ISIS
ISIS is a routing protocol which is described in ISO10589, RFC 1195, RFC 5308. ISIS is an IGP. Compared with RIP, ISIS can provide scalable network support and faster convergence times like OSPF. ISIS is widely used in large networks such as ISP and carrier backbone networks.
Configuring isisd
There are no isisd specific options. Common options can be specified (Common Invocation Options) to isisd. isisd needs to acquire interface information from zebra in order to function. Therefore zebra must be running before invoking isisd. Also, if zebra is restarted then isisd must be too.
Configuration for the daemon should be saved in the FRR integrated configuration file located in /etc/frr/frr.conf, see Integrated Config File for more information on system configuration.
Prior versions of FRR supported reading and writing per-daemon config files;
however, with the introduction of the centralized management daemon mgmtd
this could no longer be supported.
In order to allow for an orderly transition from per-daemon config files to the
integrated config file, FRR daemons will continue to try and read their
specific per-daemon configuration file as before. Additionally the config can
still be loaded directly using the -f or --config-file CLI options;
however, these files will not be updated when the configuration is written
(e.g., with the write mem command).
Warning
Per-daemon files will no longer be updated when the user issues a write
memory command. Therefore these per-daemon config files should only be used
as a mechanism for transitioning to the integrated config, and then removed.
ISIS router
To start the ISIS process you have to specify the ISIS router. As of this writing, isisd does not support multiple ISIS processes.
- router isis WORD [vrf NAME]
Enable or disable the ISIS process by specifying the ISIS domain with ‘WORD’. isisd does not yet support multiple ISIS processes but you must specify the name of ISIS process. The ISIS process name ‘WORD’ is then used for interface (see command
ip router isis WORD).
- net XX.XXXX. ... .XXX.XX
Set/Unset network entity title (NET) provided in ISO format.
- hostname dynamic
Enable support for dynamic hostname.
- area-password [clear | md5] <password> [authenticate snp <send-only|validate>]
- domain-password [clear | md5] <password> [authenticate snp <send-only|validate>]
Configure the authentication password for an area, respectively a domain, as clear text or md5 one. With
authenticate snp, the Sequence Number PDUs of the level are authenticated too: sent authenticated, and also verified withvalidate.
- area-password key-chain KEYCHAIN [strict-rfc5310] [authenticate snp <send-only|validate>]
- domain-password key-chain KEYCHAIN [strict-rfc5310] [authenticate snp <send-only|validate>]
Authenticate the LSPs of an area, respectively a domain, and with
authenticate snpits Sequence Number PDUs, with the keys of a key chain (RFC 5310);strict-rfc5310derives the HMAC key from them as section 3.3 of RFC 5310 does. See ISIS Authentication.
- attached-bit [receive ignore | send]
Set attached bit for inter-area traffic:
receive If LSP received with attached bit set, create default route to neighbor
send If L1|L2 router, set attached bit in LSP sent to L1 router
- log-adjacency-changes
Log changes in adjacency state.
- log-pdu-drops
Log any dropped PDUs.
- metric-style [narrow | transition | wide]
Set old-style (ISO 10589) or new-style packet formats:
narrow Use old style of TLVs with narrow metric
transition Send and accept both styles of TLVs during transition
wide Use new style of TLVs to carry wider metric. FRR uses this as a default value
- advertise-high-metrics
Advertise high metric value on all interfaces to gracefully shift traffic off the router. Reference: RFC 3277
For narrow metrics, the high metric value is 63; for wide metrics, 16777215; for transition metrics, 62.
- set-overload-bit
Set overload bit to avoid any transit traffic.
- set-overload-bit on-startup (0-86400)
Set overload bit on startup for the specified duration, in seconds. Reference: RFC 3277
- lsp-mtu (128-4352)
Configure the maximum size of generated LSPs, in bytes.
- advertise-passive-only
Advertise prefixes of passive interfaces only.
ISIS Timer
- lsp-gen-interval [level-1 | level-2] (1-120)
Set minimum interval in seconds between regenerating same LSP, globally, for an area (level-1) or a domain (level-2).
- lsp-refresh-interval [level-1 | level-2] (1-65235)
Set LSP refresh interval in seconds, globally, for an area (level-1) or a domain (level-2).
- max-lsp-lifetime [level-1 | level-2] (350-65535)
Set LSP maximum LSP lifetime in seconds, globally, for an area (level-1) or a domain (level-2).
- spf-interval [level-1 | level-2] (1-120)
Set minimum interval between consecutive SPF calculations in seconds.
ISIS Fast-Reroute
Unless stated otherwise, commands in this section apply to all LFA flavors (local LFA, Remote LFA and TI-LFA).
- spf prefix-priority [critical | high | medium] WORD
Assign a priority to the prefixes that match the specified access-list.
By default loopback prefixes have medium priority and non-loopback prefixes have low priority.
- fast-reroute priority-limit [critical | high | medium] [level-1 | level-2]
Limit LFA backup computation up to the specified prefix priority.
- fast-reroute lfa tiebreaker [downstream | lowest-backup-metric | node-protecting] index (1-255) [level-1 | level-2]
Configure a tie-breaker for multiple local LFA backups. Lower indexes are processed first.
- fast-reroute load-sharing disable [level-1 | level-2]
Disable load sharing across multiple LFA backups.
- fast-reroute remote-lfa prefix-list [WORD] [level-1 | level-2]
Configure a prefix-list to select eligible PQ nodes for remote LFA backups (valid for all protected interfaces).
- redistribute <ipv4 | ipv6> table (1-65535) <level-1 | level-2> [metric (0-16777215)|route-map WORD]
Redistribute routes from a given routing table into the given ISIS level database. Any default route will be ignored unless the default-information originate command is specified.
- default-information originate <ipv4 | ipv6> <level-1 | level-2> [always] [metric (0-16777214)|route-map WORD]
Originate a default route into the given ISIS level database. The always command will turn it on by default even if there is no default route in the system.
ISIS region
- is-type [level-1 | level-1-2 | level-2-only]
Define the ISIS router behavior:
level-1 Act as a station router only
level-1-2 Act as both a station router and an area router
level-2-only Act as an area router only
- leak level-1 into level-2
A level-1-2 router advertises the prefixes reachable inside its level-1 area in its level-2 LSP (RFC 1195 section 1.3), so that level-2 routers can reach them. The prefixes are taken from the level-1 LSPs of the routers that are reachable in the level-1 SPT, with the metric of the path to the advertising router added to the metric of the prefix. Prefixes that carry the up/down bit are not advertised back into level-2 (RFC 5302 section 2). This requires wide metrics. It is disabled by default, so that the level-2 routers do not learn more prefixes after an upgrade; use this command on the level-1-2 routers to enable it, and the
noform to disable it again. Only the leaking of level-1 prefixes into level-2 is supported.
ISIS interface
- <ip|ipv6> router isis WORD
Activate ISIS adjacency on this interface. Note that the name of ISIS instance must be the same as the one used to configure the ISIS process (see command
router isis WORD). To enable IPv4, issueip router isis WORD; to enable IPv6, issueipv6 router isis WORD.
BFD monitoring can be enabled for IS-IS adjacencies on an interface with
isis bfd. BFD provides faster failure detection than the IS-IS hello
timers. Use isis bfd profile BFDPROF to apply a BFD profile to these
sessions.
See also
- isis circuit-type [level-1 | level-1-2 | level-2]
Configure circuit type for interface:
level-1 Level-1 only adjacencies are formed
level-1-2 Level-1-2 adjacencies are formed
level-2-only Level-2 only adjacencies are formed
- isis csnp-interval (1-600) [level-1 | level-2]
Set CSNP interval in seconds globally, for an area (level-1) or a domain (level-2).
- isis hello padding
Add padding to IS-IS hello packets.
- isis hello padding during-adjacency-formation
Add padding to IS-IS hello packets during adjacency formation only.
- isis hello-interval [level-1 | level-2] (1-600)
Set Hello interval in seconds globally, for an area (level-1) or a domain (level-2).
- isis hello-multiplier [level-1 | level-2] (2-100)
Set multiplier for Hello holding time globally, for an area (level-1) or a domain (level-2).
- isis metric [level-1 | level-2] [(0-63) | (0-16777215)]
Set default metric value globally, for an area (level-1) or a domain (level-2). Max value depend if metric support narrow or wide value (see command
metric-style [narrow | transition | wide]).With wide metrics, the maximum metric 16777215 (2^24 - 1) is still advertised, but links advertised with it are ignored during the SPF computation, as required by RFC 5305.
- isis network point-to-point
Set network type to ‘Point-to-Point’ (broadcast by default).
- isis passive
Configure the passive mode for this interface.
- isis password [clear | md5] <password>
Configure the authentication password (clear or encoded text) for the interface.
- isis password key-chain KEYCHAIN [strict-rfc5310]
Authenticate the hellos of the interface with the keys of a key chain (RFC 5310);
strict-rfc5310derives the HMAC key from them as section 3.3 of RFC 5310 does. See ISIS Authentication.
- isis extended-sequence-number <send|verify>
Add an Extended Sequence Number (RFC 7602) to the IIHs and SNPs sent on the interface, and with
verifyalso discard those received without a newer one. See Extended Sequence Numbers.
- isis priority (0-127) [level-1 | level-2]
Set priority for Designated Router election, globally, for the area (level-1) or the domain (level-2).
- isis psnp-interval (1-120) [level-1 | level-2]
Set PSNP interval in seconds globally, for an area (level-1) or a domain (level-2).
- isis three-way-handshake
Enable or disable RFC 5303 Three-Way Handshake for P2P adjacencies. Three-Way Handshake is enabled by default.
- isis periodic-csnp
Enable or disable periodic CSNPs on a point-to-point circuit. When disabled, a complete set of CSNPs is sent only when an adjacency comes up, as described in RFC 5303, instead of every
isis csnp-interval. This saves bandwidth, at the cost of the periodic database check. Periodic CSNPs are enabled by default. While they are disabled,isis csnp-intervalis not used on point-to-point circuits, and setting a CSNP interval other than the default prints a warning. On a broadcast circuit the DIS keeps sending periodic CSNPs: the setting has no effect there, and setting it prints a warning.
- isis ipv4-over-ipv6-nexthop
Route IPv4 over this interface when it has no usable IPv4 nexthop, such as when the interface has no IPv4 address. On an interface running both
ip router isisandipv6 router isis, an adjacency whose IPv4 addressing is missing on either side is used for IPv4 through the neighbor’s IPv6 link-local address, which becomes the nexthop of the IPv4 routes installed through it (as with BGP and RFC 8950). Adjacencies with usable IPv4 addressing keep IPv4 nexthops, so the option can be enabled on some interfaces and not others.The kernel must support IPv4 routes with IPv6 nexthops (Linux 5.2 or later). Nothing in IS-IS tells a neighbor whether this is enabled, so enable it on every router attached to a link without IPv4: a router without it does not route IPv4 over that link, while its neighbors still forward IPv4 to it over it.
Segment Routing prefix SIDs and Flexible Algorithm labels work over IPv6 link-local nexthops. Adjacency SIDs, and therefore TI-LFA repair paths, are only allocated for IPv4 adjacencies that have IPv4 addresses.
LDP does not label IPv4 FECs over such links, so remote LFA cannot protect prefixes through them. With LDP-IGP synchronization the link needs LDP over IPv6: IPv4 LDP cannot form without IPv4 on the link, and the link then stays at the maximum metric. MPLS-TE advertises the link’s attributes, but a link without interface addresses gets no edge in the Traffic Engineering Database.
- isis fast-reroute lfa [level-1 | level-2]
Enable per-prefix local LFA fast reroute link protection.
- isis fast-reroute lfa [level-1 | level-2] exclude interface IFNAME
Exclude an interface from the local LFA backup nexthop computation.
- isis fast-reroute remote-lfa tunnel mpls-ldp [level-1 | level-2]
Enable per-prefix Remote LFA fast reroute link protection. Note that other routers in the network need to be configured to accept LDP targeted hello messages in order for RLFA to work.
- isis fast-reroute remote-lfa maximum-metric (1-16777215) [level-1 | level-2]
Limit Remote LFA PQ node selection within the specified metric.
- isis fast-reroute ti-lfa [level-1|level-2] [node-protection [link-fallback]]
Enable per-prefix TI-LFA fast reroute link or node protection. When node protection is used, option link-fallback enables the computation and use of link-protecting LFAs for destinations unprotected by node protection.
ISIS Authentication
Besides clear text and HMAC-MD5 (RFC 5304) passwords, the hellos of an
interface (isis password key-chain KEYCHAIN [strict-rfc5310]) and the LSPs and SNPs
of a level (area-password key-chain KEYCHAIN [strict-rfc5310] [authenticate snp <send-only|validate>],
domain-password key-chain KEYCHAIN [strict-rfc5310] [authenticate snp <send-only|validate>])
can be authenticated with the keys of a key chain, as RFC 5310 generic
cryptographic authentication.
A key is used only if it has a key-string, an HMAC-SHA
cryptographic-algorithm, and an index of at most 65535, which is sent as
the Key ID. The internal cryptographic implementation, the default, only
provides hmac-sha-256; hmac-sha-1, hmac-sha-384 and
hmac-sha-512 need FRR built with --with-crypto=openssl.
A PDU is signed with the first key of the chain, in Key ID order, that is usable and valid for sending, and verified with the key its Key ID names, which must be valid for accepting. Hellos and SNPs are not sent while the chain has no key to sign them with, as RFC 5310 requires: the adjacencies of the interface go down. LSPs stay in the database and are sent unauthenticated, for the neighbors that verify them to discard.
Keys are changed with their lifetimes. A new key needs an accept lifetime on
every router before its send lifetime starts on any, and the old one must
stay valid for accepting until every router sends with the new one. LSPs are
signed when they are generated, so isisd regenerates them whenever a send
lifetime of the key chain starts or ends, and whenever the key chain changes;
as any regeneration, that can wait for the lsp-gen-interval, which the
overlap of the two keys has to cover.
There is no transition mode: a router that authenticates PDUs discards those that are not, so authentication has to be enabled on all the routers of a link (hellos) or of a level (LSPs) at the same time.
The HMAC key is the key string, hashed when it is longer than the block size
of the hash, as plain HMAC (RFC 2104) has it. Section 3.3 of RFC 5310
hashes the keys longer than the hash instead, which none of the
implementations tested does (IOS XR, EOS, Holo); strict-rfc5310 selects
it, for peers that do. The two only differ for keys longer than the hash and
no longer than its block (33 to 64 bytes for HMAC-SHA-256), so the option
does not matter with shorter keys. Arista EOS computes RFC 5310 digests
only with the rfc-5310 keyword of its authentication commands.
The state of the authentication is shown, never the passwords: by
show isis [vrf <NAME|all>] interface [detail] [IFNAME] [json] for the
hellos of an active interface and by
show isis [vrf <NAME|all>] summary [json] for the LSPs and SNPs of each
level. They give the kind of authentication, for a key chain the Key ID and
algorithm it signs with now (or that no key is valid for sending), and how
many received PDUs failed authentication (failures) or did not carry the
kind expected (type failures).
show isis [vrf <NAME|all>] neighbor [detail] [SYSTEMID] [json] gives the
Key ID of each neighbor’s last hello and
show isis [vrf <NAME|all>] database [detail] [LSPID] [json] that of each
LSP, which shows a key change reaching every router:
r1# show isis interface detail
...
Authentication: key chain isis, sending with Key ID 2 (hmac-sha-256)
Failures: 0, type failures: 0
r1# show isis summary
...
Level-2:
...
Authentication: key chain isis, sending with Key ID 2 (hmac-sha-256), SNP authentication: validate
Failures: 0, type failures: 0
r1# show isis neighbor detail
...
Authenticated with Key ID 2
key chain isis
key 1
key-string a-secret
cryptographic-algorithm hmac-sha-256
exit
exit
!
interface eth0
ip router isis 1
isis password key-chain isis
isis extended-sequence-number verify
exit
!
router isis 1
net 49.0001.0000.0000.0001.00
area-password key-chain isis authenticate snp validate
domain-password key-chain isis authenticate snp validate
exit
Extended Sequence Numbers
Authentication does not stop a captured hello or SNP from being sent again.
With isis extended-sequence-number, the IIHs and SNPs
sent on an interface carry an Extended Sequence Number (RFC 7602): a
64-bit session number and a 32-bit packet number, counted for the whole area so
that they keep increasing however interfaces come and go. With verify, the
IIHs and SNPs received on the interface without exactly one well-formed one, or
with one not greater than the last one accepted from the same neighbor for the
same PDU type, are discarded. The number is only worth checking in a PDU that
cannot be tampered with, so verify applies to the IIHs of an interface
authenticated with HMAC-MD5 or a key chain, and to the SNPs of a level that
validates them that way (authenticate snp validate); the others are
accepted without their number being checked or remembered. LSPs never carry
one: their own sequence numbers already defeat replay.
Enable send on all the routers of a link before verify on any of
them, and disable verify on all of them before send: a router that
verifies discards the PDUs of a neighbor that does not send the number. The
mode applies to both levels of a LAN.
When isisd starts sending numbers, the session number of an area becomes the
current time, or one more than the last one, kept in the isisd state file
(isisd.json in the FRR state directory), whichever is greater
(Appendix A of RFC 7602): a restarted router keeps numbering above what its
neighbors have seen as long as either its clock or that file survives.
The last numbers accepted from a neighbor are kept with the interface, not with the adjacency, so that a replayed hello cannot bring back an adjacency with a neighbor that went away; they are forgotten when the interface goes down. A router that lost both its state file and its clock has its PDUs discarded until then.
show isis interface detail shows the mode, the session number, and
the PDUs discarded without a valid Extended Sequence Number or as replayed.
Showing ISIS information
- show isis [vrf <NAME|all>] summary [json]
Show summary information about ISIS, including the authentication of each level.
- show isis [vrf <NAME|all>] hostname
Show information about ISIS node.
- show isis [vrf <NAME|all>] interface [detail] [IFNAME] [json]
Show state and configuration of ISIS specified interface, or all interfaces if no interface is given with or without details.
- show isis [vrf <NAME|all>] neighbor [detail] [SYSTEMID] [json]
Show state and information of ISIS specified neighbor, or all neighbors if no system id is given with or without details.
- show isis [vrf <NAME|all>] database [detail] [LSPID] [json]
Show the ISIS database globally, for a specific LSP id without or with details.
- show isis [vrf <NAME|all>] topology [level-1|level-2] [algorithm [(128-255)]]
Show topology IS-IS paths to Intermediate Systems, globally, in area (level-1) or domain (level-2).
- show isis [vrf <NAME|all>] route [level-1|level-2] [prefix-sid] [backup] [algorithm [(128-255)]]
Show the ISIS routing table, as determined by the most recent SPF calculation.
- show isis [vrf <NAME|all>] fast-reroute summary [level-1|level-2]
Show information about the number of prefixes having LFA protection, and network-wide LFA coverage.
Traffic Engineering
Note
IS-IS-TE supports RFC 5305 (base TE), RFC 6119 (IPv6) and RFC 7810 / 8570 (Extended Metric) with or without Multi-Topology. All Traffic Engineering information are stored in a database formally named TED. However, best accuracy is provided without Multi-Topology due to inconsistency of Traffic Engineering Advertisement of 3rd party commercial routers when MT is enabled. At this time, FRR offers partial support for some of the routing protocol extensions that can be used with MPLS-TE. FRR does not currently support a complete RSVP-TE solution.
- mpls-te on
Enable Traffic Engineering LSP flooding.
- mpls-te router-address <A.B.C.D>
Configure stable IP address for MPLS-TE.
- mpls-te router-address ipv6 <X:X::X:X>
Configure stable IPv6 address for MPLS-TE.
- mpls-te export
Export Traffic Engineering DataBase to other daemons through the ZAPI Opaque Link State messages.
- show isis mpls-te interface
- show isis mpls-te interface INTERFACE
Show MPLS Traffic Engineering parameters for all or specified interface.
- show isis mpls-te router
Show Traffic Engineering router parameters.
- show isis [vrf <NAME|all>] mpls-te database [detail|json]
- show isis [vrf <NAME|all>] mpls-te database vertex [WORD] [detail|json]
- show isis [vrf <NAME|all>] mpls-te database edge [A.B.C.D|X:X::X:X] [detail|json]
- show isis [vrf <NAME|all>] mpls-te database subnet [A.B.C.D/M|X:X::X:X/M] [detail|json]
Show Traffic Engineering Database
See also
Segment Routing
This is an EXPERIMENTAL support of Segment Routing as per RFC8667 for MPLS dataplane. It supports IPv4, IPv6 and ECMP and has been tested against Cisco & Juniper routers.
- Known limitations:
No support for level redistribution (L1 to L2 or L2 to L1)
No support for binding SID
No support for SRMS
No support for SRLB
Only one SRGB and default SPF Algorithm is supported
- segment-routing on
Enable Segment Routing.
- segment-routing global-block (16-1048575) (16-1048575) [local-block (16-1048575) (16-1048575)]
Set the Segment Routing Global Block i.e. the label range used by MPLS to store label in the MPLS FIB for Prefix SID. Optionally sets also the Segment Routing Local Block. The negative command always unsets both.
- segment-routing node-msd (1-16)
Set the Maximum Stack Depth supported by the router. The value depend of the MPLS dataplane. E.g. for Linux kernel, since version 4.13 the maximum value is 32.
- segment-routing prefix <A.B.C.D/M|X:X::X:X/M> [algorithm (128-255)] <absolute (16-1048575)|index (0-1048575) [no-php-flag|explicit-null] [n-flag-clear]
prefix. The ‘no-php-flag’ means NO Penultimate Hop Popping that allows SR node to request to its neighbor to not pop the label. The ‘explicit-null’ flag allows SR node to request to its neighbor to send IP packet with the EXPLICIT-NULL label. The ‘n-flag-clear’ option can be used to explicitly clear the Node flag that is set by default for Prefix-SIDs associated to loopback addresses. This option is necessary to configure Anycast-SIDs.
- show isis segment-routing node [algorithm [(128-255)]]
Show detailed information about all learned Segment Routing Nodes.
Flex-Algos (Flex-Algo)
isisd supports some features of RFC 9350 on an MPLS Segment-Routing dataplane. The compatibility has been tested against Cisco.
IS-IS uses by default the Shortest-Path-First algorithm that basically calculates paths based on the shortest total metric to the destinations. Flex-Algo allows new algorithms to run in parallel to compute paths in different manners, based on metrics (IGP metric or a new type of metrics such as Traffic Engineering (TE) metric and minimum delay…) and constraints. New metric types are not yet implemented but constraints are already operational. Constraints can restrict paths to links with specific affinities or avoid links with specific affinities. Combinations of these are also possible.
The administrator can configure up to 128 Flex-Algos in an IS-IS area. To do so, it defines a set of Flex-Algo Definitions (FAD) which have the following characteristics:
a numeric identifier (ID) between 128 and 255 inclusive
- a set of constraints (basically, include or exclude a certain given set of
links, designated by a admin-group)
the calculation type (only the Shortest-Path-First is currently supported)
the metric type (only the IGP inherited metric type is currently supported)
some additional flags (not supported for the moment).
A subset of routers advertises the Flex-Algo Definitions (FAD) to the other routers within an area. In order to use a common set of FADs, each router runs a FAD election process for each locally configured algorithm and each level, among the FADs flooded in that level, using the following rules:
- If a locally configured FAD is not advertised to the area, the router does not
participate in the particular flex algorithm.
- If a given flex algorithm is running, the participation in this particular
flex algorithm stops when its advertisements are over.
- A router includes its own FAD in the election process if and only if it is
advertised to the other routers.
If only one router advertises the FAD, the FAD is elected.
- If several FADs are advertised with different priorities, the one with the
highest priority value is selected.
- If there are multiple advertisements of the FAD with the same highest
priority, the FAD of the router with the highest IS-IS system-ID is selected.
Routers only use the specifications of the elected FAD regardless of the locally configured definitions. If a router does not support one of the FAD characteristics, it stops participating in the Flex-Algo.
For each running Flex-Algo, the Segment-Routing SIDs must be configured with values unique to the algorithm. It allows routers to identify which flex algorithm they must use for a given packet.
The following commands configure Flex-Algo at the ‘router isis’ configuration level. Segment-Routing prefixes must be configured for the Flex-Algo.
- flex-algo (128-255)
Add a Flex-Algo Definition (FAD) and enter the FAD configuration level. The algorithm ID value is in the range of 128 to 255 inclusive.
- affinity-map NAME bit-position (0-255)
Add the specified ‘affinity-map’. Affinity-map definitions are used in FADs and in interfaces admin-group definition.
Affinity-maps format in advertisement TLVs use the extended admin-group format defined in the RFC7308 section 2.2. The extended admin-group uses a 256 bits field. If an affinity-map is set, the bit at the extended admin-group ‘bit-position’ is set 1, else it is set to 0.
The following commands configure Flex-Algo at the ‘router isis’ and ‘flex-algo (128-255)’ configuration level.
- advertise-definition
Advertise the current FAD to other IS-IS routers by using specific IS-IS TLVs. By default, the definition is is not shared with other routers.
A router can advertise a FAD without participating in the Flex-Algo.
- priority (0-255)
Set the specified ‘priority’ in the current FAD advertisements .
- metric-type [igp|te|delay]
Set the ‘metric-type’ for the current FAD. ‘igp’ is the default value and refers to the classic ‘Shortest-Path-First’ algorithm. If the ‘te’ or the ‘delay’ metric is selected, the value is advertised but the flex algorithm is disabled locally because these types are not currently supported.
- no metric-type
Reset the ‘metric-type’ to the default ‘igp’ metric.
- affinity exclude-any NAME
Add the specified affinity to the list of exclude-any affinities. The Flex-Algo will compute paths that exclude the segments with any of the specified affinities.
- no affinity exclude-any NAME
Remove the specified affinity to the list of exclude-any affinities.
- affinity include-all NAME
Add the specified affinity to the list of include-all affinities. The Flex-Algo will compute paths that include the segments with all the specified affinities.
- no affinity include-all NAME
Remove the specified affinity to the list of include-all affinities.
- affinity include-any NAME
Add the specified affinity to the list of include-any affinities. The Flex-Algo will compute paths that include the segments with any of the specified affinities.
- no affinity include-any NAME
Remove the specified affinity to the list of include-any affinities.
The following commands configure Flex-Algo at the ‘interface’ configuration level.
- isis affinity flex-algo NAME
Add the specified affinity to the interface.
- no isis affinity flex-algo NAME
Remove the specified affinity from the interface.
The following command show Flex-Algo information:
- show isis flex-algo [(128-255)]
Show information about the elected FADs
‘show isis route’, ‘show isis topology’ and ‘show isis segment-routing node’ includes an ‘algorithm (128-255)’ optional argument. See Showing ISIS information and Segment Routing.
Segment Routing over IPv6 (SRv6)
This feature enables extensions in IS-IS to support Segment Routing over IPv6 data plane (SRv6) as per RFC 9352.
- segment-routing srv6
Enable Segment Routing over IPv6 data plane (SRv6).
- locator NAME
Specify the SRv6 locator to use for SRv6. The locator must be configured in Zebra. Once the locator is configured, IS-IS automatically allocates prefix SID and adjacency SIDs, creates local SID entries in the data plane, and advertises them in the IGP domain.
- interface NAME
Specify the dummy interface used to install SRv6 SIDs in the Linux data plane. The interface must be created manually. By default, the interface is ‘sr0’. The interface can be created using the iproute2 utility:
ip link add sr0 type dummy ip link set sr0 up
- show isis segment-routing srv6 node
Show detailed information about all learned SRv6 Nodes.
Debugging ISIS
- debug isis adj-packets
IS-IS Adjacency related packets.
- debug isis events
IS-IS Events.
- debug isis packet-dump
IS-IS packet dump.
- debug isis route-events
IS-IS Route related events.
- debug isis snp-packets
IS-IS CSNP/PSNP packets.
- debug isis spf-events
IS-IS Shortest Path First Events.
- debug isis update-packets
Update related packets.
- debug isis te-events
IS-IS Traffic Engineering events
- debug isis sr-events
IS-IS Segment Routing events.
- debug isis lfa
IS-IS LFA events.
- debug isis flooding
- debug isis bfd
- debug isis ldp-sync
- debug isis tx-queue
- show debugging isis
Print which ISIS debug level is activate.
ISIS Configuration Examples
A simple example:
!
interface eth0
ip router isis FOO
isis network point-to-point
isis circuit-type level-2-only
!
router isis FOO
net 47.0023.0000.0000.0000.0000.0000.0000.1900.0004.00
metric-style wide
is-type level-2-only
A Traffic Engineering configuration, with Inter-ASv2 support.
First, the zebra.conf part:
hostname HOSTNAME
password PASSWORD
log file /var/log/zebra.log
!
interface eth0
ip address 10.2.2.2/24
link-params
max-bw 1.25e+07
max-rsv-bw 1.25e+06
unrsv-bw 0 1.25e+06
unrsv-bw 1 1.25e+06
unrsv-bw 2 1.25e+06
unrsv-bw 3 1.25e+06
unrsv-bw 4 1.25e+06
unrsv-bw 5 1.25e+06
unrsv-bw 6 1.25e+06
unrsv-bw 7 1.25e+06
admin-grp 0xab
!
interface eth1
ip address 10.1.1.1/24
link-params
enable
metric 100
max-bw 1.25e+07
max-rsv-bw 1.25e+06
unrsv-bw 0 1.25e+06
unrsv-bw 1 1.25e+06
unrsv-bw 2 1.25e+06
unrsv-bw 3 1.25e+06
unrsv-bw 4 1.25e+06
unrsv-bw 5 1.25e+06
unrsv-bw 6 1.25e+06
unrsv-bw 7 1.25e+06
neighbor 10.1.1.2 as 65000
Then the isisd.conf itself:
hostname HOSTNAME
password PASSWORD
log file /var/log/isisd.log
!
!
interface eth0
ip router isis FOO
!
interface eth1
ip router isis FOO
!
!
router isis FOO
isis net 47.0023.0000.0000.0000.0000.0000.0000.1900.0004.00
mpls-te on
mpls-te router-address 10.1.1.1
!
line vty
A Segment Routing configuration, with IPv4, IPv6, SRGB and MSD configuration.
hostname HOSTNAME
password PASSWORD
log file /var/log/isisd.log
!
!
interface eth0
ip router isis SR
isis network point-to-point
!
interface eth1
ip router isis SR
!
!
router isis SR
net 49.0000.0000.0000.0001.00
is-type level-1
topology ipv6-unicast
lsp-gen-interval 2
segment-routing on
segment-routing node-msd 8
segment-routing prefix 10.1.1.1/32 index 100 explicit-null
segment-routing prefix 2001:db8:1000::1/128 index 101 explicit-null
!
An SRv6 configuration:
hostname HOSTNAME
password PASSWORD
log file /var/log/isisd.log
!
!
interface eth0
ipv6 router isis FOO
ip router isis FOO
isis hello-interval 5
!
interface eth1
ip router isis FOO
!
!
router isis FOO
net 49.0001.1111.1111.1111.00
is-type level-2-only
metric-style wide
segment-routing srv6
locator loc1
!
line vty
ISIS Vrf Configuration Examples
A simple vrf example:
!
interface eth0 vrf RED
ip router isis FOO
isis network point-to-point
isis circuit-type level-2-only
!
router isis FOO vrf RED
net 47.0023.0000.0000.0000.0000.0000.0000.1900.0004.00
metric-style wide
is-type level-2-only