ceph-deploy中new模块有什么用

这篇文章主要为大家展示了"ceph-deploy中new模块有什么用"，内容简而易懂，条理清晰，希望能够帮助大家解决疑惑，下面让小编带领大家一起研究并学习一下"ceph-deploy中new模块有什么用"这篇文章吧。

ceph-deploy源码分析--new模块

ceph-deploy的new.py模块是用来开始部署新集群，创建ceph.conf、ceph.mon.keyring文件。

new 子命令格式如下

ceph-deploy new [-h] [--no-ssh-copykey] [--fsid FSID]                       [--cluster-network CLUSTER_NETWORK]                       [--public-network PUBLIC_NETWORK]                       MON [MON ...]

部署集群

make函数priority为10，子命令设置的默认函数为new函数。

@priority(10)def make(parser):    """    Start deploying a new cluster, and write a CLUSTER.conf and keyring for it.    """    parser.add_argument(        'mon',        metavar='MON',        nargs='+',        help='initial monitor hostname, fqdn, or hostname:fqdn pair',        type=arg_validators.Hostname(),        )    parser.add_argument(        '--no-ssh-copykey',        dest='ssh_copykey',        action='store_false',        default=True,        help='do not attempt to copy SSH keys',    )    parser.add_argument(        '--fsid',        dest='fsid',        help='provide an alternate FSID for ceph.conf generation',    )    parser.add_argument(        '--cluster-network',        help='specify the (internal) cluster network',        type=arg_validators.Subnet(),    )    parser.add_argument(        '--public-network',        help='specify the public network for a cluster',        type=arg_validators.Subnet(),    )    parser.set_defaults(        func=new,        )

部署新集群

new 函数开始部署新集群

• 创建ceph.conf文件,写入[global]fsid、mon_initial_members、mon_host、auth_cluster_required、auth_service_required、auth_client_required，如果参数中有public_network、cluster_network写入配置文件

• 调用new_mon_keyring函数创建ceph.mon.keyring文件

• def new(args):    if args.ceph_conf:        raise RuntimeError('will not create a Ceph conf file if attemtping to re-use with `--ceph-conf` flag')    LOG.debug('Creating new cluster named %s', args.cluster)    # 生成配置    cfg = conf.ceph.CephConf()    cfg.add_section('global')    # 获取参数中的额fsid或者自动生成    fsid = args.fsid or uuid.uuid4()    cfg.set('global', 'fsid', str(fsid))    # if networks were passed in, lets set them in the    # global section    if args.public_network:        cfg.set('global', 'public network', str(args.public_network))    if args.cluster_network:        cfg.set('global', 'cluster network', str(args.cluster_network))    # mon节点    mon_initial_members = []    # mon主机    mon_host = []    # 循环host    for (name, host) in mon_hosts(args.mon):        # Try to ensure we can ssh in properly before anything else        # ssh key copy        if args.ssh_copykey:            ssh_copy_keys(host, args.username)        # Now get the non-local IPs from the remote node        # 连接远程主机        distro = hosts.get(host, username=args.username)        # 获取主机的IP地址        remote_ips = net.ip_addresses(distro.conn)        # custom cluster names on sysvinit hosts won't work        if distro.init == 'sysvinit' and args.cluster != 'ceph':            LOG.error('custom cluster names are not supported on sysvinit hosts')            raise exc.ClusterNameError(                'host %s does not support custom cluster names' % host            )        distro.conn.exit()        # Validate subnets if we received any        if args.public_network or args.cluster_network:            # 校验IP地址            validate_host_ip(remote_ips, [args.public_network, args.cluster_network])        # Pick the IP that matches the public cluster (if we were told to do        # so) otherwise pick the first, non-local IP        LOG.debug('Resolving host %s', host)        if args.public_network:            ip = get_public_network_ip(remote_ips, args.public_network)        else:            ip = net.get_nonlocal_ip(host)        LOG.debug('Monitor %s at %s', name, ip)        mon_initial_members.append(name)        try:            socket.inet_pton(socket.AF_INET6, ip)            mon_host.append("[" + ip + "]")            LOG.info('Monitors are IPv6, binding Messenger traffic on IPv6')            cfg.set('global', 'ms bind ipv6', 'true')        except socket.error:            mon_host.append(ip)    LOG.debug('Monitor initial members are %s', mon_initial_members)    LOG.debug('Monitor addrs are %s', mon_host)    # mon_initial_members 有多个的话，中间用空格隔开    cfg.set('global', 'mon initial members', ', '.join(mon_initial_members))    # no spaces here, see http://tracker.newdream.net/issues/3145    # mon_host 有多个的话，中间没有空格    cfg.set('global', 'mon host', ','.join(mon_host))    # override undesirable defaults, needed until bobtail    # http://tracker.ceph.com/issues/6788    cfg.set('global', 'auth cluster required', 'cephx')    cfg.set('global', 'auth service required', 'cephx')    cfg.set('global', 'auth client required', 'cephx')    path = '{name}.conf'.format(        name=args.cluster,        )    # 创建mon keyring    new_mon_keyring(args)    LOG.debug('Writing initial config to %s...', path)    tmp = '%s.tmp' % path    with open(tmp, 'w') as f:        # 保存ceph配置文件        cfg.write(f)    try:        os.rename(tmp, path)    except OSError as e:        if e.errno == errno.EEXIST:            raise exc.ClusterExistsError(path)        else:            raise

  
  

注意：
mon_initial_members 有多个的话，中间用空格隔开
mon_host 有多个的话，中间没有空格

创建ceph.mon.keyring文件

new_mon_keyring函数创建ceph.mon.keyring文件

def new_mon_keyring(args):    LOG.debug('Creating a random mon key...')    mon_keyring = '[mon.]\nkey = %s\ncaps mon = allow *\n' % generate_auth_key()    keypath = '{name}.mon.keyring'.format(        name=args.cluster,        )    oldmask = os.umask(0o77)    LOG.debug('Writing monitor keyring to %s...', keypath)    try:        tmp = '%s.tmp' % keypath        with open(tmp, 'w', 0o600) as f:            f.write(mon_keyring)        try:            os.rename(tmp, keypath)        except OSError as e:            if e.errno == errno.EEXIST:                raise exc.ClusterExistsError(keypath)            else:                raise    finally:        os.umask(oldmask)

手工部署集群

以ceph-deploy部署集群：ceph-deploy new ceph-231为例，对应的手工操作。

获取ip地址

执行以下命令，通过正则表达式获取IP地址192.168.217.231

[root@ceph-231 ceph-cluster]# /usr/sbin/ip link show1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:002: eth0:  mtu 1500 qdisc pfifo_fast master ovs-system state UP mode DEFAULT qlen 1000    link/ether 02:03:e7:fc:dc:36 brd ff:ff:ff:ff:ff:ff3: ovs-system:  mtu 1500 qdisc noop state DOWN mode DEFAULT    link/ether 86:f4:14:e3:1b:b2 brd ff:ff:ff:ff:ff:ff4: xenbr0:  mtu 1500 qdisc noqueue state UNKNOWN mode DEFAULT    link/ether 02:03:e7:fc:dc:36 brd ff:ff:ff:ff:ff:ff

[root@ceph-231 ceph-cluster]# /usr/sbin/ip addr show1: lo:  mtu 65536 qdisc noqueue state UNKNOWN    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00    inet 127.0.0.1/8 scope host lo       valid_lft forever preferred_lft forever2: eth0:  mtu 1500 qdisc pfifo_fast master ovs-system state UP qlen 1000    link/ether 02:03:e7:fc:dc:36 brd ff:ff:ff:ff:ff:ff3: ovs-system:  mtu 1500 qdisc noop state DOWN    link/ether 86:f4:14:e3:1b:b2 brd ff:ff:ff:ff:ff:ff4: xenbr0:  mtu 1500 qdisc noqueue state UNKNOWN    link/ether 02:03:e7:fc:dc:36 brd ff:ff:ff:ff:ff:ff    inet 192.168.217.231/24 brd 192.168.217.255 scope global xenbr0       valid_lft forever preferred_lft forever

创建ceph.conf

[root@ceph-231 ceph-cluster]# vi ceph.conf[global]fsid = a3b9b0aa-01ab-4e1b-bba3-6f5317b0795bmon_initial_members = ceph-231mon_host = 192.168.217.231auth_cluster_required = cephxauth_service_required = cephxauth_client_required = cephxpublic_network = 192.168.217.231

创建ceph.mon.keyring

可以通过ceph-authtool命令生成

[root@ceph-231 ceph-cluster]# ceph-authtool --create-keyring /tmp/ceph.mon.keyring --gen-key -n mon. --cap mon 'allow *'creating /tmp/ceph.mon.keyring[root@ceph-231 ~]# cat /tmp/ceph.mon.keyring[mon.]        key = AQCzxEhZC7tICxAAuHK5GipD96enMuhv82CCLg==        caps mon = "allow *"

将/tmp/ceph.mon.keyring内容复制到ceph.mon.keyring

[root@ceph-231 ceph-cluster]# vi ceph.mon.keyring[mon.]key = AQCzxEhZC7tICxAAuHK5GipD96enMuhv82CCLg==caps mon = allow

以上是"ceph-deploy中new模块有什么用"这篇文章的所有内容，感谢各位的阅读！相信大家都有了一定的了解，希望分享的内容对大家有所帮助，如果还想学习更多知识，欢迎关注行业资讯频道！