IPv6 addresses are 128 bits in length. | |
IPv6 addresses are represented as eight 16-bit hexadecimal segments separated by colons. | |
The two rules for compacting IPv6 addresses are
| |
Using more than one double colon ambiguates the address; the exact length of each string of zeroes cannot be determined. | |
Both addresses are all zeros. ::/0 is the default address, whereas ::/128 is the unspecified address. | |
The part of a unicast IPv6 address that specifies the host is the Interface ID, and it is usually 64 bits in length. | |
The Subnet ID of the unicast IPv6 address is 16 bits long. | |
An IPv6 address beginning with FF80::/10 is a link-local address. | |
This is a global unicast address, identified by the first three bits of 001. | |
An anycast address is an address that represents a service rather than a device, and can therefore appear on more than one device. | |
A multicast address is an address that represents a group of devices rather than a single device. | |
The IPv6 header is 40 bytes in length. | |
The Flow Label field, by labeling individual flows (packets with the same source and destination address and the same source and destination ports) in the header, is intended to allow highly granular load balancing without having to pay a performance penalty from having to look into the packet payload. | |
The IPv6 Next Header field corresponds to the IPv4 Protocol Number field. It is named differently because the value of the field might specify a following protocol header or it might specify an IPv6 extension header. | |
The Hop Limit field corresponds to the IPv4 Time to Live (TTL) field. The name is changed because routers have never decremented the field according to transit time; rather, every transit router decrements the field by 1, marking a hop instead of a transit time. | |
The IPv6 Next Header field is like the IPv4 Protocol Number field in that it is an 8-bit field that can, if the next header is an upper-layer protocol header, specify the protocol number. But it is different from the Protocol Number field in that it can also specify, if the next header is an IPv6 extension header, that header’s type number. | |
Extension headers make the IPv6 header more efficient by being specialized to specific functions and only being included when the specific function is used. | |
The Next Header value of ICMPv6 (corresponding to a Protocol Number) is 58. | |
Aside from the use of the Fragment extension header, the significant difference of IPv6 fragmentation from IPv4 fragmentation is that IPv6 routers do not fragment packets. It is up to the originating host to either fragment packets or ensure that no packet it originates is too large. | |
The five ICMPv6 messages used by NDP are Router Solicitation (RS), Router Advertisement (RA), Neighbor Solicitation (NS), Neighbor Advertisement (NA), and Redirect. | |
The M flag, when set, tells hosts to use DHCPv6 to configure its address. The O flag tells hosts to used DHCPv6 to find other link parameters. | |
The Reachable Timer field specifies the time, in milliseconds, that a node should assume a neighbor is reachable after the node has confirmed reachability. | |
The Retransmit Timer field specifies the period, in milliseconds, that a node should wait between successive transmissions of an NS. | |
A Router Lifetime value of 0 in the RA indicates that the originating router should not be added to a host’s default router list. | |
The S flag, when set, indicates that the NA was sent in response to an NS. Two-way reachability is confirmed, and a neighbor address changed to Reachable state in the neighbor cache, only if the NA is in response to a solicitation; so the reception of an NA with the S bit cleared, indicating that it is unsolicited, does not change the state of a neighbor cache entry. | |
Stateful address autoconfiguration relies on DHCPv6 to allocate an address to the host. Stateless address autoconfiguration uses RAs to determine a prefix of larger scope than link-local, plus MAC-to-EUI64 conversion, to determine a host’s address. | |
MAC-to-EUI64 conversion inserts a value of 0xFFFE in the middle of a MAC address, then flips the U/L bit to 1, to create a 64-bit Interface ID from a 48-bit MAC address. | |
Duplicate address detection must never be performed on an anycast address. | |
The prefix FF02:0:0:0:0:1:FF00::/104 is used for solicited node multicast addresses. It is prepended to the last 24 bits of an address that is being solicited. | |
IPv6 uses the NDP function Neighbor Address Resolution instead of ARP, and a neighbor cache instead of an ARP cache. | |
A privacy address is one in which the Interface ID is randomly generated, and changed both at some regular period and whenever a host acquires a new prefix. It is used in conjunction with an autoconfigured public address to ensure anonymity of the host. The public address is used for reachability, but the private address is used as the source address of any packets the host originates. | |
An Incomplete state indicates that Neighbor Address Resolution for the entry is in progress. | |
A Probe state indicates that an NS has been sent to verify two-way reachability of a Stale entry, but a responding NA has not yet been received. | |
Neighbor Unreachability Detection verifies two-way reachability of a neighbor either by “hints” from an upper-layer protocol that has received an acknowledgment of a sent message, or by actively probing the neighbor with an NS. |
At a minimum, each entry of the routing table must include a destination address and the address of a next-hop router or an indication that the destination address is directly connected. | |
Variably subnetted means that the router knows of more than one subnet mask for subnets of the same major IP address. | |
Discontiguous subnets are two or more subnets of a major IP network address that are separated by a different major IP address. | |
show ip route is used to examine the routing table of a Cisco router. | |
The command show ipv6 route displays the IPv6 route table. Prefixes, prefix lengths, and the next-hop address or outgoing interface are displayed, as are the administrative distance and route metric. | |
The first bracketed number is the administrative distance of the routing protocol by which the route was learned. The second number is the metric of the route. | |
When a static route is configured to reference an exit interface instead of a next-hop address, the destination address will be entered into the routing table as directly connected. | |
A summary route is a single route entry that points to multiple subnets or major IP addresses. In the context of static routes, summary routes can reduce the number of static routes that must be configured. | |
An administrative distance is a rating of preference for a routing protocol or a static route. Every routing protocol and every static route has an administrative distance associated with it. When a router learns of a destination via more than one routing protocol or static route, it will use the route with the lowest administrative distance. | |
A floating static route is an alternative route to a destination. The administrative distance is set high enough that the floating static route is used only if a more-preferred route becomes unavailable. | |
Equal-cost load sharing distributes traffic equally among multiple paths with equal metrics. Unequal-cost load sharing distributes packets among multiple paths with different metrics. The traffic will be distributed inversely proportional to the cost of the routes. | |
If an inbound interface is configured for CEF, the packet will be switched with CEF, and CEF load balancing rules will apply: per destination or per packet for IPv4, depending upon the configuration, and per destination for IPv6. If CEF is not configured on the inbound interface, the outbound interface determines the switching mode, and thus the method of load sharing. If an outbound interface is fast switched, per destination load sharing is performed. If an interface is process switched, per packet load sharing is performed. | |
A recursive routing table lookup occurs when a router cannot acquire all the information it needs to forward a packet with a single routing table lookup. For example, the router may perform one lookup to find the route to a destination and then perform another lookup to find a route to the next-hop router of the first route. |
A routing protocol is a “language” that routers speak to each other to share information about network destinations. | |
At a minimum, a routing protocol should define procedures for
| |
A route metric, also called a route cost or a route distance, is used to determine the best path to a destination. Best is defined by the type of metric used. | |
Convergence time is the time a group of routers takes to complete the exchange of routing information. | |
Load balancing is the process of sending packets over multiple paths to the same destination. Four types of load balancing are
| |
A distance vector protocol is a routing protocol in which each router calculates routes based on the routes of its neighbors and then passes its routes to other neighbors. | |
Several problems associated with distance vector protocols are
| |
Neighbors are routers connected to the same data link. | |
Route invalidation timers delete routes from a route table if they exceed a certain age. | |
Simple split horizon does not send route information back to the source of the route information. Split horizon with poisoned reverse sends the information back to the source but sets the metric to unreachable. | |
Counting to infinity occurs when routes update a route over a loop; each router increases the metric of the route until the metric reaches infinity. The effects of counting to infinity are controlled by defining infinity as a fairly low metric so that infinity is reached fairly quickly and the route is declared unreachable. | |
Holddown timers help prevent routing loops. If a route is declared unreachable or if the metric increases beyond a certain threshold, a router will not accept any other information about that route until the holddown timer expires. This approach prevents the router from accepting possibly bad routing information while the network is reconverging. | |
A distance vector router sends its entire route table, but it only sends the table to directly connected neighbors. A link state router sends only information about its directly connected links, but it floods the information throughout the networking area. Distance vector protocols usually use a variant of the Bellman-Ford algorithm to calculate routes, and link state protocols usually use a variant of the Dijkstra algorithm to calculate routes. | |
A topological database holds the link state information originated by all routers in the link state routing domain. | |
Each router floods a link state information advertisement describing its links, the states of its links, and any neighboring routers connected to those links, throughout the networking area. All routers store all received copies of the link state advertisement in a link state database. Each router calculates a shortest path tree from the information in the topological database and enters routes in its routing tables based on the shortest path tree. | |
Sequence numbers help a router differentiate between multiple copies of the same link state advertisement and also prevent flooded link state advertisements from circulating endlessly throughout the network. | |
Aging prevents old, possibly obsolete, link state information from residing in a topological database or from being accepted by a router. | |
A router builds a shortest path tree by first adding itself as the root. Using the information in the topological database, the router creates a list of all of its directly connected neighbors. The lowest-cost link to a neighbor becomes a branch of the tree, and that router’s neighbors are added to the list. The list is checked for duplicate paths, and if they exist, the higher-cost paths are removed from the list. The lowest-cost router on the list is added to the tree, that router’s neighbors are added to the list, and the list is again checked for duplicate paths. This process continues until no routers remain on the list. | |
Within a routing domain, areas are subdomains. They make link state routing more efficient by limiting the size of the link state database of each router in the area. | |
Depending on the usage, an autonomous system can be defined as a network under a common administrative domain or a single routing domain. | |
An Interior Gateway Protocol is a routing protocol that routes within an autonomous system. An Exterior Gateway Protocol is a routing protocol that routes between autonomous systems. |
RIP uses UDP port 520. | |
RIP uses a hop count metric. An unreachable network is indicated by setting the hop count to 16, which RIP interprets as an infinite distance. | |
RIP sends periodic updates every 30 seconds minus a small random variable that prevents the updates of neighboring routers from becoming synchronized. | |
A route entry is marked as unreachable if six updates are missed. | |
The garbage collection timer, or flush timer, is set when a route is declared unreachable. When the timer expires, the route is flushed from the route table. This process allows an unreachable route to remain in the routing table long enough for neighbors to be notified of its status. | |
The random timer, whose range is one to five seconds, prevents a “storm” of triggered updates during a topology change. | |
A Request message asks a router for an update. A Response message is an update. | |
A Request message might either ask for a full update, or in some special cases, it might ask for specific routes. | |
A Response is sent when the update timer expires, or upon reception of a Request message. | |
RIP updates do not include the subnet mask of the destination address, so a RIP router depends on the subnet masks of its own interfaces to determine how an attached major network address is subnetted. If a router does not have an attachment to a particular major network address, it has no way to know how that major network is subnetted. Therefore, no subnets of a major network address can be advertised into another major network. |
The Route Tag field, the Subnet Mask field, and the Next Hop field are RIPv2 extensions that do not exist in RIPv1 messages. The basic format of the RIP message remains unchanged between the two versions; Version 2 merely uses fields that are unused in Version 1. | |
In addition to the functions that use the new fields, RIPv2 supports authentication and multicast updates. | |
RIPv2 uses the multicast address 224.0.0.9. Multicasting of routing messages is better than broadcasting because hosts and non-RIPv2 routers will ignore the multicast messages. | |
When another routing protocol uses the RIPv2 domain as a transit domain, the protocol external to RIPv2 can use the Route Tag field to communicate information to its peers on the other side of the RIPv2 domain. | |
The Next Hop field is used to inform other routers of a next-hop address on the same multiaccess network that is metrically closer to the destination than the originating router. | |
RIPv2 uses the same UDP port number as RIPv1, port number 520. | |
RIPng uses the UDP port number 521. | |
A classless routing protocol does not consider the major network address in its route lookups, but just looks for the longest match. | |
To support VLSM, a routing protocol must be able to include the subnet mask of each destination address in its updates. | |
The Cisco implementation of RIPv2 supports clear-text authentication and MD5 authentication. Only clear-text authentication is defined in RFC 2453. |
EIGRP is a distance vector protocol. | |
By default, EIGRP uses no more than 50 percent of the link’s bandwidth, based on the bandwidth configured on the router’s interface. This percentage can be changed with the command ip bandwidth-percent eigrp. | |
EIGRP and IGRP use the same formula to calculate their composite metrics, but EIGRP scales the metric by a factor of 256. The four basic components of EIGRP are
| |
Reliable delivery means EIGRP packets are guaranteed to be delivered, and they are delivered in order. RTP uses a reliable multicast, in which received packets are acknowledged, to guarantee delivery; sequence numbers are used to ensure that they are delivered in order. | |
Sequence numbers ensure that a router is receiving the most recent route entry. | |
EIGRP uses the multicast address 224.0.0.10. | |
The packet types used by EIGRP are
| |
The default EIGRP Hello interval is 5 seconds, except on some slow-speed (T1 and below) interfaces, where the default is 60 seconds. | |
The EIGRP default hold time is three times the Hello interval. | |
The neighbor table stores information about EIGRP-speaking neighbors; the topology table lists all known routes that have feasible successors. | |
The feasible distance to a destination is a router’s lowest calculated distance to the destination. | |
The feasibility condition is the rule by which feasible successors are chosen for a destination. The feasibility condition is satisfied if a neighbor’s advertised distance to a destination is lower than the receiving router’s feasible distance to the destination. In other words, a router’s neighbor meets the feasibility condition if the neighbor is metrically closer to the destination than the router. Another way to describe this is that the neighbor is “downstream” relative to the destination. | |
A feasible successor to a destination is a neighbor that satisfies the feasibility condition for that destination. | |
A successor to a destination is a feasible successor that is currently being used as the next hop to the destination. | |
A route is active on a particular router if the router has queried its neighbors for a feasible successor and has not yet received a reply from every queried neighbor. The route is passive when there are no outstanding queries. | |
A route becomes active when no feasible successor exists in its topology table. | |
An active route becomes passive when a reply has been received from every queried neighbor. | |
If a router does not receive a reply from a queried neighbor within the active time (three minutes, by default), the route is declared stuck-in-active. A response with an infinite metric is entered on the neighbor’s behalf to satisfy DUAL, and the neighbor is deleted from the neighbor table. | |
Subnetting is the practice of creating a group of subnet addresses from a single IP network address. Address aggregation is the practice of summarizing a group of network or subnet addresses with a single IP network address. | |
From the perspective of an OSPF router, a neighbor is another OSPF router that is attached to one of the first router’s directly connected links. | |
An OSPF adjacency is a conceptual link to a neighbor over which LSAs can be sent. | |
The five OSPF packet types, and their purposes, are
| |
A router originates a link-state advertisement to describe one or more destinations. An OSPF Update packet transports LSAs from one neighbor to another. Although LSAs are flooded throughout an area or OSPF domain, Update packets never leave a data link. | |
The most common LSA types and their purposes are
| |
The link state database is where a router stores all the OSPF LSAs it knows of, including its own. Database synchronization is the process of ensuring that all routers within an area have identical link-state databases. | |
The default OSPF HelloInterval is 10 seconds. | |
The default RouterDeadInterval is four times the HelloInterval. | |
A Router ID is an address by which an OSPF router identifies itself. It is either the numerically highest IP address of all the router’s loopback interfaces, or if no loopback interfaces are configured, it is the numerically highest IP address of all the router’s LAN interfaces. It can also be manually configured. | |
An area is an OSPF sub-domain, within which all routers have an identical link-state database. | |
Area 0 is the backbone area. All other areas must send their inter-area traffic through the backbone. | |
MaxAge, one hour, is the age at which an LSA is considered to be obsolete. | |
The four OSPF router types are
| |
The four OSPF path types are
| |
What are the five OSPF network types? The five OSPF network types are
| |
A Designated Router is a router that represents a multiaccess network, and the routers connected to the network, to the rest of the OSPF domain. | |
Cisco IOS calculates the outgoing cost of an interface as 108/BW, where BW is the configured bandwidth of the interface. 108 can be changed with the OSPF command auto-cost reference-bandwidth. | |
An area is partitioned if one or more of its routers cannot send a packet to the area’s other routers without sending the packet out of the area. | |
A virtual link is a tunnel that extends an OSPF backbone connection through a non-backbone area. | |
A stub area is an area into which no type 5 LSAs are flooded. A totally stubby area is an area into which no type 3, 4, or 5 LSAs are flooded, with the exception of type 3 LSAs to advertise a default route. Not-so-stubby areas are areas through which external destinations are advertised into the OSPF domain, but into which no type 5 LSAs are sent by the ABR. | |
OSPF network entries are entries in the route table, describing IP destinations. OSPF router entries are entries in a separate route table that record only routes to ABRs and ASBRs. | |
Type 2 authentication uses MD5 encryption, whereas type 1 authentication uses clear-text passwords. | |
The three fields in the LSA header that distinguish different LSAs are the Type, Advertising Router, and the Link State ID fields. The three fields in the LSA header that distinguish different instances of the same LSA are the Sequence Number, Age, and Checksum fields. |
OSPFv3 cannot support IPv4 at the time of this writing. To route both IPv4 and IPv6 with OSPF, you must run both OSPFv2 and OSPFv3. | |
Multiple instance per link means that there can be separate adjacencies among different routers all connected to the same broadcast link, so that different OSPFv3 routing domains can use the same shared link without interfering with or having knowledge of each other. The Instance ID field in the OSPFv3 header makes this possible. | |
OSPFv3 packets are authenticated using the built-in IPv6 authentication (by means of the IPv6 Authentication extension header). OSPFv3 does not have its own authentication mechanism as OSPFv2 does. | |
The OSPFv3 Next Header number is the same as the OSPFv2 Protocol Number, 89. | |
OSPFv3 uses the reserved multicast addresses FF02::5 (AllSPFRouters) and FF02::6 (AllDRouters). | |
No. OSPFv3 uses the same five message types as OSPFv2. | |
The first bit is the U bit, specifying how the receiving router should treat the LSA if its type is unknown. The second and third bits are the S bits, indicating the flooding scope of the LSA. | |
OSPFv3 supports a link-local flooding scope that is not supported by OSPFv2. The Link LSA uses this flooding scope. | |
OSPFv3 Router and Network LSAs do not advertise prefixes, as OSPFv2 Router and Network LSAs do. | |
The Intra-Area Prefix LSA carries IPv6 prefixes connected to the originating router. | |
The Link LSA carries information that is only significant between two directly connected neighbors. |
An Intermediate System is the ISO term for a router. | |
A Network Protocol Data Unit is the ISO term for a packet. | |
An L1 router has no direct connections to another area. An L2 router has no adjacencies with L1 routers. An L1/L2 router routes both inter-area and intra-area traffic and acts as an inter-area gateway for L1 routers. | |
Cisco routers by default are L1/L2. | |
The borders of IS-IS areas are between routers, on links. The borders of OSPF areas are defined by the routers themselves. | |
Two L1/L2 routers with the same AIDs will form both an L1 and an L2 adjacency. Two L1/L2 routers with different AIDs will form an L2 adjacency. | |
Two L2-only routers will form an L2 adjacency, whether the AIDs are the same or different. | |
The Network Entity Title is an address by which a router identifies both itself and the area in which it resides. | |
The NSAP Selector should be set to 0x00 in a NET. | |
The System ID uniquely identifies a router within an IS-IS domain. | |
The portion of the NET preceding the last seven octets is the area address. | |
IS-IS does not elect a BDR. | |
The Pseudonode ID is the last octet of a LAN ID. Its purpose is to distinguish LAN IDs that are originated by a single router that is the DR on multiple LANs. | |
The MaxAge of an IS-IS LSP is 1200 seconds (20 minutes). The MaxAge (or beginning Remaining Lifetime) can be configured up to 65,535 seconds. | |
OSPF increments the age up to MaxAge; IS-IS decrements the age down to 0. A new OSPF LSA has an age of 0, whereas a new IS-IS LSP has an age of MaxAge. | |
The refresh rate of an IS-IS router is 900 seconds (15 minutes). | |
A Complete Sequence Number Packet contains a full listing of all LSPs in a database. A CSNP is periodically sent by the Designated Router on a broadcast network to maintain database synchronization. | |
A Partial Sequence Number Packet contains a listing of one or more LSPs. It has two uses: On point-to-point networks, it is used to acknowledge the receipt of LSPs. On broadcast networks, it is used to request LSPs. | |
An IS-IS router uses the Overload bit to inform its neighbors that it is experiencing a memory overload and cannot store the entire link-state database. | |
The Attached bit is used by L1/L2 routers to inform L1 routers that it is attached to the L2 backbone. | |
The Up/down bit is used to distinguish between an address that originated within an area, or an address that was leaked into an area. | |
The ISO specifies four metrics: Default, Expense, Delay, and Error. Cisco supports only the Default metric. | |
The two metric styles are narrow and wide. The narrow metric has a maximum value of 63. The wide metric has a maximum value of 16777214. | |
The maximum metric value of an IS-IS route is 1023 for narrow metrics and 4261412864 for wide metrics. | |
L1 IS-IS metrics apply to intra-area routes, and L2 IS-IS metrics apply to inter-area routes. | |
Internal metrics apply to routes to destinations within the IS-IS domain. External metrics apply to routes to destinations external to the IS-IS domain. | |
A single adjacency is formed between two routers, even if both IPv4 and IPv6 are configured in multi-topology mode. | |
L1 areas may be configured on a router. One L2 area is configured. | |
The two active mesh group modes are Blocked and Set (Numbered). Blocked mode offers the most reduced flooding, but at the possible cost of the most reduced redundancy and increased convergence time. Set mode, or numbered mesh groups, do not reduce flooding load as much as Blocked mode, but also have less potential impact on redundancy and convergence time. |
Routes that are learned from another routing protocol, between two processes of the same routing protocol, from static routes, or from a direct connection to the destination network can be redistributed into a routing domain. Routes can also be redistributed between IS-IS levels 1 and 2. | |
In contrast to metrics, which are used to determine the best path among multiple routes to the same destination discovered by the same routing protocol, administrative distances are used to determine the best path among multiple routes to the same destination discovered by different routing protocols. | |
A route to a destination within a routing domain with a higher administrative distance can be redistributed into a routing domain with a lower administrative distance. If that route is redistributed back into the higher-distance domain, packets might be misrouted into the lower-distance domain. | |
Redistributing variably subnetted destination addresses from a classless domain into a classful domain can cause problems. The classful domain might not be able to recognize all the subnets attempting to be redistributed from the classless domain. | |
OSPF and IS-IS understand the default metric. RIP, IGRP, and EIGRP do not. | |
The metric command assigns a metric to specific redistribution statements. The default-metric command assigns a metric to all redistribution commands that do not include the metric command. | |
Without the subnets keyword, only major network addresses that are not directly connected to the router will be redistributed. | |
A router that originates a summary route should use the null interface as the next hop of the summary route. Any packets that match the summary route, but for which there is no more-specific route to the packet’s destination address, will be dropped. This prevents the router from forwarding “lost” packets. |
The IPv4 default route address is 0.0.0.0. | |
The IPv6 default prefix/prefix length is ::/0. | |
EIGRP advertises a default address as an external address type. | |
Yes. | |
A stub router is a router with only a single link to another router. A stub network is a network with only one attached router. | |
Using a default route rather than a full route table can conserve router memory by keeping the table small and can save router processing cycles by limiting the routing information that must be processed. | |
Using a full route table rather than a default route can make routing more accurate. | |
ODR uses Cisco Discovery Protocol (CDP) to discover routes. | |
ODR is available in IOS 11.2 and later. | |
The medium over which ODR is to run must support SNAP. |
Route maps are similar to access lists in that they define match criteria and an action to take in the event of a match. Route maps are different from access lists in that they not only specify match criteria but also specify set criteria. The set action can modify a route, or route a packet according to the parameters of the packet. | |
Policy routes are static routes that use route maps to determine which packets should be routed and where the packets should be routed. | |
Route tags are fields within routing information packets that allow external information to be carried through the routing domain. | |
Route tags have no effect on the routing protocols that carry them. |