Oracle Solaris 10 IPv6 Implementation

This section describes the files, commands, and daemons that enable IPv6 in Oracle Solaris.

IPv6 Configuration Files

This section describes the configuration files that are part of an IPv6 implementation:

• ndpd.conf Configuration File

• IPv6 Interface Configuration File

• /etc/inet/ipaddrsel.conf Configuration File

ndpd.conf Configuration File

The /etc/inet/ndpd.conf file is used to configure options that are used by the in.ndpd Neighbor Discovery daemon. For a router, you primarily use ndpd.conf to configure the site prefix to be advertised to the link. For a host, you use ndpd.conf to turn off address autoconfiguration or to configure temporary addresses.

The next table shows the keywords that are used in the ndpd.conf file.

In the ndpd.conf file, you use the keywords in this table with a set of router configuration variables. These variables are defined in detail in RFC 2461, Neighbor Discovery for IP Version 6 (IPv6).

The next table shows the variables for configuring an interface, along with brief definitions.

The next table shows the variables that are used for configuring IPv6 prefixes.

Example 11–1 /etc/inet/ndpd.conf File

The following example shows how the keywords and configuration variables are used in the ndpd.conf file. Remove the comment (#) to activate the variable.

# ifdefault      [variable-value ]*
# prefixdefault [variable-value ]*
# if ifname   [variable-value ]*
# prefix prefix/length ifname
#
#  Per interface configuration variables
#
#DupAddrDetectTransmits
#AdvSendAdvertisements
#MaxRtrAdvInterval
#MinRtrAdvInterval
#AdvManagedFlag
#AdvOtherConfigFlag
#AdvLinkMTU
#AdvReachableTime
#AdvRetransTimer
#AdvCurHopLimit
#AdvDefaultLifetime
#
# Per Prefix:  AdvPrefixList configuration variables
#
#
#AdvValidLifetime
#AdvOnLinkFlag
#AdvPreferredLifetime
#AdvAutonomousFlag
#AdvValidExpiration
#AdvPreferredExpiration

ifdefault AdvReachableTime 30000 AdvRetransTimer 2000
prefixdefault AdvValidLifetime 240m AdvPreferredLifetime 120m

if qe0 AdvSendAdvertisements 1
prefix 2:0:0:56::/64 qe0
prefix fec0:0:0:56::/64 qe0

if qe1 AdvSendAdvertisements 1
prefix 2:0:0:55::/64 qe1
prefix fec0:0:0:56::/64 qe1

if hme1 AdvSendAdvertisements 1
prefix  2002:8192:56bb:1::/64 qfe0 

if hme1 AdvSendAdvertisements 1
prefix  2002:8192:56bb:2::/64 hme1

IPv6 Interface Configuration File

IPv6 uses the /etc/hostname6.interface file at start up to automatically define IPv6 logical interfaces. When you select the IPv6 Enabled option during Oracle Solaris installation, the installation program creates an /etc/hostname6.interface file for the primary network interface, in addition to the /etc/hostname.interface file.

If more than one physical interface is detected during installation, you are prompted as to whether you want to configure these interfaces. The installation program creates IPv4 physical interface configuration files and IPv6 logical interface configuration files for each additional interface that you indicate.

As with IPv4 interfaces, you can also configure IPv6 interfaces manually, after Oracle Solaris installation. You create/etc/hostname6.interface files for the new interfaces. For instructions for manually configuring interfaces, refer to Administering Interfaces in Solaris 10 3/05 or Chapter 6, Administering Network Interfaces (Tasks).

The network interface configuration file names have the following syntax:

hostname.interface
hostname6.interface

The interface variable has the following syntax:

dev[.module[.module ...]]PPA

dev

Indicates a network interface device. The device can be a physical network interface, such as eri or qfe, or a logical interface, such as a tunnel. See IPv6 Interface Configuration File for more details.

Module

Lists one or more STREAMS modules to be pushed onto the device when the device is plumbed.

PPA

Indicates the physical point of attachment.

The syntax [.[.]] is also accepted.

Example 11–2 IPv6 Interface Configuration Files

The following are examples of valid IPv6 configuration file names:

hostname6.qfe0
hostname.ip.tun0
hostname.ip6.tun0
hostname6.ip6to4tun0
hostname6.ip.tun0
hostname6.ip6.tun0

/etc/inet/ipaddrsel.conf Configuration File

The /etc/inet/ipaddrsel.conf file contains the IPv6 default address selection policy table. When you install Oracle Solaris with IPv6 enabled, this file contains the contents that are shown in Table 11–5.

You can edit the contents of /etc/inet/ipaddrsel.conf. However, in most cases, you should refrain from modifying this file. If modification is necessary, refer to the procedure How to Administer the IPv6 Address Selection Policy Table. For more information on ippaddrsel.conf, refer to Reasons for Modifying the IPv6 Address Selection Policy Table and the ipaddrsel.conf(4) man page.

IPv6-Related Commands

This section describes commands that are added with the Oracle Solaris IPv6 implementation. The text also describes modifications to existing commands to support IPv6.

ipaddrsel Command

The ipaddrsel command enables you to modify the IPv6 default address selection policy table.

The Oracle Solaris kernel uses the IPv6 default address selection policy table to perform destination address ordering and source address selection for an IPv6 packet header. The /etc/inet/ipaddrsel.conf file contains the policy table.

The following table lists the default address formats and their priorities for the policy table. You can find technical details for IPv6 address selection in the inet6(7P) man page.

In this table, IPv6 prefixes (::1/128 and ::/0) take precedence over 6to4 addresses (2002::/16) and IPv4 addresses (::/96 and ::ffff:0:0/96). Therefore, by default, the kernel selects the global IPv6 address of the interface for packets going to another IPv6 destination. The IPv4 address of the interface has a lower priority, particularly for packets going to an IPv6 destination. Given the selected IPv6 source address, the kernel also uses the IPv6 format for the destination address.

Reasons for Modifying the IPv6 Address Selection Policy Table

Under most instances, you do not need to change the IPv6 default address selection policy table. If you do need to administer the policy table, you use the ipaddrsel command.

You might want to modify the policy table under the following circumstances:

• If the system has an interface that is used for a 6to4 tunnel, you can give higher priority to 6to4 addresses.

• If you want a particular source address to be used only in communications with a particular destination address, you can add these addresses to the policy table. Then, you can use ifconfig to flag these addresses as preferred.

• If you want IPv4 addresses to take precedence over IPv6 addresses, you can change the priority of ::ffff:0:0/96 to a higher number.

• If you need to assign a higher priority to deprecated addresses, you can add the deprecated address to the policy table. For example, site-local addresses are now deprecated in IPv6. These addresses have the prefix fec0::/10. You can change the policy table to give higher priority to site-local addresses.

For details about the ipaddrsel command, refer to the ipaddrsel(1M) man page.

6to4relay Command

6to4 tunneling enables communication between isolated 6to4 sites. However, to transfer packets with a native, non-6to4 IPv6 site, the 6to4 router must establish a tunnel with a 6to4 relay router. The 6to4 relay router then forwards the 6to4 packets to the IPv6 network and ultimately, to the native IPv6 site. If your 6to4-enabled site must exchange data with a native IPv6 site, you use the 6to4relay command to enable the appropriate tunnel.

Because the use of relay routers is insecure, tunneling to a relay router is disabled by default in Oracle Solaris. Carefully consider the issues that are involved in creating a tunnel to a 6to4 relay router before deploying this scenario. For detailed information on 6to4 relay routers, refer to Considerations for Tunnels to a 6to4 Relay Router. If you decide to enable 6to4 relay router support, you can find the related procedures in How to Configure a 6to4 Tunnel.

Syntax of 6to4relay

The 6to4relay command has the following syntax:

6to4relay -e [-a IPv4-address] -d -h

-e

Enables support for tunnels between the 6to4 router and an anycast 6to4 relay router. The tunnel endpoint address is then set to 192.88.99.1, the default address for the anycast group of 6to4 relay routers.

-a IPv4-address

Enables support for tunnels between the 6to4 router and a 6to4 relay router with the specified IPv4-address.

-d

Disables support for tunneling to the 6to4 relay router, the default for Oracle Solaris.

-h

Displays help for 6to4relay.

For more information, refer to the 6to4relay(1M) man page.

Example 11–3 Default Status Display of 6to4 Relay Router Support

The 6to4relay command, without arguments, shows the current status of 6to4 relay router support. This example shows the default for the Oracle Solaris implementation of IPv6.

# /usr/sbin/6to4relay
6to4relay:6to4 Relay Router communication support is disabled

Example 11–4 Status Display With 6to4 Relay Router Support Enabled

If relay router support is enabled, 6to4relay displays the following output:

# /usr/sbin/6to4relay
6to4relay:6to4 Relay Router communication support is enabled
IPv4 destination address of Relay Router=192.88.99.1

Example 11–5 Status Display With a 6to4 Relay Router Specified

If you specify the -a option and an IPv4 address to the 6to4relay command, the IPv4 address that you give with -a is displayed instead of 192.88.99.1.

6to4relay does not report successful execution of the -d, -e, and-a IPv4 address options. However, 6to4relay does display any error messages that might be generated when you run these options.

ifconfig Command Extensions for IPv6 Support

The ifconfig command enables IPv6 interfaces and the tunneling module to be plumbed. ifconfig uses an extended set of ioctls to configure both IPv4 and IPv6 network interfaces. The following describes ifconfig options that support IPv6 operations. See Monitoring the Interface Configuration With the ifconfig Command for a range of both IPv4 and IPv6 tasks that involve ifconfig.

index

Sets the interface index.

tsrc/tdst

Sets the tunnel source or destination.

addif

Creates the next available logical interface.

removeif

Deletes a logical interface with a specific IP address.

destination

Sets the point-to-point destination address for an interface.

set

Sets an address, netmask, or both for an interface.

subnet

Sets the subnet address of an interface.

xmit/-xmit

Enables or disables packet transmission on an interface.

Chapter 7, Configuring an IPv6 Network (Tasks) provides IPv6 configuration procedures.

Example 11–6 Adding a Logical IPv6 Interface With the -addif Option of the ifconfig Command

The following form of the ifconfig command creates the hme0:3 logical interface:

# ifconfig hme0 inet6 addif up
Created new logical interface hme0:3

This form of ifconfig verifies the creation of the new interface:

# ifconfig hme0:3 inet6
hme0:3: flags=2000841<UP,RUNNING,MULTICAST,IPv6> mtu 1500 index 2
		inet6  inet6 fe80::203:baff:fe11:b321/10

Example 11–7 Removing a Logical IPv6 Interface With the -removeif Option of the ifconfig Command

The following form of the ifconfig command removes the hme0:3 logical interface.

# ifconfig hme0:3 inet6 down

# ifconfig hme0 inet6 removeif 1234::5678

Example 11–8 Using ifconfig to Configure an IPv6 Tunnel Source

# ifconfig ip.tun0 inet6 plumb index 13

Opens the tunnel to be associated with the physical interface name.

# ifconfig ip.tun0 inet6
ip.tun0: flags=2200850<POINTOPOINT,RUNNING,MULTICAST,NONUD,
#IPv6> mtu 1480 index 13
		inet tunnel src 0.0.0.0 
		inet6 fe80::/10 --> :: 

Configures the streams that are needed for TCP/IP to use the tunnel device and report the status of the device.

# ifconfig ip.tun0 inet6 tsrc 120.46.86.158 tdst 120.46.86.122

Configures the source and the destination address for the tunnel.

# ifconfig ip.tun0 inet6
ip.tun0: flags=2200850<POINTOPOINT,RUNNING,MULTICAST,NONUD,
IPv6> mtu 1480 index 13
		inet tunnel src 120.46.86.158  tunnel dst 120.46.86.122
		inet6 fe80::8192:569e/10 --> fe80::8192:567a

Reports the new status of the device after the configuration.

Example 11–9 Configuring a 6to4 Tunnel Through ifconfig (Long Form)

This example of a 6to4 pseudo-interface configuration uses the subnet ID of 1 and specifies the host ID, in hexadecimal form.

# ifconfig ip.6to4tun0 inet6 plumb
# ifconfig ip.6to4tun0 inet tsrc 129.146.86.187 \
2002:8192:56bb:1::8192:56bb/64 up

# ifconfig ip.6to4tun0 inet6
ip.6to4tun0: flags=2200041<UP,RUNNING,NONUD,IPv6>mtu 1480 index 11
        inet tunnel src 129.146.86.187 
        tunnel hop limit 60 
        inet6 2002:8192:56bb:1::8192:56bb/64 

Example 11–10 Configuring a 6to4 Tunnel Through ifconfig (Short Form)

This example shows the short form for configuring a 6to4 tunnel.

# ifconfig ip.6to4tun0 inet6 plumb
# ifconfig ip.6to4tun0 inet tsrc 129.146.86.187 up

# ifconfig ip.6to4tun0 inet6
ip.6to4tun0: flags=2200041<UP,RUNNING,NONUD,IPv6>mtu 1480 index 11
        inet tunnel src 129.146.86.187 
        tunnel hop limit 60 
        inet6 2002:8192:56bb::1/64 

netstat Command Modifications for IPv6 Support

The netstat command displays both IPv4 and IPv6 network status. You can choose which protocol information to display by setting the DEFAULT_IP value in the /etc/default/inet_type file or by using the -f command-line option. With a permanent setting of DEFAULT_IP, you can ensure that netstat displays only IPv4 information. You can override this setting by using the -f option. For more information on the inet_type file, see the inet_type(4) man page.

The -p option of the netstat command displays the net-to-media table, which is the ARP table for IPv4 and the neighbor cache for IPv6. See the netstat(1M) man page for details. See How to Display the Status of Sockets for descriptions of procedures that use this command.

snoop Command Modifications for IPv6 Support

The snoop command can capture both IPv4 and IPv6 packets. This command can display IPv6 headers, IPv6 extension headers, ICMPv6 headers, and Neighbor Discovery protocol data. By default, the snoop command displays both IPv4 and IPv6 packets. If you specify the ip or ip6 protocol keyword, the snoop command displays only IPv4 or IPv6 packets. The IPv6 filter option enables you to filter through all packets, both IPv4 and IPv6, displaying only the IPv6 packets. See the snoop(1M) man page for details. See How to Monitor IPv6 Network Traffic for procedures that use the snoop command.

route Command Modifications for IPv6 Support

The route command operates on both IPv4 and IPv6 routes, with IPv4 routes as the default. If you use the -inet6 option on the command line immediately after the route command, operations are performed on IPv6 routes. See the route(1M) man page for details.

ping Command Modifications for IPv6 Support

The ping command can use both IPv4 and IPv6 protocols to probe target hosts. Protocol selection depends on the addresses that are returned by the name server for the specific target host. By default, if the name server returns an IPv6 address for the target host, the ping command uses the IPv6 protocol. If the server returns only an IPv4 address, the ping command uses the IPv4 protocol. You can override this action by using the -A command-line option to specify which protocol to use.

For detailed information, see the ping(1M) man page. For procedures that use ping, refer to Probing Remote Hosts With the ping Command.

traceroute Command Modifications for IPv6 Support

You can use the traceroute command to trace both the IPv4 and IPv6 routes to a specific host. From a protocol perspective, traceroute uses the same algorithm as ping. Use the -A command-line option to override this selection. You can trace each individual route to every address of a multihomed host by using the -a command-line option.

For detailed information, see the traceroute(1M) man page. For procedures that use traceroute, refer to Displaying Routing Information With the traceroute Command.

IPv6-Related Daemons

This section discusses the IPv6-related daemons.

in.ndpd Daemon, for Neighbor Discovery

Thein.ndpd daemon implements the IPv6 Neighbor Discovery protocol and router discovery. The daemon also implements address autoconfiguration for IPv6. The following shows the supported options of in.ndpd.

-d

Turns on debugging.

-D

Turns on debugging for specific events.

-f

Specifies a file to read configuration data from, instead of the default /etc/inet/ndpd.conf file.

-I

Prints related information for each interface.

-n

Does not loop back router advertisements.

-r

Ignores received packets.

-v

Specifies verbose mode, reporting various types of diagnostic messages.

-t

Turns on packet tracing.

The in.ndpd daemon is controlled by parameters that are set in the /etc/inet/ndpd.conf configuration file and any applicable parameters in the /var/inet/ndpd_state.interface startup file.

When the /etc/inet/ndpd.conf file exists, the file is parsed and used to configure a node as a router. Table 11–2 lists the valid keywords that might appear in this file. When a host is booted, routers might not be immediately available. Advertised packets by the router might be dropped. Also, advertised packets might not reach the host.

The /var/inet/ndpd_state.interface file is a state file. This file is updated periodically by each node. When the node fails and is restarted, the node can configure its interfaces in the absence of routers. This file contains the interface address, the last time that the file was updated, and how long the file is valid. This file also contains other parameters that are “learned” from previous router advertisements.

You do not need to alter the contents of state files. The in.ndpd daemon automatically maintains state files.

See the in.ndpd(1M) man page and the ndpd.conf(4) man page for lists of configuration variables and allowable values.

in.ripngd Daemon, for IPv6 Routing

The in.ripngd daemon implements the Routing Information Protocol next-generation for IPv6 routers (RIPng). RIPng defines the IPv6 equivalent of RIP. When you configure an IPv6 router with the routeadm command and turn on IPv6 routing, the in.ripngd daemon implements RIPng on the router.

The following shows the supported options of RIPng.

-p n

n specifies the alternate port number that is used to send or receive RIPnG packets.

-q

Suppresses routing information.

-s

Forces routing information even if the daemon is acting as a router.

-P

Suppresses use of poison reverse.

-S

If in.ripngd does not act as a router, the daemon enters only a default route for each router.

inetd Daemon and IPv6 Services

An IPv6-enabled server application can handle both IPv4 requests and IPv6 requests, or IPv6 requests only. The server always handles requests through an IPv6 socket. Additionally, the server uses the same protocol that the corresponding client uses. To add or modify a service for IPv6, use the commands available from the Service Management Facility (SMF).

• For information about the SMF commands, refer to SMF Command-Line Administrative Utilities in System Administration Guide: Basic Administration.

• For an example task that uses SMF to configure an IPv4 service manifest that runs over SCTP, refer to How to Add Services That Use the SCTP Protocol.

To configure an IPv6 service, you must ensure that the proto field value in the inetadm profile for that service lists the appropriate value:

• For a service that handles both IPv4 and IPv6 requests, choose tcp6, udp6, or sctp. A proto value of tcp6, udp6, or sctp6 causes inetd to pass on an IPv6 socket to the server. The server contains an IPv4-mapped address in case a IPv4 client has a request.

• For a service that handles only IPv6 requests, choose tcp6only or udp6only. With either of these values for proto, inetd passes the server an IPv6 socket.

If you replace an Oracle Solaris command with another implementation, you must verify that the implementation of that service supports IPv6. If the implementation does not support IPv6, then you must specify the proto value as either tcp, udp, or sctp.

Here is a profile that results from running inetadm for an echo service manifest that supports both IPv4 and IPv6 and runs over SCTP:

# inetadm -l svc:/network/echo:sctp_stream
	SCOPE    NAME=VALUE	  name="echo"
	         endpoint_type="stream"
	         proto="sctp6"
	         isrpc=FALSE
	         wait=FALSE
	         exec="/usr/lib/inet/in.echod -s"
	         user="root"
	default  bind_addr=""
	default  bind_fail_max=-1
	default  bind_fail_interval=-1
	default  max_con_rate=-1
	default  max_copies=-1
	default  con_rate_offline=-1
	default  failrate_cnt=40
	default  failrate_interval=60
	default  inherit_env=TRUE
	default  tcp_trace=FALSE
	default  tcp_wrappers=FALSE

To change the value of the proto field, use the following syntax:

# inetadm -m FMRI proto="transport-protocols"

All servers that are provided with Oracle Solaris software require only one profile entry that specifies proto as tcp6, udp6, or sctp6. However, the remote shell server (shell) and the remote execution server (exec) now are composed of a single service instance, which requires a proto value containing both the tcp and tcp6only values. For example, to set the proto value for shell, you would issue the following command:

# inetadm -m network/shell:default proto="tcp,tcp6only"

See IPv6 extensions to the Socket API in Programming Interfaces Guide for more details on writing IPv6-enabled servers that use sockets.

Considerations When Configuring a Service for IPv6

When you add or modify a service for IPv6, keep in mind the following caveats:

• You need to specify the proto value as tcp6, sctp6, or udp6 to enable both IPv4 or IPv6 connections. If you specify the value for proto as tcp, sctp, or udp, the service uses only IPv4.

• Though you can add a service instance that uses one-to-many style SCTP sockets for inetd, this is not recommended. inetd does not work with one-to-many style SCTP sockets.

• If a service requires two entries because its wait-status or exec properties differ, then you must create two instances/services from the original service.