not wrap: self.remove_chains.add(name) chain_set.remove(name) if not wrap: self.remove_rules += filter(lambda r: r.chain == name, self.rules) self.rules = filter(lambda r: r.chain != name, self.rules) if wrap: jump_snippet = '-j %s-%s' % (binary_name, name) else: jump_snippet = '-j %s' % (name,) if not wrap: self.remove_rules += filter(lambda r: jump_snippet in r.rule, self.rules) self.rules = filter(lambda r: jump_snippet not in r.rule, self.rules) def add_rule(self, chain, rule, wrap=True, top=False): """Add a rule to the table. This is just like what you'd feed to iptables, just without the '-A ' bit at the start. However, if you need to jump to one of your wrapped chains, prepend its name with a '$' which will ensure the wrapping is applied correctly. """ if wrap and chain not in self.chains: raise ValueError(_('Unknown chain: %r') % chain) if '$' in rule: rule = ' '.join(map(self._wrap_target_chain, rule.split(' '))) rule_obj = IptablesRule(chain, rule, wrap, top) if rule_obj in self.rules: LOG.debug("Skipping duplicate iptables rule addition. " "%(rule)r already in %(rules)r", {'rule': rule_obj, 'rules': self.rules}) else: self.rules.append(IptablesRule(chain, rule, wrap, top)) self.dirty = True def _wrap_target_chain(self, s): if s.startswith('$'): return '%s-%s' % (binary_name, s[1:]) return s def remove_rule(self, chain, rule, wrap=True, top=False): """Remove a rule from a chain. Note: The rule must be exactly identical to the one that was added. You cannot switch arguments around like you can with the iptables CLI tool. """ try: self.rules.remove(IptablesRule(chain, rule, wrap, top)) if not wrap: self.remove_rules.append(IptablesRule(chain, rule, wrap, top)) self.dirty = True except ValueError: LOG.warning(_LW('Tried to remove rule that was not there:' ' %(chain)r %(rule)r %(wrap)r %(top)r'), {'chain': chain, 'rule': rule, 'top': top, 'wrap': wrap}) def remove_rules_regex(self, regex): """Remove all rules matching regex.""" if isinstance(regex, six.string_types): regex = re.compile(regex) num_rules = len(self.rules) self.rules = filter(lambda r: not regex.match(str(r)), self.rules) removed = num_rules - len(self.rules) if removed > 0: self.dirty = True return removed def empty_chain(self, chain, wrap=True): """Remove all rules from a chain.""" chained_rules = [rule for rule in self.rules if rule.chain == chain and rule.wrap == wrap] if chained_rules: self.dirty = True for rule in chained_rules: self.rules.remove(rule) class IptablesManager(object): """Wrapper for iptables. See IptablesTable for some usage docs A number of chains are set up to begin with. First, nova-filter-top. It's added at the top of FORWARD and OUTPUT. Its name is not wrapped, so it's shared between the various nova workers. It's intended for rules that need to live at the top of the FORWARD and OUTPUT chains. It's in both the ipv4 and ipv6 set of tables. For ipv4 and ipv6, the built-in INPUT, OUTPUT, and FORWARD filter chains are wrapped, meaning that the "real" INPUT chain has a rule that jumps to the wrapped INPUT chain, etc. Additionally, there's a wrapped chain named "local" which is jumped to from nova-filter-top. For ipv4, the built-in PREROUTING, OUTPUT, and POSTROUTING nat chains are wrapped in the same was as the built-in filter chains. Additionally, there's a snat chain that is applied after the POSTROUTING chain. """ def __init__(self, execute=None): if not execute: self.execute = _execute else: self.execute = execute self.ipv4 = {'filter': IptablesTable(), 'nat': IptablesTable(), 'mangle': IptablesTable()} self.ipv6 = {'filter': IptablesTable()} self.iptables_apply_deferred = False # Add a nova-filter-top chain. It's intended to be shared # among the various nova components. It sits at the very top # of FORWARD and OUTPUT. for tables in [self.ipv4, self.ipv6]: tables['filter'].add_chain('nova-filter-top', wrap=False) tables['filter'].add_rule('FORWARD', '-j nova-filter-top', wrap=False, top=True) tables['filter'].add_rule('OUTPUT', '-j nova-filter-top', wrap=False, top=True) tables['filter'].add_chain('local') tables['filter'].add_rule('nova-filter-top', '-j $local', wrap=False) # Wrap the built-in chains builtin_chains = {4: {'filter': ['INPUT', 'OUTPUT', 'FORWARD'], 'nat': ['PREROUTING', 'OUTPUT', 'POSTROUTING'], 'mangle': ['POSTROUTING']}, 6: {'filter': ['INPUT', 'OUTPUT', 'FORWARD']}} for ip_version in builtin_chains: if ip_version == 4: tables = self.ipv4 elif ip_version == 6: tables = self.ipv6 for table, chains in six.iteritems(builtin_chains[ip_version]): for chain in chains: tables[table].add_chain(chain) tables[table].add_rule(chain, '-j $%s' % (chain,), wrap=False) # Add a nova-postrouting-bottom chain. It's intended to be shared # among the various nova components. We set it as the last chain # of POSTROUTING chain. self.ipv4['nat'].add_chain('nova-postrouting-bottom', wrap=False) self.ipv4['nat'].add_rule('POSTROUTING', '-j nova-postrouting-bottom', wrap=False) # We add a snat chain to the shared nova-postrouting-bottom chain # so that it's applied last. self.ipv4['nat'].add_chain('snat') self.ipv4['nat'].add_rule('nova-postrouting-bottom', '-j $snat', wrap=False) # And then we add a float-snat chain and jump to first thing in # the snat chain. self.ipv4['nat'].add_chain('float-snat') self.ipv4['nat'].add_rule('snat', '-j $float-snat') def defer_apply_on(self): self.iptables_apply_deferred = True def defer_apply_off(self): self.iptables_apply_deferred = False self.apply() def dirty(self): for table in six.itervalues(self.ipv4): if table.dirty: return True if CONF.use_ipv6: for table in six.itervalues(self.ipv6): if table.dirty: return True return False def apply(self): if self.iptables_apply_deferred: return if self.dirty(): self._apply() else: LOG.debug("Skipping apply due to lack of new rules") @utils.synchronized('iptables', external=True) def _apply(self): """Apply the current in-memory set of iptables rules. This will blow away any rules left over from previous runs of the same component of Nova, and replace them with our current set of rules. This happens atomically, thanks to iptables-restore. """ s = [('iptables', self.ipv4)] if CONF.use_ipv6: s += [('ip6tables', self.ipv6)] for cmd, tables in s: all_tables, _err = self.execute('%s-save' % (cmd,), '-c', run_as_root=True, attempts=5) all_lines = all_tables.split('\n') for table_name, table in six.iteritems(tables): start, end = self._find_table(all_lines, table_name) all_lines[start:end] = self._modify_rules( all_lines[start:end], table, table_name) table.dirty = False self.execute('%s-restore' % (cmd,), '-c', run_as_root=True, process_input='\n'.join(all_lines), attempts=5) LOG.debug("IPTablesManager.apply completed with success") def _find_table(self, lines, table_name): if len(lines) < 3: # length only <2 when fake iptables return (0, 0) try: start = lines.index('*%s' % table_name) - 1 except ValueError: # Couldn't find table_name return (0, 0) end = lines[start:].index('COMMIT') + start + 2 return (start, end) def _modify_rules(self, current_lines, table, table_name): unwrapped_chains = table.unwrapped_chains chains = sorted(table.chains) remove_chains = table.remove_chains rules = table.rules remove_rules = table.remove_rules if not current_lines: fake_table = ['#Generated by nova', '*' + table_name, 'COMMIT', '#Completed by nova'] current_lines = fake_table # Remove any trace of our rules new_filter = filter(lambda line: binary_name not in line, current_lines) top_rules = [] bottom_rules = [] if CONF.iptables_top_regex: regex = re.compile(CONF.iptables_top_regex) temp_filter = filter(lambda line: regex.search(line), new_filter) for rule_str in temp_filter: new_filter = filter(lambda s: s.strip() != rule_str.strip(), new_filter) top_rules = temp_filter if CONF.iptables_bottom_regex: regex = re.compile(CONF.iptables_bottom_regex) temp_filter = filter(lambda line: regex.search(line), new_filter) for rule_str in temp_filter: new_filter = filter(lambda s: s.strip() != rule_str.strip(), new_filter) bottom_rules = temp_filter seen_chains = False rules_index = 0 for rules_index, rule in enumerate(new_filter): if not seen_chains: if rule.startswith(':'): seen_chains = True else: if not rule.startswith(':'): break if not seen_chains: rules_index = 2 our_rules = top_rules bot_rules = [] for rule in rules: rule_str = str(rule) if rule.top: # rule.top == True means we want this rule to be at the top. # Further down, we weed out duplicates from the bottom of the # list, so here we remove the dupes ahead of time. # We don't want to remove an entry if it has non-zero # [packet:byte] counts and replace it with [0:0], so let's # go look for a duplicate, and over-ride our table rule if # found. # ignore [packet:byte] counts at beginning of line if rule_str.startswith('['): rule_str = rule_str.split(']', 1)[1] dup_filter = filter(lambda s: rule_str.strip() in s.strip(), new_filter) new_filter = filter(lambda s: rule_str.strip() not in s.strip(), new_filter) # if no duplicates, use original rule if dup_filter: # grab the last entry, if there is one dup = dup_filter[-1] rule_str = str(dup) else: rule_str = str(rule) rule_str.strip() our_rules += [rule_str] else: bot_rules += [rule_str] our_rules += bot_rules new_filter[rules_index:rules_index] = our_rules new_filter[rules_index:rules_index] = [':%s - [0:0]' % (name,) for name in unwrapped_chains] new_filter[rules_index:rules_index] = [':%s-%s - [0:0]' % (binary_name, name,) for name in chains] commit_index = new_filter.index('COMMIT') new_filter[commit_index:commit_index] = bottom_rules seen_lines = set() def _weed_out_duplicates(line): # ignore [packet:byte] counts at beginning of lines if line.startswith('['): line = line.split(']', 1)[1] line = line.strip() if line in seen_lines: return False else: seen_lines.add(line) return True def _weed_out_removes(line): # We need to find exact matches here if line.startswith(':'): # it's a chain, for example, ":nova-billing - [0:0]" # strip off everything except the chain name line = line.split(':')[1] line = line.split('- [')[0] line = line.strip() for chain in remove_chains: if chain == line: remove_chains.remove(chain) return False elif line.startswith('['): # it's a rule # ignore [packet:byte] counts at beginning of lines line = line.split(']', 1)[1] line = line.strip() for rule in remove_rules: # ignore [packet:byte] counts at beginning of rules rule_str = str(rule) rule_str = rule_str.split(' ', 1)[1] rule_str = rule_str.strip() if rule_str == line: remove_rules.remove(rule) return False # Leave it alone return True # We filter duplicates, letting the *last* occurrence take # precedence. We also filter out anything in the "remove" # lists. new_filter.reverse() new_filter = filter(_weed_out_duplicates, new_filter) new_filter = filter(_weed_out_removes, new_filter) new_filter.reverse() # flush lists, just in case we didn't find something remove_chains.clear() for rule in remove_rules: remove_rules.remove(rule) return new_filter # NOTE(jkoelker) This is just a nice little stub point since mocking # builtins with mox is a nightmare def write_to_file(file, data, mode='w'): with open(file, mode) as f: f.write(data) def is_pid_cmdline_correct(pid, match): """Ensure that the cmdline for a pid seems sane Because pids are recycled, blindly killing by pid is something to avoid. This provides the ability to include a substring that is expected in the cmdline as a safety check. """ try: with open('/proc/%d/cmdline' % pid) as f: cmdline = f.read() return match in cmdline except EnvironmentError: return False def metadata_forward(): """Create forwarding rule for metadata.""" if CONF.metadata_host != '127.0.0.1': iptables_manager.ipv4['nat'].add_rule('PREROUTING', '-s 0.0.0.0/0 -d 169.254.169.254/32 ' '-p tcp -m tcp --dport 80 -j DNAT ' '--to-destination %s:%s' % (CONF.metadata_host, CONF.metadata_port)) else: iptables_manager.ipv4['nat'].add_rule('PREROUTING', '-s 0.0.0.0/0 -d 169.254.169.254/32 ' '-p tcp -m tcp --dport 80 ' '-j REDIRECT --to-ports %s' % CONF.metadata_port) iptables_manager.apply() def _iptables_dest(ip): if ((netaddr.IPAddress(ip).version == 4 and ip == '127.0.0.1') or ip == '::1'): return '-m addrtype --dst-type LOCAL' else: return '-d %s' % ip def metadata_accept(): """Create the filter accept rule for metadata.""" rule = ('-p tcp -m tcp --dport %s %s -j ACCEPT' % (CONF.metadata_port, _iptables_dest(CONF.metadata_host))) if netaddr.IPAddress(CONF.metadata_host).version == 4: iptables_manager.ipv4['filter'].add_rule('INPUT', rule) else: iptables_manager.ipv6['filter'].add_rule('INPUT', rule) iptables_manager.apply() def add_snat_rule(ip_range, is_external=False): if CONF.routing_source_ip: if is_external: if CONF.force_snat_range: snat_range = CONF.force_snat_range else: snat_range = [] else: snat_range = ['0.0.0.0/0'] for dest_range in snat_range: rule = ('-s %s -d %s -j SNAT --to-source %s' % (ip_range, dest_range, CONF.routing_source_ip)) if not is_external and CONF.public_interface: rule += ' -o %s' % CONF.public_interface iptables_manager.ipv4['nat'].add_rule('snat', rule) iptables_manager.apply() def init_host(ip_range, is_external=False): """Basic networking setup goes here.""" # NOTE(devcamcar): Cloud public SNAT entries and the default # SNAT rule for outbound traffic. add_snat_rule(ip_range, is_external) rules = [] if is_external: for snat_range in CONF.force_snat_range: rules.append('PREROUTING -p ipv4 --ip-src %s --ip-dst %s ' '-j redirect --redirect-target ACCEPT' % (ip_range, snat_range)) if rules: ensure_ebtables_rules(rules, 'nat') iptables_manager.ipv4['nat'].add_rule('POSTROUTING', '-s %s -d %s/32 -j ACCEPT' % (ip_range, CONF.metadata_host)) for dmz in CONF.dmz_cidr: iptables_manager.ipv4['nat'].add_rule('POSTROUTING', '-s %s -d %s -j ACCEPT' % (ip_range, dmz)) iptables_manager.ipv4['nat'].add_rule('POSTROUTING', '-s %(range)s -d %(range)s ' '-m conntrack ! --ctstate DNAT ' '-j ACCEPT' % {'range': ip_range}) iptables_manager.apply() def send_arp_for_ip(ip, device, count): out, err = _execute('arping', '-U', ip, '-A', '-I', device, '-c', str(count), run_as_root=True, check_exit_code=False) if err: LOG.debug('arping error for ip %s', ip) def bind_floating_ip(floating_ip, device): """Bind ip to public interface.""" _execute('ip', 'addr', 'add', str(floating_ip) + '/32', 'dev', device, run_as_root=True, check_exit_code=[0, 2, 254]) if CONF.send_arp_for_ha and CONF.send_arp_for_ha_count > 0: send_arp_for_ip(floating_ip, device, CONF.send_arp_for_ha_count) def unbind_floating_ip(floating_ip, device): """Unbind a public ip from public interface.""" _execute('ip', 'addr', 'del', str(floating_ip) + '/32', 'dev', device, run_as_root=True, check_exit_code=[0, 2, 254]) def ensure_metadata_ip(): """Sets up local metadata ip.""" _execute('ip', 'addr', 'add', '169.254.169.254/32', 'scope', 'link', 'dev', 'lo', run_as_root=True, check_exit_code=[0, 2, 254]) def ensure_vpn_forward(public_ip, port, private_ip): """Sets up forwarding rules for vlan.""" iptables_manager.ipv4['filter'].add_rule('FORWARD', '-d %s -p udp ' '--dport 1194 ' '-j ACCEPT' % private_ip) iptables_manager.ipv4['nat'].add_rule('PREROUTING', '-d %s -p udp ' '--dport %s -j DNAT --to %s:1194' % (public_ip, port, private_ip)) iptables_manager.ipv4['nat'].add_rule('OUTPUT', '-d %s -p udp ' '--dport %s -j DNAT --to %s:1194' % (public_ip, port, private_ip)) iptables_manager.apply() def ensure_floating_forward(floating_ip, fixed_ip, device, network): """Ensure floating ip forwarding rule.""" # NOTE(vish): Make sure we never have duplicate rules for the same ip regex = '.*\s+%s(/32|\s+|$)' % floating_ip num_rules = iptables_manager.ipv4['nat'].remove_rules_regex(regex) if num_rules: msg = _LW('Removed %(num)d duplicate rules for floating ip %(float)s') LOG.warn(msg, {'num': num_rules, 'float': floating_ip}) for chain, rule in floating_forward_rules(floating_ip, fixed_ip, device): iptables_manager.ipv4['nat'].add_rule(chain, rule) iptables_manager.apply() if device != network['bridge']: ensure_ebtables_rules(*floating_ebtables_rules(fixed_ip, network)) def remove_floating_forward(floating_ip, fixed_ip, device, network): """Remove forwarding for floating ip.""" for chain, rule in floating_forward_rules(floating_ip, fixed_ip, device): iptables_manager.ipv4['nat'].remove_rule(chain, rule) iptables_manager.apply() if device != network['bridge']: remove_ebtables_rules(*floating_ebtables_rules(fixed_ip, network)) def floating_ebtables_rules(fixed_ip, network): """Makes sure only in-network traffic is bridged.""" return (['PREROUTING --logical-in %s -p ipv4 --ip-src %s ' '! --ip-dst %s -j redirect --redirect-target ACCEPT' % (network['bridge'], fixed_ip, network['cidr'])], 'nat') def floating_forward_rules(floating_ip, fixed_ip, device): rules = [] rule = '-s %s -j SNAT --to %s' % (fixed_ip, floating_ip) if device: rules.append(('float-snat', rule + ' -d %s' % fixed_ip)) rules.append(('float-snat', rule + ' -o %s' % device)) else: rules.append(('float-snat', rule)) rules.append( ('PREROUTING', '-d %s -j DNAT --to %s' % (floating_ip, fixed_ip))) rules.append( ('OUTPUT', '-d %s -j DNAT --to %s' % (floating_ip, fixed_ip))) rules.append(('POSTROUTING', '-s %s -m conntrack --ctstate DNAT -j SNAT ' '--to-source %s' % (fixed_ip, floating_ip))) return rules def clean_conntrack(fixed_ip): try: _execute('conntrack', '-D', '-r', fixed_ip, run_as_root=True, check_exit_code=[0, 1]) except processutils.ProcessExecutionError: LOG.exception(_LE('Error deleting conntrack entries for %s'), fixed_ip) def _enable_ipv4_forwarding(): sysctl_key = 'net.ipv4.ip_forward' stdout, stderr = _execute('sysctl', '-n', sysctl_key) if stdout.strip() is not '1': _execute('sysctl', '-w', '%s=1' % sysctl_key, run_as_root=True) @utils.synchronized('lock_gateway', external=True) def initialize_gateway_device(dev, network_ref): if not network_ref: return _enable_ipv4_forwarding() # NOTE(vish): The ip for dnsmasq has to be the first address on the # bridge for it to respond to requests properly try: prefix = network_ref.cidr.prefixlen except AttributeError: prefix = network_ref['cidr'].rpartition('/')[2] full_ip = '%s/%s' % (network_ref['dhcp_server'], prefix) new_ip_params = [[full_ip, 'brd', network_ref['broadcast']]] old_ip_params = [] out, err = _execute('ip', 'addr', 'show', 'dev', dev, 'scope', 'global') for line in out.split('\n'): fields = line.split() if fields and fields[0] == 'inet': if fields[-2] in ('secondary', 'dynamic'): ip_params = fields[1:-2] else: ip_params = fields[1:-1] old_ip_params.append(ip_params) if ip_params[0] != full_ip: new_ip_params.append(ip_params) if not old_ip_params or old_ip_params[0][0] != full_ip: old_routes = [] result = _execute('ip', 'route', 'show', 'dev', dev) if result: out, err = result for line in out.split('\n'): fields = line.split() if fields and 'via' in fields: old_routes.append(fields) _execute('ip', 'route', 'del', fields[0], 'dev', dev, run_as_root=True) for ip_params in old_ip_params: _execute(*_ip_bridge_cmd('del', ip_params, dev), run_as_root=True, check_exit_code=[0, 2, 254]) for ip_params in new_ip_params: _execute(*_ip_bridge_cmd('add', ip_params, dev), run_as_root=True, check_exit_code=[0, 2, 254]) for fields in old_routes: _execute('ip', 'route', 'add', *fields, run_as_root=True) if CONF.send_arp_for_ha and CONF.send_arp_for_ha_count > 0: send_arp_for_ip(network_ref['dhcp_server'], dev, CONF.send_arp_for_ha_count) if CONF.use_ipv6: _execute('ip', '-f', 'inet6', 'addr', 'change', network_ref['cidr_v6'], 'dev', dev, run_as_root=True) def get_dhcp_leases(context, network_ref): """Return a network's hosts config in dnsmasq leasefile format.""" hosts = [] host = None if network_ref['multi_host']: host = CONF.host for fixedip in objects.FixedIPList.get_by_network(context, network_ref, host=host): # NOTE(cfb): Don't return a lease entry if the IP isn't # already leased if fixedip.leased: hosts.append(_host_lease(fixedip)) return '\n'.join(hosts) def get_dhcp_hosts(context, network_ref, fixedips): """Get network's hosts config in dhcp-host format.""" hosts = [] macs = set() for fixedip in fixedips: if fixedip.allocated: if fixedip.virtual_interface.address not in macs: hosts.append(_host_dhcp(fixedip)) macs.add(fixedip.virtual_interface.address) return '\n'.join(hosts) def get_dns_hosts(context, network_ref): """Get network's DNS hosts in hosts format.""" hosts = [] for fixedip in objects.FixedIPList.get_by_network(context, network_ref): if fixedip.allocated: hosts.append(_host_dns(fixedip)) return '\n'.join(hosts) def _add_dnsmasq_accept_rules(dev): """Allow DHCP and DNS traffic through to dnsmasq.""" table = iptables_manager.ipv4['filter'] for port in [67, 53]: for proto in ['udp', 'tcp']: args = {'dev': dev, 'port': port, 'proto': proto} table.add_rule('INPUT', '-i %(dev)s -p %(proto)s -m %(proto)s ' '--dport %(port)s -j ACCEPT' % args) iptables_manager.apply() def _remove_dnsmasq_accept_rules(dev): """Remove DHCP and DNS traffic allowed through to dnsmasq.""" table = iptables_manager.ipv4['filter'] for port in [67, 53]: for proto in ['udp', 'tcp']: args = {'dev': dev, 'port': port, 'proto': proto} table.remove_rule('INPUT', '-i %(dev)s -p %(proto)s -m %(proto)s ' '--dport %(port)s -j ACCEPT' % args) iptables_manager.apply() # NOTE(russellb) Curious why this is needed? Check out this explanation from # markmc: https://bugzilla.redhat.com/show_bug.cgi?id=910619#c6 def _add_dhcp_mangle_rule(dev): table = iptables_manager.ipv4['mangle'] table.add_rule('POSTROUTING', '-o %s -p udp -m udp --dport 68 -j CHECKSUM ' '--checksum-fill' % dev) iptables_manager.apply() def _remove_dhcp_mangle_rule(dev): table = iptables_manager.ipv4['mangle'] table.remove_rule('POSTROUTING', '-o %s -p udp -m udp --dport 68 -j CHECKSUM ' '--checksum-fill' % dev) iptables_manager.apply() def get_dhcp_opts(context, network_ref, fixedips): """Get network's hosts config in dhcp-opts format.""" gateway = network_ref['gateway'] # NOTE(vish): if we are in multi-host mode and we are not sharing # addresses, then we actually need to hand out the # dhcp server address as the gateway. if network_ref['multi_host'] and not (network_ref['share_address'] or CONF.share_dhcp_address): gateway = network_ref['dhcp_server'] hosts = [] if CONF.use_single_default_gateway: for fixedip in fixedips: if fixedip.allocated: vif_id = fixedip.virtual_interface_id if fixedip.default_route: hosts.append(_host_dhcp_opts(vif_id, gateway)) else: hosts.append(_host_dhcp_opts(vif_id)) else: hosts.append(_host_dhcp_opts(None, gateway)) return '\n'.join(hosts) def release_dhcp(dev, address, mac_address): if device_exists(dev): try: utils.execute('dhcp_release', dev, address, mac_address, run_as_root=True) except processutils.ProcessExecutionError: raise exception.NetworkDhcpReleaseFailed(address=address, mac_address=mac_address) def update_dhcp(context, dev, network_ref): conffile = _dhcp_file(dev, 'conf') host = None if network_ref['multi_host']: host = CONF.host fixedips = objects.FixedIPList.get_by_network(context, network_ref, host=host) write_to_file(conffile, get_dhcp_hosts(context, network_ref, fixedips)) restart_dhcp(context, dev, network_ref, fixedips) def update_dns(context, dev, network_ref): hostsfile = _dhcp_file(dev, 'hosts') host = None if network_ref['multi_host']: host = CONF.host fixedips = objects.FixedIPList.get_by_network(context, network_ref, host=host) write_to_file(hostsfile, get_dns_hosts(context, network_ref)) restart_dhcp(context, dev, network_ref, fixedips) def update_dhcp_hostfile_with_text(dev, hosts_text): conffile = _dhcp_file(dev, 'conf') write_to_file(conffile, hosts_text) def kill_dhcp(dev): pid = _dnsmasq_pid_for(dev) if pid: # Check that the process exists and looks like a dnsmasq process conffile = _dhcp_file(dev, 'conf') if is_pid_cmdline_correct(pid, conffile.split('/')[-1]): _execute('kill', '-9', pid, run_as_root=True) else: LOG.debug('Pid %d is stale, skip killing dnsmasq', pid) _remove_dnsmasq_accept_rules(dev) _remove_dhcp_mangle_rule(dev) # NOTE(ja): Sending a HUP only reloads the hostfile, so any # configuration options (like dchp-range, vlan, ...) # aren't reloaded. @utils.synchronized('dnsmasq_start') def restart_dhcp(context, dev, network_ref, fixedips): """(Re)starts a dnsmasq server for a given network. If a dnsmasq instance is already running then send a HUP signal causing it to reload, otherwise spawn a new instance. """ conffile = _dhcp_file(dev, 'conf') optsfile = _dhcp_file(dev, 'opts') write_to_file(optsfile, get_dhcp_opts(context, network_ref, fixedips)) os.chmod(optsfile, 0o644) _add_dhcp_mangle_rule(dev) # Make sure dnsmasq can actually read it (it setuid()s to "nobody") os.chmod(conffile, 0o644) pid = _dnsmasq_pid_for(dev) # if dnsmasq is already running, then tell it to reload if pid: if is_pid_cmdline_correct(pid, conffile.split('/')[-1]): try: _execute('kill', '-HUP', pid, run_as_root=True) _add_dnsmasq_accept_rules(dev) return except Exception as exc: LOG.error(_LE('kill -HUP dnsmasq threw %s'), exc) else: LOG.debug('Pid %d is stale, relaunching dnsmasq', pid) cmd = ['env', 'CONFIG_FILE=%s' % jsonutils.dumps(CONF.dhcpbridge_flagfile), 'NETWORK_ID=%s' % str(network_ref['id']), 'dnsmasq', '--strict-order', '--bind-interfaces', '--conf-file=%s' % CONF.dnsmasq_config_file, '--pid-file=%s' % _dhcp_file(dev, 'pid'), '--dhcp-optsfile=%s' % _dhcp_file(dev, 'opts'), '--listen-address=%s' % network_ref['dhcp_server'], '--except-interface=lo', '--dhcp-range=set:%s,%s,static,%s,%ss' % (network_ref['label'], network_ref['dhcp_start'], network_ref['netmask'], CONF.dhcp_lease_time), '--dhcp-lease-max=%s' % len(netaddr.IPNetwork(network_ref['cidr'])), '--dhcp-hostsfile=%s' % _dhcp_file(dev, 'conf'), '--dhcp-script=%s' % CONF.dhcpbridge, '--no-hosts', '--leasefile-ro'] # dnsmasq currently gives an error for an empty domain, # rather than ignoring. So only specify it if defined. if CONF.dhcp_domain: cmd.append('--domain=%s' % CONF.dhcp_domain) dns_servers = CONF.dns_server if CONF.use_network_dns_servers: if network_ref.get('dns1'): dns_servers.append(network_ref.get('dns1')) if network_ref.get('dns2'): dns_servers.append(network_ref.get('dns2')) if network_ref['multi_host']: cmd.append('--addn-hosts=%s' % _dhcp_file(dev, 'hosts')) if dns_servers: cmd.append('--no-resolv') for dns_server in dns_servers: cmd.append('--server=%s' % dns_server) _execute(*cmd, run_as_root=True) _add_dnsmasq_accept_rules(dev) @utils.synchronized('radvd_start') def update_ra(context, dev, network_ref): conffile = _ra_file(dev, 'conf') conf_str = """ interface %s { AdvSendAdvert on; MinRtrAdvInterval 3; MaxRtrAdvInterval 10; prefix %s { AdvOnLink on; AdvAutonomous on; }; }; """ % (dev, network_ref['cidr_v6']) write_to_file(conffile, conf_str) # Make sure radvd can actually read it (it setuid()s to "nobody") os.chmod(conffile, 0o644) pid = _ra_pid_for(dev) # if radvd is already running, then tell it to reload if pid: if is_pid_cmdline_correct(pid, conffile): try: _execute('kill', pid, run_as_root=True) except Exception as exc: LOG.error(_LE('killing radvd threw %s'), exc) else: LOG.debug('Pid %d is stale, relaunching radvd', pid) cmd = ['radvd', '-C', '%s' % _ra_file(dev, 'conf'), '-p', '%s' % _ra_file(dev, 'pid')] _execute(*cmd, run_as_root=True) def _host_lease(fixedip): """Return a host string for an address in leasefile format.""" timestamp = timeutils.utcnow() seconds_since_epoch = calendar.timegm(timestamp.utctimetuple()) return '%d %s %s %s *' % (seconds_since_epoch + CONF.dhcp_lease_time, fixedip.virtual_interface.address, fixedip.address, fixedip.instance.hostname or '*') def _host_dhcp_network(vif_id): return 'NW-%s' % vif_id def _host_dhcp(fixedip): """Return a host string for an address in dhcp-host format.""" # NOTE(cfb): dnsmasq on linux only supports 64 characters in the hostname # field (LP #1238910). Since the . counts as a character we need # to truncate the hostname to only 63 characters. hostname = fixedip.instance.hostname if len(hostname) > 63: LOG.warning(_LW('hostname %s too long, truncating.') % (hostname)) hostname = fixedip.instance.hostname[:2] + '-' +\ fixedip.instance.hostname[-60:] if CONF.use_single_default_gateway: net = _host_dhcp_network(fixedip.virtual_interface_id) return '%s,%s.%s,%s,net:%s' % (fixedip.virtual_interface.address, hostname, CONF.dhcp_domain, fixedip.address, net) else: return '%s,%s.%s,%s' % (fixedip.virtual_interface.address, hostname, CONF.dhcp_domain, fixedip.address) def _host_dns(fixedip): return '%s\t%s.%s' % (fixedip.address, fixedip.instance.hostname, CONF.dhcp_domain) def _host_dhcp_opts(vif_id=None, gateway=None): """Return an empty gateway option.""" values = [] if vif_id is not None: values.append(_host_dhcp_network(vif_id)) # NOTE(vish): 3 is the dhcp option for gateway. values.append('3') if gateway: values.append('%s' % gateway) return ','.join(values) def _execute(*cmd, **kwargs): """Wrapper around utils._execute for fake_network.""" if CONF.fake_network: LOG.debug('FAKE NET: %s', ' '.join(map(str, cmd))) return 'fake', 0 else: return utils.execute(*cmd, **kwargs) def device_exists(device): """Check if ethernet device exists.""" return os.path.exists('/sys/class/net/%s' % device) def _dhcp_file(dev, kind): """Return path to a pid, leases, hosts or conf file for a bridge/device.""" fileutils.ensure_tree(CONF.networks_path) return os.path.abspath('%s/nova-%s.%s' % (CONF.networks_path, dev, kind)) def _ra_file(dev, kind): """Return path to a pid or conf file for a bridge/device.""" fileutils.ensure_tree(CONF.networks_path) return os.path.abspath('%s/nova-ra-%s.%s' % (CONF.networks_path, dev, kind)) def _dnsmasq_pid_for(dev): """Returns the pid for prior dnsmasq instance for a bridge/device. Returns None if no pid file exists. If machine has rebooted pid might be incorrect (caller should check). """ pid_file = _dhcp_file(dev, 'pid') if os.path.exists(pid_file): try: with open(pid_file, 'r') as f: return int(f.read()) except (ValueError, IOError): return None def _ra_pid_for(dev): """Returns the pid for prior radvd instance for a bridge/device. Returns None if no pid file exists. If machine has rebooted pid might be incorrect (caller should check). """ pid_file = _ra_file(dev, 'pid') if os.path.exists(pid_file): with open(pid_file, 'r') as f: return int(f.read()) def _ip_bridge_cmd(action, params, device): """Build commands to add/del ips to bridges/devices.""" cmd = ['ip', 'addr', action] cmd.extend(params) cmd.extend(['dev', device]) return cmd def _set_device_mtu(dev, mtu=None): """Set the device MTU.""" if not mtu: mtu = CONF.network_device_mtu if mtu: utils.execute('ip', 'link', 'set', dev, 'mtu', mtu, run_as_root=True, check_exit_code=[0, 1, 2, 254]) def _create_veth_pair(dev1_name, dev2_name): """Create a pair of veth devices with the specified names, deleting any previous devices with those names. """ for dev in [dev1_name, dev2_name]: delete_net_dev(dev) utils.execute('ip', 'link', 'add', dev1_name, 'type', 'veth', 'peer', 'name', dev2_name, run_as_root=True) for dev in [dev1_name, dev2_name]: utils.execute('ip', 'link', 'set', dev, 'up', run_as_root=True) utils.execute('ip', 'link', 'set', dev, 'promisc', 'on', run_as_root=True) _set_device_mtu(dev) def _ovs_vsctl(args): full_args = ['ovs-vsctl', '--timeout=%s' % CONF.ovs_vsctl_timeout] + args try: return utils.execute(*full_args, run_as_root=True) except Exception as e: LOG.error(_LE("Unable to execute %(cmd)s. Exception: %(exception)s"), {'cmd': full_args, 'exception': e}) raise exception.AgentError(method=full_args) def create_ovs_vif_port(bridge, dev, iface_id, mac, instance_id): _ovs_vsctl(['--', '--if-exists', 'del-port', dev, '--', 'add-port', bridge, dev, '--', 'set', 'Interface', dev, 'external-ids:iface-id=%s' % iface_id, 'external-ids:iface-status=active', 'external-ids:attached-mac=%s' % mac, 'external-ids:vm-uuid=%s' % instance_id]) _set_device_mtu(dev) def delete_ovs_vif_port(bridge, dev): _ovs_vsctl(['--', '--if-exists', 'del-port', bridge, dev]) delete_net_dev(dev) def ovs_set_vhostuser_port_type(dev): _ovs_vsctl(['--', 'set', 'Interface', dev, 'type=dpdkvhostuser']) def create_ivs_vif_port(dev, iface_id, mac, instance_id): utils.execute('ivs-ctl', 'add-port', dev, run_as_root=True) def delete_ivs_vif_port(dev): utils.execute('ivs-ctl', 'del-port', dev, run_as_root=True) utils.execute('ip', 'link', 'delete', dev, run_as_root=True) def create_tap_dev(dev, mac_address=None): if not device_exists(dev): try: # First, try with 'ip' utils.execute('ip', 'tuntap', 'add', dev, 'mode', 'tap', run_as_root=True, check_exit_code=[0, 2, 254]) except processutils.ProcessExecutionError: # Second option: tunctl utils.execute('tunctl', '-b', '-t', dev, run_as_root=True) if mac_address: utils.execute('ip', 'link', 'set', dev, 'address', mac_address, run_as_root=True, check_exit_code=[0, 2, 254]) utils.execute('ip', 'link', 'set', dev, 'up', run_as_root=True, check_exit_code=[0, 2, 254]) def delete_net_dev(dev): """Delete a network device only if it exists.""" if device_exists(dev): try: utils.execute('ip', 'link', 'delete', dev, run_as_root=True, check_exit_code=[0, 2, 254]) LOG.debug("Net device removed: '%s'", dev) except processutils.ProcessExecutionError: with excutils.save_and_reraise_exception(): LOG.error(_LE("Failed removing net device: '%s'"), dev) def delete_bridge_dev(dev): """Delete a network bridge.""" if device_exists(dev): try: utils.execute('ip', 'link', 'set', dev, 'down', run_as_root=True) utils.execute('brctl', 'delbr', dev, run_as_root=True) except processutils.ProcessExecutionError: with excutils.save_and_reraise_exception(): LOG.error(_LE("Failed removing bridge device: '%s'"), dev) # Similar to compute virt layers, the Linux network node # code uses a flexible driver model to support different ways # of creating ethernet interfaces and attaching them to the network. # In the case of a network host, these interfaces # act as gateway/dhcp/vpn/etc. endpoints not VM interfaces. interface_driver = None def _get_interface_driver(): global interface_driver if not interface_driver: interface_driver = importutils.import_object( CONF.linuxnet_interface_driver) return interface_driver def plug(network, mac_address, gateway=True): return _get_interface_driver().plug(network, mac_address, gateway) def unplug(network): return _get_interface_driver().unplug(network) def get_dev(network): return _get_interface_driver().get_dev(network) class LinuxNetInterfaceDriver(object): """Abstract class that defines generic network host API for all Linux interface drivers. """ def plug(self, network, mac_address): """Create Linux device, return device name.""" raise NotImplementedError() def unplug(self, network): """Destroy Linux device, return device name.""" raise NotImplementedError() def get_dev(self, network): """Get device name.""" raise NotImplementedError() # plugs interfaces using Linux Bridge class LinuxBridgeInterfaceDriver(LinuxNetInterfaceDriver): def plug(self, network, mac_address, gateway=True): vlan = network.get('vlan') if vlan is not None: iface = CONF.vlan_interface or network['bridge_interface'] LinuxBridgeInterfaceDriver.ensure_vlan_bridge( vlan, network['bridge'], iface, network, mac_address, network.get('mtu')) iface = 'vlan%s' % vlan else: iface = CONF.flat_interface or network['bridge_interface'] LinuxBridgeInterfaceDriver.ensure_bridge( network['bridge'], iface, network, gateway) if network['share_address'] or CONF.share_dhcp_address: isolate_dhcp_address(iface, network['dhcp_server']) # NOTE(vish): applying here so we don't get a lock conflict iptables_manager.apply() return network['bridge'] def unplug(self, network, gateway=True): vlan = network.get('vlan') if vlan is not None: iface = 'vlan%s' % vlan LinuxBridgeInterfaceDriver.remove_vlan_bridge(vlan, network['bridge']) else: iface = CONF.flat_interface or network['bridge_interface'] LinuxBridgeInterfaceDriver.remove_bridge(network['bridge'], gateway) if network['share_address'] or CONF.share_dhcp_address: remove_isolate_dhcp_address(iface, network['dhcp_server']) iptables_manager.apply() return self.get_dev(network) def get_dev(self, network): return network['bridge'] @staticmethod def ensure_vlan_bridge(vlan_num, bridge, bridge_interface, net_attrs=None, mac_address=None, mtu=None): """Create a vlan and bridge unless they already exist.""" interface = LinuxBridgeInterfaceDriver.ensure_vlan(vlan_num, bridge_interface, mac_address, mtu) LinuxBridgeInterfaceDriver.ensure_bridge(bridge, interface, net_attrs) return interface @staticmethod def remove_vlan_bridge(vlan_num, bridge): """Delete a bridge and vlan.""" LinuxBridgeInterfaceDriver.remove_bridge(bridge) LinuxBridgeInterfaceDriver.remove_vlan(vlan_num) @staticmethod @utils.synchronized('lock_vlan', external=True) def ensure_vlan(vlan_num, bridge_interface, mac_address=None, mtu=None, interface=None): """Create a vlan unless it already exists.""" if interface is None: interface = 'vlan%s' % vlan_num if not device_exists(interface): LOG.debug('Starting VLAN interface %s', interface) _execute('ip', 'link', 'add', 'link', bridge_interface, 'name', interface, 'type', 'vlan', 'id', vlan_num, run_as_root=True, check_exit_code=[0, 2, 254]) # (danwent) the bridge will inherit this address, so we want to # make sure it is the value set from the NetworkManager if mac_address: _execute('ip', 'link', 'set', interface, 'address', mac_address, run_as_root=True, check_exit_code=[0, 2, 254]) _execute('ip', 'link', 'set', interface, 'up', run_as_root=True, check_exit_code=[0, 2, 254]) # NOTE(vish): set mtu every time to ensure that changes to mtu get # propogated _set_device_mtu(interface, mtu) return interface @staticmethod @utils.synchronized('lock_vlan', external=True) def remove_vlan(vlan_num): """Delete a vlan.""" vlan_interface = 'vlan%s' % vlan_num delete_net_dev(vlan_interface) @staticmethod @utils.synchronized('lock_bridge', external=True) def ensure_bridge(bridge, interface, net_attrs=None, gateway=True, filtering=True): """Create a bridge unless it already exists. :param interface: the interface to create the bridge on. :param net_attrs: dictionary with attributes used to create bridge. :param gateway: whether or not the bridge is a gateway. :param filtering: whether or not to create filters on the bridge. If net_attrs is set, it will add the net_attrs['gateway'] to the bridge using net_attrs['broadcast'] and net_attrs['cidr']. It will also add the ip_v6 address specified in net_attrs['cidr_v6'] if use_ipv6 is set. The code will attempt to move any ips that already exist on the interface onto the bridge and reset the default gateway if necessary. """ if not device_exists(bridge): LOG.debug('Starting Bridge %s', bridge) out, err = _execute('brctl', 'addbr', bridge, check_exit_code=False, run_as_root=True) if (err and err != "device %s already exists; can't create " "bridge with the same name\n" % (bridge)): msg = _('Failed to add bridge: %s') % err raise exception.NovaException(msg) _execute('brctl', 'setfd', bridge, 0, run_as_root=True) # _execute('brctl setageing %s 10' % bridge, run_as_root=True) _execute('brctl', 'stp', bridge, 'off', run_as_root=True) _execute('ip', 'link', 'set', bridge, 'up', run_as_root=True) if interface: LOG.debug('Adding interface %(interface)s to bridge %(bridge)s', {'interface': interface, 'bridge': bridge}) out, err = _execute('brctl', 'addif', bridge, interface, check_exit_code=False, run_as_root=True) if (err and err != "device %s is already a member of a bridge; " "can't enslave it to bridge %s.\n" % (interface, bridge)): msg = _('Failed to add interface: %s') % err raise exception.NovaException(msg) # NOTE(apmelton): Linux bridge's default behavior is to use the # lowest mac of all plugged interfaces. This isn't a problem when # it is first created and the only interface is the bridged # interface. But, as instance interfaces are plugged, there is a # chance for the mac to change. So, set it here so that it won't # change in the future. if not CONF.fake_network: interface_addrs = netifaces.ifaddresses(interface) interface_mac = interface_addrs[netifaces.AF_LINK][0]['addr'] _execute('ip', 'link', 'set', bridge, 'address', interface_mac, run_as_root=True) out, err = _execute('ip', 'link', 'set', interface, 'up', check_exit_code=False, run_as_root=True) # NOTE(vish): This will break if there is already an ip on the # interface, so we move any ips to the bridge # NOTE(danms): We also need to copy routes to the bridge so as # not to break existing connectivity on the interface old_routes = [] out, err = _execute('ip', 'route', 'show', 'dev', interface) for line in out.split('\n'): fields = line.split() if fields and 'via' in fields: old_routes.append(fields) _execute('ip', 'route', 'del', *fields, run_as_root=True) out, err = _execute('ip', 'addr', 'show', 'dev', interface, 'scope', 'global') for line in out.split('\n'): fields = line.split() if fields and fields[0] == 'inet': if fields[-2] in ('secondary', 'dynamic', ): params = fields[1:-2] else: params = fields[1:-1] _execute(*_ip_bridge_cmd('del', params, fields[-1]), run_as_root=True, check_exit_code=[0, 2, 254]) _execute(*_ip_bridge_cmd('add', params, bridge), run_as_root=True, check_exit_code=[0, 2, 254]) for fields in old_routes: _execute('ip', 'route', 'add', *fields, run_as_root=True) if filtering: # Don't forward traffic unless we were told to be a gateway ipv4_filter = iptables_manager.ipv4['filter'] if gateway: for rule in get_gateway_rules(bridge): ipv4_filter.add_rule(*rule) else: ipv4_filter.add_rule('FORWARD', ('--in-interface %s -j %s' % (bridge, CONF.iptables_drop_action))) ipv4_filter.add_rule('FORWARD', ('--out-interface %s -j %s' % (bridge, CONF.iptables_drop_action))) @staticmethod @utils.synchronized('lock_bridge', external=True) def remove_bridge(bridge, gateway=True, filtering=True): """Delete a bridge.""" if not device_exists(bridge): return else: if filtering: ipv4_filter = iptables_manager.ipv4['filter'] if gateway: for rule in get_gateway_rules(bridge): ipv4_filter.remove_rule(*rule) else: drop_actions = ['DROP'] if CONF.iptables_drop_action != 'DROP': drop_actions.append(CONF.iptables_drop_action) for drop_action in drop_actions: ipv4_filter.remove_rule('FORWARD', ('--in-interface %s -j %s' % (bridge, drop_action))) ipv4_filter.remove_rule('FORWARD', ('--out-interface %s -j %s' % (bridge, drop_action))) delete_bridge_dev(bridge) # NOTE(cfb): This is a temporary fix to LP #1316621. We really want to call # ebtables with --concurrent. In order to do that though we need # libvirt to support this. Additionally since ebtables --concurrent # will hang indefinitely waiting on the lock we need to teach # oslo_concurrency.processutils how to timeout a long running # process first. Once those are complete we can replace all of this # with calls to ebtables --concurrent and a reasonable timeout. def _exec_ebtables(*cmd, **kwargs): check_exit_code = kwargs.pop('check_exit_code', True) # List of error strings to re-try. retry_strings = ( 'Multiple ebtables programs', ) # We always try at least once attempts = CONF.ebtables_exec_attempts if attempts <= 0: attempts = 1 count = 1 while count <= attempts: # Updated our counters if needed sleep = CONF.ebtables_retry_interval * count count += 1 # NOTE(cfb): ebtables reports all errors with a return code of 255. # As such we can't know if we hit a locking error, or some # other error (like a rule doesn't exist) so we have to # to parse stderr. try: _execute(*cmd, check_exit_code=[0], **kwargs) except processutils.ProcessExecutionError as exc: # See if we can retry the error. if any(error in exc.stderr for error in retry_strings): if count > attempts and check_exit_code: LOG.warning(_LW('%s failed. Not Retrying.'), ' '.join(cmd)) raise else: # We need to sleep a bit before retrying LOG.warning(_LW("%(cmd)s failed. Sleeping %(time)s " "seconds before retry."), {'cmd': ' '.join(cmd), 'time': sleep}) time.sleep(sleep) else: # Not eligible for retry if check_exit_code: LOG.warning(_LW('%s failed. Not Retrying.'), ' '.join(cmd)) raise else: return else: # Success return @utils.synchronized('ebtables', external=True) def ensure_ebtables_rules(rules, table='filter'): for rule in rules: cmd = ['ebtables', '-t', table, '-D'] + rule.split() _exec_ebtables(*cmd, check_exit_code=False, run_as_root=True) cmd[3] = '-I' _exec_ebtables(*cmd, run_as_root=True) @utils.synchronized('ebtables', external=True) def remove_ebtables_rules(rules, table='filter'): for rule in rules: cmd = ['ebtables', '-t', table, '-D'] + rule.split() _exec_ebtables(*cmd, check_exit_code=False, run_as_root=True) def isolate_dhcp_address(interface, address): # block arp traffic to address across the interface rules = [] rules.append('INPUT -p ARP -i %s --arp-ip-dst %s -j DROP' % (interface, address)) rules.append('OUTPUT -p ARP -o %s --arp-ip-src %s -j DROP' % (interface, address)) rules.append('FORWARD -p IPv4 -i %s --ip-protocol udp ' '--ip-destination-port 67:68 -j DROP' % interface) rules.append('FORWARD -p IPv4 -o %s --ip-protocol udp ' '--ip-destination-port 67:68 -j DROP' % interface) # NOTE(vish): the above is not possible with iptables/arptables ensure_ebtables_rules(rules) def remove_isolate_dhcp_address(interface, address): # block arp traffic to address across the interface rules = [] rules.append('INPUT -p ARP -i %s --arp-ip-dst %s -j DROP' % (interface, address)) rules.append('OUTPUT -p ARP -o %s --arp-ip-src %s -j DROP' % (interface, address)) rules.append('FORWARD -p IPv4 -i %s --ip-protocol udp ' '--ip-destination-port 67:68 -j DROP' % interface) rules.append('FORWARD -p IPv4 -o %s --ip-protocol udp ' '--ip-destination-port 67:68 -j DROP' % interface) remove_ebtables_rules(rules) # NOTE(vish): the above is not possible with iptables/arptables def get_gateway_rules(bridge): interfaces = CONF.forward_bridge_interface if 'all' in interfaces: return [('FORWARD', '-i %s -j ACCEPT' % bridge), ('FORWARD', '-o %s -j ACCEPT' % bridge)] rules = [] for iface in CONF.forward_bridge_interface: if iface: rules.append(('FORWARD', '-i %s -o %s -j ACCEPT' % (bridge, iface))) rules.append(('FORWARD', '-i %s -o %s -j ACCEPT' % (iface, bridge))) rules.append(('FORWARD', '-i %s -o %s -j ACCEPT' % (bridge, bridge))) rules.append(('FORWARD', '-i %s -j %s' % (bridge, CONF.iptables_drop_action))) rules.append(('FORWARD', '-o %s -j %s' % (bridge, CONF.iptables_drop_action))) return rules # plugs interfaces using Open vSwitch class LinuxOVSInterfaceDriver(LinuxNetInterfaceDriver): def plug(self, network, mac_address, gateway=True): dev = self.get_dev(network) if not device_exists(dev): bridge = CONF.linuxnet_ovs_integration_bridge _ovs_vsctl(['--', '--may-exist', 'add-port', bridge, dev, '--', 'set', 'Interface', dev, 'type=internal', '--', 'set', 'Interface', dev, 'external-ids:iface-id=%s' % dev, '--', 'set', 'Interface', dev, 'external-ids:iface-status=active', '--', 'set', 'Interface', dev, 'external-ids:attached-mac=%s' % mac_address]) _execute('ip', 'link', 'set', dev, 'address', mac_address, run_as_root=True) _set_device_mtu(dev, network.get('mtu')) _execute('ip', 'link', 'set', dev, 'up', run_as_root=True) if not gateway: # If we weren't instructed to act as a gateway then add the # appropriate flows to block all non-dhcp traffic. _execute('ovs-ofctl', 'add-flow', bridge, 'priority=1,actions=drop', run_as_root=True) _execute('ovs-ofctl', 'add-flow', bridge, 'udp,tp_dst=67,dl_dst=%s,priority=2,actions=normal' % mac_address, run_as_root=True) # .. and make sure iptbles won't forward it as well. iptables_manager.ipv4['filter'].add_rule('FORWARD', '--in-interface %s -j %s' % (bridge, CONF.iptables_drop_action)) iptables_manager.ipv4['filter'].add_rule('FORWARD', '--out-interface %s -j %s' % (bridge, CONF.iptables_drop_action)) else: for rule in get_gateway_rules(bridge): iptables_manager.ipv4['filter'].add_rule(*rule) return dev def unplug(self, network): dev = self.get_dev(network) bridge = CONF.linuxnet_ovs_integration_bridge _ovs_vsctl(['--', '--if-exists', 'del-port', bridge, dev]) return dev def get_dev(self, network): dev = 'gw-' + str(network['uuid'][0:11]) return dev # plugs interfaces using Linux Bridge when using NeutronManager class NeutronLinuxBridgeInterfaceDriver(LinuxNetInterfaceDriver): BRIDGE_NAME_PREFIX = 'brq' GATEWAY_INTERFACE_PREFIX = 'gw-' def plug(self, network, mac_address, gateway=True): dev = self.get_dev(network) bridge = self.get_bridge(network) if not gateway: # If we weren't instructed to act as a gateway then add the # appropriate flows to block all non-dhcp traffic. # .. and make sure iptbles won't forward it as well. iptables_manager.ipv4['filter'].add_rule('FORWARD', ('--in-interface %s -j %s' % (bridge, CONF.iptables_drop_action))) iptables_manager.ipv4['filter'].add_rule('FORWARD', ('--out-interface %s -j %s' % (bridge, CONF.iptables_drop_action))) return bridge else: for rule in get_gateway_rules(bridge): iptables_manager.ipv4['filter'].add_rule(*rule) create_tap_dev(dev, mac_address) if not device_exists(bridge): LOG.debug("Starting bridge %s ", bridge) utils.execute('brctl', 'addbr', bridge, run_as_root=True) utils.execute('brctl', 'setfd', bridge, str(0), run_as_root=True) utils.execute('brctl', 'stp', bridge, 'off', run_as_root=True) utils.execute('ip', 'link', 'set', bridge, 'address', mac_address, run_as_root=True, check_exit_code=[0, 2, 254]) utils.execute('ip', 'link', 'set', bridge, 'up', run_as_root=True, check_exit_code=[0, 2, 254]) LOG.debug("Done starting bridge %s", bridge) full_ip = '%s/%s' % (network['dhcp_server'], network['cidr'].rpartition('/')[2]) utils.execute('ip', 'address', 'add', full_ip, 'dev', bridge, run_as_root=True, check_exit_code=[0, 2, 254]) return dev def unplug(self, network): dev = self.get_dev(network) if not device_exists(dev): return None else: delete_net_dev(dev) return dev def get_dev(self, network): dev = self.GATEWAY_INTERFACE_PREFIX + str(network['uuid'][0:11]) return dev def get_bridge(self, network): bridge = self.BRIDGE_NAME_PREFIX + str(network['uuid'][0:11]) return bridge # provide compatibility with existing configs QuantumLinuxBridgeInterfaceDriver = NeutronLinuxBridgeInterfaceDriver iptables_manager = IptablesManager() def set_vf_interface_vlan(pci_addr, mac_addr, vlan=0): pf_ifname = pci_utils.get_ifname_by_pci_address(pci_addr, pf_interface=True) vf_ifname = pci_utils.get_ifname_by_pci_address(pci_addr) vf_num = pci_utils.get_vf_num_by_pci_address(pci_addr) # Set the VF's mac address and vlan exit_code = [0, 2, 254] port_state = 'up' if vlan > 0 else 'down' utils.execute('ip', 'link', 'set', pf_ifname, 'vf', vf_num, 'mac', mac_addr, 'vlan', vlan, run_as_root=True, check_exit_code=exit_code) # Bring up/down the VF's interface utils.execute('ip', 'link', 'set', vf_ifname, port_state, run_as_root=True, check_exit_code=exit_code)