A host performs several steps to autoconfigureits interfaces in IPv6. The autoconfiguration process creates a link-localaddress. The autoconfiguration process verifies its uniqueness on a link.The process also determines which information should be autoconfigured, addresses,other information, or both. The process determines if the addresses shouldbe obtained through the stateless mechanism, the stateful mechanism, or bothmechanisms. This section describes the process for generating a link-localaddress. This section also describes the process for generating site-localand global addresses by stateless address autoconfiguration. Finally, thissection describes the procedure for duplicate address detection.
Stateless Autoconfiguration Requirements
IPv6 defines mechanisms for both stateful address and stateless addressautoconfiguration. Stateless autoconfiguration requires no manual configurationof hosts, minimal (if any) configuration of routers, and no additional servers.The stateless mechanism enables a host to generate its own addresses. Thestateless mechanism uses local information as well as non-local informationthat is advertised by routers to generate the addresses. Routers advertiseprefixes that identify the subnet or subnets that are associated with a link. Hosts generate an interface identifier that uniquelyidentifies an interface on a subnet. An address is formed by combining theprefix and the interface identifier. In the absence of routers, a host cangenerate only link-local addresses. However, link-local addresses are onlysufficient for allowing communication among nodes that are attached to thesame link.
Stateful Autoconfiguration Model
In the stateful autoconfiguration model, hosts obtain interfaceaddresses or configuration information and parameters from a server. Serversmaintain a database that checks which addresses have been assigned to whichhosts. The stateful autoconfiguration protocol allows hosts to obtain addressesand other configuration information from a server. Stateless and statefulautoconfiguration complement each other. For example, a host can use statelessautoconfiguration to configure its own addresses, but use stateful autoconfigurationto obtain other information.
When to Use Stateless and Stateful Approaches
The stateless approach is used when a site is not concerned with theexact addresses that hosts use. However, the addresses must be unique. Theaddresses must also be properly routable. The stateful approach is used whena site requires more precise control over exact address assignments. Statefuland stateless address autoconfiguration can be used simultaneously. The siteadministrator specifies which type of autoconfiguration to use through thesetting of appropriate fields in router advertisem*nt messages.
IPv6addresses are leased to an interface for a fixed, possibly infinite, lengthof time. Each address has an associated lifetime that indicates how long theaddress is bound to an interface. When a lifetime expires, the binding, andaddress, become invalid and the address can be reassigned to another interfaceelsewhere. To handle the expiration of address bindings gracefully, an addressexperiences two distinct phases while the address is assigned to an interface.Initially, an address is preferred, meaning that its use in arbitrary communicationis unrestricted. Later, an address becomes deprecatedin anticipation that its current interface binding becomes invalid. When theaddress is in a deprecated state, the use of the address is discouraged, butnot strictly forbidden. New communication, for example, the opening of a newTCP connection, should use a preferred address when possible. A deprecatedaddress should be used only by applications that have been using the address.Applications that cannot switch to another address without a service disruptioncan use a deprecated address.
Duplicate Address Detection Algorithm
To ensure that all configured addresses arelikely to be unique on a particular link, nodes run a duplicateaddress detection algorithm on addresses. The nodes must run thealgorithm before assigning the addresses to an interface. The duplicate addressdetection algorithm is performed on all addresses.
The autoconfiguration process that is specified in this document appliesonly to hosts and not routers. Because host autoconfiguration uses informationthat is advertised by routers, routers need to be configured by some othermeans. However, routers probably generate link-local addresses by using themechanism that is described in this document. In addition, routers are expectedto pass successfully the duplicate address detection procedure on all addressesprior to assigning the address to an interface.
Autoconfiguration Process
This section provides an overview ofthe typical steps that are performed by an interface during autoconfiguration.Autoconfiguration is performed only on multicast-capable links. Autoconfigurationbegins when a multicast-capable interface is enabled, for example, duringsystem startup. Nodes, both hosts and routers, begin the autoconfigurationprocess by generating a link-local address for the interface. A link-localaddress is formed by appending the interface's identifier to the well-knownlink-local prefix.
A node must attempt to verify that a tentative link-local address isnot already in use by another node on the link. After verification, the link-localaddress can be assigned to an interface. Specifically, the node sends a neighborsolicitation message that contains the tentative address as the target. Ifanother node is already using that address, the node returns a neighbor advertisem*ntsaying that the node is using that address. If another node is also attemptingto use the same address, the node also sends a neighbor solicitation for thetarget. The number of neighbor solicitation transmissions or retransmissions,and the delay between consecutive solicitations, are link specific. Theseparameters can be set by system management.
If a node determines that its tentative link-local address is not unique,autoconfiguration stops and manual configuration of the interface is required.To simplify recovery in this instance, an administrator can supply an alternateinterface identifier that overrides the default identifier. Then, the autoconfigurationmechanism can be applied by using the new, presumably unique, interface identifier.Alternatively, link-local and other addresses need to be configured manually.
After a node determines that its tentative link-local address is unique,the node assigns the address to the interface. At this point, the node hasIP-level connectivity with neighboring nodes. The remaining autoconfigurationsteps are performed only by hosts.
Obtaining Router Advertisem*nt
The next phase of autoconfiguration involves obtaininga router advertisem*nt or determining that no routers are present. If routersare present, the routers send router advertisem*nts that specify what typeof autoconfiguration a host should perform. If no routers are present, statefulautoconfiguration is invoked.
Routerssend router advertisem*nts periodically. However, the delay between successiveadvertisem*nts is generally longer than a host that performs autoconfigurationcan wait. To obtain an advertisem*nt quickly, a host sends one or more routersolicitations to the all-routers multicast group. Router advertisem*nts containtwo flags that indicate what type of stateful autoconfiguration (if any) shouldbe performed. A managed address configuration flag indicateswhether hosts should use stateful autoconfiguration to obtain addresses. An other stateful configuration flag indicates whether hosts shoulduse stateful autoconfiguration to obtain additional information, excludingaddresses.
Prefix Information
Router advertisem*nts also contain zero or moreprefix information options that contain information that stateless addressautoconfiguration uses to generate site-local and global addresses. The statelessaddress and stateful address autoconfiguration fields in router advertisem*ntsare processed independently. A host can use both stateful address and statelessaddress autoconfiguration simultaneously. One option field that contains prefixinformation, the autonomous address-configuration flag,indicates whether the option even applies to stateless autoconfiguration.If the option field does apply, additional option fields contain a subnetprefix with lifetime values. These values indicate how long addresses thatare created from the prefix remain preferred and valid.
Because routers generate router advertisem*nts periodically, hosts continuallyreceive new advertisem*nts. Hosts process the information that is containedin each advertisem*nt as described previously. Hosts add to the information.Hosts also refresh the information that is received in previous advertisem*nts.
Address Uniqueness
For safety, all addresses must be tested for uniqueness priorto their assignment to an interface. The situation is different for addressesthat are created through stateless autoconfiguration. The uniqueness of anaddress is determined primarily by the portion of the address that is formedfrom an interface identifier. Thus, if a node has already verified the uniquenessof a link-local address, additional addresses need not be tested individually.The addresses must be created from the same interface identifier. In contrast,all addresses that are obtained manually should be tested individually foruniqueness. The same is true for addresses that are obtained by stateful addressautoconfiguration. Some sites believe that the overhead of performing duplicateaddress detection outweighs its benefits. For these sites, the use of duplicateaddress detection can be disabled by setting a per-interface configurationflag.
To accelerate the autoconfiguration process, a host can generate itslink-local address, and verify its uniqueness, while the host waits for arouter advertisem*nt. A router might delay a response to a router solicitationfor a few seconds. Consequently, the total time necessary to complete autoconfigurationcan be significantly longer if the two steps are done serially.
