diff options
| author | Devin Carlen <devin.carlen@gmail.com> | 2011-02-12 18:34:36 -0800 |
|---|---|---|
| committer | Devin Carlen <devin.carlen@gmail.com> | 2011-02-12 18:34:36 -0800 |
| commit | 2e75026ef11dfac37f1449a54c0d6f85ff8bfca6 (patch) | |
| tree | 5af966fce0491b62f3e2871472dfbce1cf8dfc71 | |
| parent | 6f6a09216458ffdba17d1960bbad723a3e71e7b2 (diff) | |
| parent | 2df3182b6e9637fe0e9ce9358a60ee874a97acb3 (diff) | |
| download | nova-2e75026ef11dfac37f1449a54c0d6f85ff8bfca6.tar.gz nova-2e75026ef11dfac37f1449a54c0d6f85ff8bfca6.tar.xz nova-2e75026ef11dfac37f1449a54c0d6f85ff8bfca6.zip | |
Merged trunk
| -rw-r--r-- | .mailmap | 1 | ||||
| -rw-r--r-- | Authors | 2 | ||||
| -rw-r--r-- | doc/source/adminguide/multi.node.install.rst | 38 | ||||
| -rw-r--r-- | nova/api/ec2/cloud.py | 7 | ||||
| -rw-r--r-- | nova/compute/instance_types.py | 4 | ||||
| -rw-r--r-- | nova/db/sqlalchemy/api.py | 8 | ||||
| -rw-r--r-- | nova/db/sqlalchemy/session.py | 10 | ||||
| -rw-r--r-- | nova/flags.py | 6 | ||||
| -rw-r--r-- | nova/network/linux_net.py | 84 | ||||
| -rw-r--r-- | nova/network/manager.py | 6 | ||||
| -rw-r--r-- | nova/rpc.py | 6 | ||||
| -rw-r--r-- | nova/utils.py | 36 | ||||
| -rw-r--r-- | nova/volume/driver.py | 10 | ||||
| -rw-r--r-- | nova/volume/manager.py | 2 | ||||
| -rw-r--r-- | nova/volume/san.py | 335 |
15 files changed, 515 insertions, 40 deletions
@@ -33,3 +33,4 @@ <corywright@gmail.com> <cory.wright@rackspace.com> <ant@openstack.org> <amesserl@rackspace.com> <chiradeep@cloud.com> <chiradeep@chiradeep-lt2> +<justin@fathomdb.com> <superstack@superstack.org> @@ -3,9 +3,11 @@ Anne Gentle <anne@openstack.org> Anthony Young <sleepsonthefloor@gmail.com> Antony Messerli <ant@openstack.org> Armando Migliaccio <Armando.Migliaccio@eu.citrix.com> +Bilal Akhtar <bilalakhtar@ubuntu.com> Chiradeep Vittal <chiradeep@cloud.com> Chmouel Boudjnah <chmouel@chmouel.com> Chris Behrens <cbehrens@codestud.com> +Christian Berendt <berendt@b1-systems.de> Cory Wright <corywright@gmail.com> David Pravec <David.Pravec@danix.org> Dan Prince <dan.prince@rackspace.com> diff --git a/doc/source/adminguide/multi.node.install.rst b/doc/source/adminguide/multi.node.install.rst index f2f25b060..c53455e3e 100644 --- a/doc/source/adminguide/multi.node.install.rst +++ b/doc/source/adminguide/multi.node.install.rst @@ -37,20 +37,22 @@ From a server you intend to use as a cloud controller node, use this command to :: - wget --no-check-certificate https://github.com/dubsquared/OpenStack-NOVA-Installer-Script/raw/master/Nova_CC_Installer_v0.1 + wget --no-check-certificate https://github.com/dubsquared/OpenStack-NOVA-Installer-Script/raw/master/nova-CC-install-v1.1.sh Ensure you can execute the script by modifying the permissions on the script file. :: - sudo chmod 755 Nova_CC_Installer_v0.1 + sudo chmod 755 nova-CC-install-v1.1.sh :: - sudo ./Nova_CC_Installer_v0.1 + sudo ./nova-CC-install-v1.1.sh -Next, from a server you intend to use as a compute node (doesn't contain the database), install the nova services. Copy the nova.conf from the cloud controller node to the compute node. +Next, from a server you intend to use as a compute node (doesn't contain the database), install the nova services. You can use the nova-NODE-installer.sh script from the above github-hosted project for the compute node installation. + +Copy the nova.conf from the cloud controller node to the compute node. Restart related services:: @@ -247,7 +249,7 @@ Here is an example of what this looks like with real data:: Note: The nova-manage service assumes that the first IP address is your network (like 192.168.0.0), that the 2nd IP is your gateway (192.168.0.1), and that the broadcast is the very last IP in the range you defined (192.168.0.255). If this is not the case you will need to manually edit the sql db 'networks' table.o. -On running this command, entries are made in the 'networks' and 'fixed_ips' table. However, one of the networks listed in the 'networks' table needs to be marked as bridge in order for the code to know that a bridge exists. The Network is marked as bridged automatically based on the type of network manager selected. This is ONLY necessary if you chose FlatManager as your network type. More information can be found at the end of this document discussing setting up the bridge device. +On running the "nova-manage network create" command, entries are made in the 'networks' and 'fixed_ips' table. However, one of the networks listed in the 'networks' table needs to be marked as bridge in order for the code to know that a bridge exists. The Network is marked as bridged automatically based on the type of network manager selected. You only need to mark the network as a bridge if you chose FlatManager as your network type. More information can be found at the end of this document discussing setting up the bridge device. Step 2 - Create Nova certifications @@ -288,9 +290,35 @@ Another common issue is you cannot ping or SSH your instances after issusing the killall dnsmasq service nova-network restart +To avoid issues with KVM and permissions with Nova, run the following commands to ensure we have VM's that are running optimally:: + + chgrp kvm /dev/kvm + chmod g+rwx /dev/kvm + +If you want to use the 10.04 Ubuntu Enterprise Cloud images that are readily available at http://uec-images.ubuntu.com/releases/10.04/release/, you may run into delays with booting. Any server that does not have nova-api running on it needs this iptables entry so that UEC images can get metadata info. On compute nodes, configure the iptables with this next step:: + + # iptables -t nat -A PREROUTING -d 169.254.169.254/32 -p tcp -m tcp --dport 80 -j DNAT --to-destination $NOVA_API_IP:8773 + Testing the Installation ```````````````````````` +You can confirm that your compute node is talking to your cloud controller. From the cloud controller, run this database query:: + + mysql -u$MYSQL_USER -p$MYSQL_PASS nova -e 'select * from services;' + +In return, you should see something similar to this:: + +---------------------+---------------------+------------+---------+----+----------+----------------+-----------+--------------+----------+-------------------+ + | created_at | updated_at | deleted_at | deleted | id | host | binary | topic | report_count | disabled | availability_zone | + +---------------------+---------------------+------------+---------+----+----------+----------------+-----------+--------------+----------+-------------------+ + | 2011-01-28 22:52:46 | 2011-02-03 06:55:48 | NULL | 0 | 1 | osdemo02 | nova-network | network | 46064 | 0 | nova | + | 2011-01-28 22:52:48 | 2011-02-03 06:55:57 | NULL | 0 | 2 | osdemo02 | nova-compute | compute | 46056 | 0 | nova | + | 2011-01-28 22:52:52 | 2011-02-03 06:55:50 | NULL | 0 | 3 | osdemo02 | nova-scheduler | scheduler | 46065 | 0 | nova | + | 2011-01-29 23:49:29 | 2011-02-03 06:54:26 | NULL | 0 | 4 | osdemo01 | nova-compute | compute | 37050 | 0 | nova | + | 2011-01-30 23:42:24 | 2011-02-03 06:55:44 | NULL | 0 | 9 | osdemo04 | nova-compute | compute | 28484 | 0 | nova | + | 2011-01-30 21:27:28 | 2011-02-03 06:54:23 | NULL | 0 | 8 | osdemo05 | nova-compute | compute | 29284 | 0 | nova | + +---------------------+---------------------+------------+---------+----+----------+----------------+-----------+--------------+----------+-------------------+ +You can see that 'osdemo0{1,2,4,5} are all running 'nova-compute.' When you start spinning up instances, they will allocate on any node that is running nova-compute from this list. + You can then use `euca2ools` to test some items:: euca-describe-images diff --git a/nova/api/ec2/cloud.py b/nova/api/ec2/cloud.py index f1d3ecdd3..b16503589 100644 --- a/nova/api/ec2/cloud.py +++ b/nova/api/ec2/cloud.py @@ -512,8 +512,11 @@ class CloudController(object): def get_console_output(self, context, instance_id, **kwargs): LOG.audit(_("Get console output for instance %s"), instance_id, context=context) - # instance_id is passed in as a list of instances - ec2_id = instance_id[0] + # instance_id may be passed in as a list of instances + if type(instance_id) == list: + ec2_id = instance_id[0] + else: + ec2_id = instance_id instance_id = ec2_id_to_id(ec2_id) output = self.compute_api.get_console_output( context, instance_id=instance_id) diff --git a/nova/compute/instance_types.py b/nova/compute/instance_types.py index 196d6a8df..309313fd0 100644 --- a/nova/compute/instance_types.py +++ b/nova/compute/instance_types.py @@ -38,8 +38,8 @@ def get_by_type(instance_type): if instance_type is None: return FLAGS.default_instance_type if instance_type not in INSTANCE_TYPES: - raise exception.ApiError(_("Unknown instance type: %s"), - instance_type) + raise exception.ApiError(_("Unknown instance type: %s") % \ + instance_type, "Invalid") return instance_type diff --git a/nova/db/sqlalchemy/api.py b/nova/db/sqlalchemy/api.py index 85250d56e..02855e7a9 100644 --- a/nova/db/sqlalchemy/api.py +++ b/nova/db/sqlalchemy/api.py @@ -579,7 +579,7 @@ def fixed_ip_disassociate_all_by_timeout(_context, host, time): 'AND instance_id IS NOT NULL ' 'AND allocated = 0', {'host': host, - 'time': time.isoformat()}) + 'time': time}) return result.rowcount @@ -719,6 +719,7 @@ def instance_get_all(context): return session.query(models.Instance).\ options(joinedload_all('fixed_ip.floating_ips')).\ options(joinedload('security_groups')).\ + options(joinedload_all('fixed_ip.network')).\ filter_by(deleted=can_read_deleted(context)).\ all() @@ -729,6 +730,7 @@ def instance_get_all_by_user(context, user_id): return session.query(models.Instance).\ options(joinedload_all('fixed_ip.floating_ips')).\ options(joinedload('security_groups')).\ + options(joinedload_all('fixed_ip.network')).\ filter_by(deleted=can_read_deleted(context)).\ filter_by(user_id=user_id).\ all() @@ -740,6 +742,7 @@ def instance_get_all_by_host(context, host): return session.query(models.Instance).\ options(joinedload_all('fixed_ip.floating_ips')).\ options(joinedload('security_groups')).\ + options(joinedload_all('fixed_ip.network')).\ filter_by(host=host).\ filter_by(deleted=can_read_deleted(context)).\ all() @@ -753,6 +756,7 @@ def instance_get_all_by_project(context, project_id): return session.query(models.Instance).\ options(joinedload_all('fixed_ip.floating_ips')).\ options(joinedload('security_groups')).\ + options(joinedload_all('fixed_ip.network')).\ filter_by(project_id=project_id).\ filter_by(deleted=can_read_deleted(context)).\ all() @@ -766,6 +770,7 @@ def instance_get_all_by_reservation(context, reservation_id): return session.query(models.Instance).\ options(joinedload_all('fixed_ip.floating_ips')).\ options(joinedload('security_groups')).\ + options(joinedload_all('fixed_ip.network')).\ filter_by(reservation_id=reservation_id).\ filter_by(deleted=can_read_deleted(context)).\ all() @@ -773,6 +778,7 @@ def instance_get_all_by_reservation(context, reservation_id): return session.query(models.Instance).\ options(joinedload_all('fixed_ip.floating_ips')).\ options(joinedload('security_groups')).\ + options(joinedload_all('fixed_ip.network')).\ filter_by(project_id=context.project_id).\ filter_by(reservation_id=reservation_id).\ filter_by(deleted=False).\ diff --git a/nova/db/sqlalchemy/session.py b/nova/db/sqlalchemy/session.py index dc885f138..4a9a28f43 100644 --- a/nova/db/sqlalchemy/session.py +++ b/nova/db/sqlalchemy/session.py @@ -20,6 +20,7 @@ Session Handling for SQLAlchemy backend """ from sqlalchemy import create_engine +from sqlalchemy import pool from sqlalchemy.orm import sessionmaker from nova import exception @@ -37,9 +38,14 @@ def get_session(autocommit=True, expire_on_commit=False): global _MAKER if not _MAKER: if not _ENGINE: + kwargs = {'pool_recycle': FLAGS.sql_idle_timeout, + 'echo': False} + + if FLAGS.sql_connection.startswith('sqlite'): + kwargs['poolclass'] = pool.NullPool + _ENGINE = create_engine(FLAGS.sql_connection, - pool_recycle=FLAGS.sql_idle_timeout, - echo=False) + **kwargs) _MAKER = (sessionmaker(bind=_ENGINE, autocommit=autocommit, expire_on_commit=expire_on_commit)) diff --git a/nova/flags.py b/nova/flags.py index 43bc174d2..3ba3fe6fa 100644 --- a/nova/flags.py +++ b/nova/flags.py @@ -208,7 +208,7 @@ def _get_my_ip(): (addr, port) = csock.getsockname() csock.close() return addr - except socket.gaierror as ex: + except socket.error as ex: return "127.0.0.1" @@ -286,8 +286,8 @@ DEFINE_string('state_path', os.path.join(os.path.dirname(__file__), '../'), DEFINE_string('sql_connection', 'sqlite:///$state_path/nova.sqlite', 'connection string for sql database') -DEFINE_string('sql_idle_timeout', - '3600', +DEFINE_integer('sql_idle_timeout', + 3600, 'timeout for idle sql database connections') DEFINE_integer('sql_max_retries', 12, 'sql connection attempts') DEFINE_integer('sql_retry_interval', 10, 'sql connection retry interval') diff --git a/nova/network/linux_net.py b/nova/network/linux_net.py index cdd1f666a..ed37e8ba7 100644 --- a/nova/network/linux_net.py +++ b/nova/network/linux_net.py @@ -20,6 +20,7 @@ Implements vlans, bridges, and iptables rules using linux utilities. import os from nova import db +from nova import exception from nova import flags from nova import log as logging from nova import utils @@ -37,6 +38,9 @@ FLAGS = flags.FLAGS flags.DEFINE_string('dhcpbridge_flagfile', '/etc/nova/nova-dhcpbridge.conf', 'location of flagfile for dhcpbridge') +flags.DEFINE_string('dhcp_domain', + 'novalocal', + 'domain to use for building the hostnames') flags.DEFINE_string('networks_path', '$state_path/networks', 'Location to keep network config files') @@ -152,6 +156,8 @@ def ensure_floating_forward(floating_ip, fixed_ip): """Ensure floating ip forwarding rule""" _confirm_rule("PREROUTING", "-t nat -d %s -j DNAT --to %s" % (floating_ip, fixed_ip)) + _confirm_rule("OUTPUT", "-t nat -d %s -j DNAT --to %s" + % (floating_ip, fixed_ip)) _confirm_rule("SNATTING", "-t nat -s %s -j SNAT --to %s" % (fixed_ip, floating_ip)) @@ -160,6 +166,8 @@ def remove_floating_forward(floating_ip, fixed_ip): """Remove forwarding for floating ip""" _remove_rule("PREROUTING", "-t nat -d %s -j DNAT --to %s" % (floating_ip, fixed_ip)) + _remove_rule("OUTPUT", "-t nat -d %s -j DNAT --to %s" + % (floating_ip, fixed_ip)) _remove_rule("SNATTING", "-t nat -s %s -j SNAT --to %s" % (fixed_ip, floating_ip)) @@ -177,32 +185,77 @@ def ensure_vlan(vlan_num): LOG.debug(_("Starting VLAN inteface %s"), interface) _execute("sudo vconfig set_name_type VLAN_PLUS_VID_NO_PAD") _execute("sudo vconfig add %s %s" % (FLAGS.vlan_interface, vlan_num)) - _execute("sudo ifconfig %s up" % interface) + _execute("sudo ip link set %s up" % interface) return interface def ensure_bridge(bridge, interface, net_attrs=None): - """Create a bridge unless it already exists""" + """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 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 interface for %s"), interface) _execute("sudo brctl addbr %s" % bridge) _execute("sudo brctl setfd %s 0" % bridge) # _execute("sudo brctl setageing %s 10" % bridge) _execute("sudo brctl stp %s off" % bridge) - if interface: - _execute("sudo brctl addif %s %s" % (bridge, interface)) + _execute("sudo ip link set %s up" % bridge) if net_attrs: - _execute("sudo ifconfig %s %s broadcast %s netmask %s up" % \ - (bridge, - net_attrs['gateway'], - net_attrs['broadcast'], - net_attrs['netmask'])) + # NOTE(vish): The ip for dnsmasq has to be the first address on the + # bridge for it to respond to reqests properly + suffix = net_attrs['cidr'].rpartition('/')[2] + out, err = _execute("sudo ip addr add %s/%s brd %s dev %s" % + (net_attrs['gateway'], + suffix, + net_attrs['broadcast'], + bridge), + check_exit_code=False) + if err and err != "RTNETLINK answers: File exists\n": + raise exception.Error("Failed to add ip: %s" % err) if(FLAGS.use_ipv6): _execute("sudo ip -f inet6 addr change %s dev %s" % (net_attrs['cidr_v6'], bridge)) - _execute("sudo ifconfig %s up" % bridge) - else: - _execute("sudo ifconfig %s up" % bridge) + # NOTE(vish): If the public interface is the same as the + # bridge, then the bridge has to be in promiscuous + # to forward packets properly. + if(FLAGS.public_interface == bridge): + _execute("sudo ip link set dev %s promisc on" % bridge) + if interface: + # NOTE(vish): This will break if there is already an ip on the + # interface, so we move any ips to the bridge + gateway = None + out, err = _execute("sudo route -n") + for line in out.split("\n"): + fields = line.split() + if fields and fields[0] == "0.0.0.0" and fields[-1] == interface: + gateway = fields[1] + out, err = _execute("sudo ip addr show dev %s scope global" % + interface) + for line in out.split("\n"): + fields = line.split() + if fields and fields[0] == "inet": + params = ' '.join(fields[1:-1]) + _execute("sudo ip addr del %s dev %s" % (params, fields[-1])) + _execute("sudo ip addr add %s dev %s" % (params, bridge)) + if gateway: + _execute("sudo route add 0.0.0.0 gw %s" % gateway) + out, err = _execute("sudo brctl addif %s %s" % + (bridge, interface), + check_exit_code=False) + + if (err and err != "device %s is already a member of a bridge; can't " + "enslave it to bridge %s.\n" % (interface, bridge)): + raise exception.Error("Failed to add interface: %s" % err) + if FLAGS.use_nova_chains: (out, err) = _execute("sudo iptables -N nova_forward", check_exit_code=False) @@ -313,8 +366,9 @@ interface %s def _host_dhcp(fixed_ip_ref): """Return a host string for an address""" instance_ref = fixed_ip_ref['instance'] - return "%s,%s.novalocal,%s" % (instance_ref['mac_address'], + return "%s,%s.%s,%s" % (instance_ref['mac_address'], instance_ref['hostname'], + FLAGS.dhcp_domain, fixed_ip_ref['address']) @@ -329,7 +383,8 @@ def _execute(cmd, *args, **kwargs): def _device_exists(device): """Check if ethernet device exists""" - (_out, err) = _execute("ifconfig %s" % device, check_exit_code=False) + (_out, err) = _execute("ip link show dev %s" % device, + check_exit_code=False) return not err @@ -359,6 +414,7 @@ def _dnsmasq_cmd(net): ' --strict-order', ' --bind-interfaces', ' --conf-file=', + ' --domain=%s' % FLAGS.dhcp_domain, ' --pid-file=%s' % _dhcp_file(net['bridge'], 'pid'), ' --listen-address=%s' % net['gateway'], ' --except-interface=lo', diff --git a/nova/network/manager.py b/nova/network/manager.py index fbcbea131..8eb9f041b 100644 --- a/nova/network/manager.py +++ b/nova/network/manager.py @@ -118,6 +118,10 @@ class NetworkManager(manager.Manager): super(NetworkManager, self).__init__(*args, **kwargs) def init_host(self): + """Do any initialization that needs to be run if this is a + standalone service. + """ + self.driver.init_host() # Set up networking for the projects for which we're already # the designated network host. ctxt = context.get_admin_context() @@ -395,7 +399,6 @@ class FlatDHCPManager(FlatManager): standalone service. """ super(FlatDHCPManager, self).init_host() - self.driver.init_host() self.driver.metadata_forward() def setup_compute_network(self, context, instance_id): @@ -465,7 +468,6 @@ class VlanManager(NetworkManager): standalone service. """ super(VlanManager, self).init_host() - self.driver.init_host() self.driver.metadata_forward() def allocate_fixed_ip(self, context, instance_id, *args, **kwargs): diff --git a/nova/rpc.py b/nova/rpc.py index 01fc6d44b..2b1f7298b 100644 --- a/nova/rpc.py +++ b/nova/rpc.py @@ -46,7 +46,7 @@ LOG = logging.getLogger('nova.rpc') class Connection(carrot_connection.BrokerConnection): """Connection instance object""" @classmethod - def instance(cls, new=False): + def instance(cls, new=True): """Returns the instance""" if new or not hasattr(cls, '_instance'): params = dict(hostname=FLAGS.rabbit_host, @@ -246,7 +246,7 @@ def msg_reply(msg_id, reply=None, failure=None): LOG.error(_("Returning exception %s to caller"), message) LOG.error(tb) failure = (failure[0].__name__, str(failure[1]), tb) - conn = Connection.instance(True) + conn = Connection.instance() publisher = DirectPublisher(connection=conn, msg_id=msg_id) try: publisher.send({'result': reply, 'failure': failure}) @@ -319,7 +319,7 @@ def call(context, topic, msg): self.result = data['result'] wait_msg = WaitMessage() - conn = Connection.instance(True) + conn = Connection.instance() consumer = DirectConsumer(connection=conn, msg_id=msg_id) consumer.register_callback(wait_msg) diff --git a/nova/utils.py b/nova/utils.py index 5f5225289..8d7ff1f64 100644 --- a/nova/utils.py +++ b/nova/utils.py @@ -152,6 +152,42 @@ def execute(cmd, process_input=None, addl_env=None, check_exit_code=True): return result +def ssh_execute(ssh, cmd, process_input=None, + addl_env=None, check_exit_code=True): + LOG.debug(_("Running cmd (SSH): %s"), cmd) + if addl_env: + raise exception.Error("Environment not supported over SSH") + + if process_input: + # This is (probably) fixable if we need it... + raise exception.Error("process_input not supported over SSH") + + stdin_stream, stdout_stream, stderr_stream = ssh.exec_command(cmd) + channel = stdout_stream.channel + + #stdin.write('process_input would go here') + #stdin.flush() + + # NOTE(justinsb): This seems suspicious... + # ...other SSH clients have buffering issues with this approach + stdout = stdout_stream.read() + stderr = stderr_stream.read() + stdin_stream.close() + + exit_status = channel.recv_exit_status() + + # exit_status == -1 if no exit code was returned + if exit_status != -1: + LOG.debug(_("Result was %s") % exit_status) + if check_exit_code and exit_status != 0: + raise exception.ProcessExecutionError(exit_code=exit_status, + stdout=stdout, + stderr=stderr, + cmd=cmd) + + return (stdout, stderr) + + def abspath(s): return os.path.join(os.path.dirname(__file__), s) diff --git a/nova/volume/driver.py b/nova/volume/driver.py index da7307733..82f4c2f54 100644 --- a/nova/volume/driver.py +++ b/nova/volume/driver.py @@ -294,8 +294,10 @@ class ISCSIDriver(VolumeDriver): self._execute("sudo ietadm --op delete --tid=%s" % iscsi_target) - def _get_name_and_portal(self, volume_name, host): + def _get_name_and_portal(self, volume): """Gets iscsi name and portal from volume name and host.""" + volume_name = volume['name'] + host = volume['host'] (out, _err) = self._execute("sudo iscsiadm -m discovery -t " "sendtargets -p %s" % host) for target in out.splitlines(): @@ -307,8 +309,7 @@ class ISCSIDriver(VolumeDriver): def discover_volume(self, volume): """Discover volume on a remote host.""" - iscsi_name, iscsi_portal = self._get_name_and_portal(volume['name'], - volume['host']) + iscsi_name, iscsi_portal = self._get_name_and_portal(volume) self._execute("sudo iscsiadm -m node -T %s -p %s --login" % (iscsi_name, iscsi_portal)) self._execute("sudo iscsiadm -m node -T %s -p %s --op update " @@ -319,8 +320,7 @@ class ISCSIDriver(VolumeDriver): def undiscover_volume(self, volume): """Undiscover volume on a remote host.""" - iscsi_name, iscsi_portal = self._get_name_and_portal(volume['name'], - volume['host']) + iscsi_name, iscsi_portal = self._get_name_and_portal(volume) self._execute("sudo iscsiadm -m node -T %s -p %s --op update " "-n node.startup -v manual" % (iscsi_name, iscsi_portal)) diff --git a/nova/volume/manager.py b/nova/volume/manager.py index 6f8e25e19..6e70ec881 100644 --- a/nova/volume/manager.py +++ b/nova/volume/manager.py @@ -87,7 +87,7 @@ class VolumeManager(manager.Manager): if volume['status'] in ['available', 'in-use']: self.driver.ensure_export(ctxt, volume) else: - LOG.info(_("volume %s: skipping export"), volume_ref['name']) + LOG.info(_("volume %s: skipping export"), volume['name']) def create_volume(self, context, volume_id): """Creates and exports the volume.""" diff --git a/nova/volume/san.py b/nova/volume/san.py new file mode 100644 index 000000000..26d6125e7 --- /dev/null +++ b/nova/volume/san.py @@ -0,0 +1,335 @@ +# vim: tabstop=4 shiftwidth=4 softtabstop=4 + +# Copyright 2011 Justin Santa Barbara +# 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. +""" +Drivers for san-stored volumes. +The unique thing about a SAN is that we don't expect that we can run the volume + controller on the SAN hardware. We expect to access it over SSH or some API. +""" + +import os +import paramiko + +from nova import exception +from nova import flags +from nova import log as logging +from nova.utils import ssh_execute +from nova.volume.driver import ISCSIDriver + +LOG = logging.getLogger("nova.volume.driver") +FLAGS = flags.FLAGS +flags.DEFINE_boolean('san_thin_provision', 'true', + 'Use thin provisioning for SAN volumes?') +flags.DEFINE_string('san_ip', '', + 'IP address of SAN controller') +flags.DEFINE_string('san_login', 'admin', + 'Username for SAN controller') +flags.DEFINE_string('san_password', '', + 'Password for SAN controller') +flags.DEFINE_string('san_privatekey', '', + 'Filename of private key to use for SSH authentication') + + +class SanISCSIDriver(ISCSIDriver): + """ Base class for SAN-style storage volumes + (storage providers we access over SSH)""" + #Override because SAN ip != host ip + def _get_name_and_portal(self, volume): + """Gets iscsi name and portal from volume name and host.""" + volume_name = volume['name'] + + # TODO(justinsb): store in volume, remerge with generic iSCSI code + host = FLAGS.san_ip + + (out, _err) = self._execute("sudo iscsiadm -m discovery -t " + "sendtargets -p %s" % host) + + location = None + find_iscsi_name = self._build_iscsi_target_name(volume) + for target in out.splitlines(): + if find_iscsi_name in target: + (location, _sep, iscsi_name) = target.partition(" ") + break + if not location: + raise exception.Error(_("Could not find iSCSI export " + " for volume %s") % + volume_name) + + iscsi_portal = location.split(",")[0] + LOG.debug("iscsi_name=%s, iscsi_portal=%s" % + (iscsi_name, iscsi_portal)) + return (iscsi_name, iscsi_portal) + + def _build_iscsi_target_name(self, volume): + return "%s%s" % (FLAGS.iscsi_target_prefix, volume['name']) + + # discover_volume is still OK + # undiscover_volume is still OK + + def _connect_to_ssh(self): + ssh = paramiko.SSHClient() + #TODO(justinsb): We need a better SSH key policy + ssh.set_missing_host_key_policy(paramiko.AutoAddPolicy()) + if FLAGS.san_password: + ssh.connect(FLAGS.san_ip, + username=FLAGS.san_login, + password=FLAGS.san_password) + elif FLAGS.san_privatekey: + privatekeyfile = os.path.expanduser(FLAGS.san_privatekey) + # It sucks that paramiko doesn't support DSA keys + privatekey = paramiko.RSAKey.from_private_key_file(privatekeyfile) + ssh.connect(FLAGS.san_ip, + username=FLAGS.san_login, + pkey=privatekey) + else: + raise exception.Error("Specify san_password or san_privatekey") + return ssh + + def _run_ssh(self, command, check_exit_code=True): + #TODO(justinsb): SSH connection caching (?) + ssh = self._connect_to_ssh() + + #TODO(justinsb): Reintroduce the retry hack + ret = ssh_execute(ssh, command, check_exit_code=check_exit_code) + + ssh.close() + + return ret + + def ensure_export(self, context, volume): + """Synchronously recreates an export for a logical volume.""" + pass + + def create_export(self, context, volume): + """Exports the volume.""" + pass + + def remove_export(self, context, volume): + """Removes an export for a logical volume.""" + pass + + def check_for_setup_error(self): + """Returns an error if prerequisites aren't met""" + if not (FLAGS.san_password or FLAGS.san_privatekey): + raise exception.Error("Specify san_password or san_privatekey") + + if not (FLAGS.san_ip): + raise exception.Error("san_ip must be set") + + +def _collect_lines(data): + """ Split lines from data into an array, trimming them """ + matches = [] + for line in data.splitlines(): + match = line.strip() + matches.append(match) + + return matches + + +def _get_prefixed_values(data, prefix): + """Collect lines which start with prefix; with trimming""" + matches = [] + for line in data.splitlines(): + line = line.strip() + if line.startswith(prefix): + match = line[len(prefix):] + match = match.strip() + matches.append(match) + + return matches + + +class SolarisISCSIDriver(SanISCSIDriver): + """Executes commands relating to Solaris-hosted ISCSI volumes. + Basic setup for a Solaris iSCSI server: + pkg install storage-server SUNWiscsit + svcadm enable stmf + svcadm enable -r svc:/network/iscsi/target:default + pfexec itadm create-tpg e1000g0 ${MYIP} + pfexec itadm create-target -t e1000g0 + + Then grant the user that will be logging on lots of permissions. + I'm not sure exactly which though: + zfs allow justinsb create,mount,destroy rpool + usermod -P'File System Management' justinsb + usermod -P'Primary Administrator' justinsb + + Also make sure you can login using san_login & san_password/san_privatekey + """ + + def _view_exists(self, luid): + cmd = "pfexec /usr/sbin/stmfadm list-view -l %s" % (luid) + (out, _err) = self._run_ssh(cmd, + check_exit_code=False) + if "no views found" in out: + return False + + if "View Entry:" in out: + return True + + raise exception.Error("Cannot parse list-view output: %s" % (out)) + + def _get_target_groups(self): + """Gets list of target groups from host.""" + (out, _err) = self._run_ssh("pfexec /usr/sbin/stmfadm list-tg") + matches = _get_prefixed_values(out, 'Target group: ') + LOG.debug("target_groups=%s" % matches) + return matches + + def _target_group_exists(self, target_group_name): + return target_group_name not in self._get_target_groups() + + def _get_target_group_members(self, target_group_name): + (out, _err) = self._run_ssh("pfexec /usr/sbin/stmfadm list-tg -v %s" % + (target_group_name)) + matches = _get_prefixed_values(out, 'Member: ') + LOG.debug("members of %s=%s" % (target_group_name, matches)) + return matches + + def _is_target_group_member(self, target_group_name, iscsi_target_name): + return iscsi_target_name in ( + self._get_target_group_members(target_group_name)) + + def _get_iscsi_targets(self): + cmd = ("pfexec /usr/sbin/itadm list-target | " + "awk '{print $1}' | grep -v ^TARGET") + (out, _err) = self._run_ssh(cmd) + matches = _collect_lines(out) + LOG.debug("_get_iscsi_targets=%s" % (matches)) + return matches + + def _iscsi_target_exists(self, iscsi_target_name): + return iscsi_target_name in self._get_iscsi_targets() + + def _build_zfs_poolname(self, volume): + #TODO(justinsb): rpool should be configurable + zfs_poolname = 'rpool/%s' % (volume['name']) + return zfs_poolname + + def create_volume(self, volume): + """Creates a volume.""" + if int(volume['size']) == 0: + sizestr = '100M' + else: + sizestr = '%sG' % volume['size'] + + zfs_poolname = self._build_zfs_poolname(volume) + + thin_provision_arg = '-s' if FLAGS.san_thin_provision else '' + # Create a zfs volume + self._run_ssh("pfexec /usr/sbin/zfs create %s -V %s %s" % + (thin_provision_arg, + sizestr, + zfs_poolname)) + + def _get_luid(self, volume): + zfs_poolname = self._build_zfs_poolname(volume) + + cmd = ("pfexec /usr/sbin/sbdadm list-lu | " + "grep -w %s | awk '{print $1}'" % + (zfs_poolname)) + + (stdout, _stderr) = self._run_ssh(cmd) + + luid = stdout.strip() + return luid + + def _is_lu_created(self, volume): + luid = self._get_luid(volume) + return luid + + def delete_volume(self, volume): + """Deletes a volume.""" + zfs_poolname = self._build_zfs_poolname(volume) + self._run_ssh("pfexec /usr/sbin/zfs destroy %s" % + (zfs_poolname)) + + def local_path(self, volume): + # TODO(justinsb): Is this needed here? + escaped_group = FLAGS.volume_group.replace('-', '--') + escaped_name = volume['name'].replace('-', '--') + return "/dev/mapper/%s-%s" % (escaped_group, escaped_name) + + def ensure_export(self, context, volume): + """Synchronously recreates an export for a logical volume.""" + #TODO(justinsb): On bootup, this is called for every volume. + # It then runs ~5 SSH commands for each volume, + # most of which fetch the same info each time + # This makes initial start stupid-slow + self._do_export(volume, force_create=False) + + def create_export(self, context, volume): + self._do_export(volume, force_create=True) + + def _do_export(self, volume, force_create): + # Create a Logical Unit (LU) backed by the zfs volume + zfs_poolname = self._build_zfs_poolname(volume) + + if force_create or not self._is_lu_created(volume): + cmd = ("pfexec /usr/sbin/sbdadm create-lu /dev/zvol/rdsk/%s" % + (zfs_poolname)) + self._run_ssh(cmd) + + luid = self._get_luid(volume) + iscsi_name = self._build_iscsi_target_name(volume) + target_group_name = 'tg-%s' % volume['name'] + + # Create a iSCSI target, mapped to just this volume + if force_create or not self._target_group_exists(target_group_name): + self._run_ssh("pfexec /usr/sbin/stmfadm create-tg %s" % + (target_group_name)) + + # Yes, we add the initiatior before we create it! + # Otherwise, it complains that the target is already active + if force_create or not self._is_target_group_member(target_group_name, + iscsi_name): + self._run_ssh("pfexec /usr/sbin/stmfadm add-tg-member -g %s %s" % + (target_group_name, iscsi_name)) + if force_create or not self._iscsi_target_exists(iscsi_name): + self._run_ssh("pfexec /usr/sbin/itadm create-target -n %s" % + (iscsi_name)) + if force_create or not self._view_exists(luid): + self._run_ssh("pfexec /usr/sbin/stmfadm add-view -t %s %s" % + (target_group_name, luid)) + + def remove_export(self, context, volume): + """Removes an export for a logical volume.""" + + # This is the reverse of _do_export + luid = self._get_luid(volume) + iscsi_name = self._build_iscsi_target_name(volume) + target_group_name = 'tg-%s' % volume['name'] + + if self._view_exists(luid): + self._run_ssh("pfexec /usr/sbin/stmfadm remove-view -l %s -a" % + (luid)) + + if self._iscsi_target_exists(iscsi_name): + self._run_ssh("pfexec /usr/sbin/stmfadm offline-target %s" % + (iscsi_name)) + self._run_ssh("pfexec /usr/sbin/itadm delete-target %s" % + (iscsi_name)) + + # We don't delete the tg-member; we delete the whole tg! + + if self._target_group_exists(target_group_name): + self._run_ssh("pfexec /usr/sbin/stmfadm delete-tg %s" % + (target_group_name)) + + if self._is_lu_created(volume): + self._run_ssh("pfexec /usr/sbin/sbdadm delete-lu %s" % + (luid)) |
