Does route table mix routes of different protocol?

johnifanx98johnifanx98 Posts: 329Member
Say, a router has RIP and EIGRP run together, will I see mixed routes like "R" and "D" running "show ip route"?

Comments

  • IristheangelIristheangel ABL - Always Be Labbin' Pasadena, CAPosts: 4,112Mod Mod
    Yes. You'll also see connected and static routes
    BS, MS, and CCIE #50931
    Blog: www.network-node.com
    Bonus TE Fun: Nerd Photos
  • johnifanx98johnifanx98 Posts: 329Member
    Ok. Assume you understood my question, but just put wrong wording. My actual puzzle is: when an EIGRP sends update to its neighbor, it will send BW and delay of each EIGRP route. However, how does it determine BW/delay for those non-EIGRP route, like RIP?
  • IristheangelIristheangel ABL - Always Be Labbin' Pasadena, CAPosts: 4,112Mod Mod
    Perhaps I'm not understanding your question. EIGRP uses bandwidth and delay as part of it's calculation to determine it's metric, but other protocols don't use the same set of criteria to determine the metric for the routes. With RIP, it's a distance vector protocol so instead of using BW and delay, it uses hop count. If you're talking about redistributing a RIP route into EIGRP (which I don't believe is on the CCNA exam so I don't think you are asking this), you actually configure the metric during the redistribute command or configure a default metric. Again, route redistribution is outside of the scope of the CCNA exam so not sure if you're asking that or not
    BS, MS, and CCIE #50931
    Blog: www.network-node.com
    Bonus TE Fun: Nerd Photos
  • johnifanx98johnifanx98 Posts: 329Member
    Thanks a lot for your clarification. I was asking about redistribution. I doubt it's in icnd2, but just not certain.
  • Mrock4Mrock4 Posts: 2,360Banned
    It's important to realize that the EIGRP metric (bandwidth/delay by default as Iris said) is not sent with updates. It is a local value, known as the "K values" which is configured using the command "metric weights" command under router config mode. A router receives the metric from it's neighbor (which was derived from the neighbors bandwidth/delay).
  • YFZbluYFZblu Posts: 1,462Member ■■■■■■■■□□
    Thanks a lot for your clarification. I was asking about redistribution. I doubt it's in icnd2, but just not certain.
    I don't remember Odom covering it, but I believe Lammle does.
  • theodoxatheodoxa Posts: 1,340Member
    1) Administrative Distance determines which routing protocol's route gets put in the table. If you receive IDENTICAL prefixes from EIGRP and RIP, the EIGRP route will get put in the routing table and not the RIP route because EIGRP has a default AD of 90 vs. RIP's 120.

    2) If you are asking about redistribution...when you redistribute a route, you can specify the metric to be used or component values used to calculate the metric depending on the routing protocol.

    Redistribution is covered in [CCNP] ROUTE.
    R&S: CCENT CCNA CCNP CCIE [ ]
    Security: CCNA [ ]
    Virtualization: VCA-DCV [ ]
  • instant000instant000 Posts: 1,745Member
    johnifanx98:

    Keep in mind that the routing table that you see from
    "show ip route" only shows the BEST routes for each destination shown.

    Each routing protocol makes its own table of valid routes.
    That is, there is a RIP-specific table. "show ip rip database"
    There is also an EIGRP-specific table: "show ip eigrp topology"

    The FIB that you see from "show ip route" includes the collection of the valid longest-prefix, lowest-AD routes. that come from all the concerned input tables, and includes connected and static routes.

    this is how this table is populated:
    longest-prefix is preferred (most specific match to the route)
    if that ties, then obviously, you want lowest AD (most trustworthy routing protocol for the route)
    if that ties, then you want the lowest metric (most efficient route)
    if you are still tied, load-balance (within routing protocol parameters, some allow you to fudge this, within constraints)

    I am not sure if this was part of your question or not, but I hope it helps.
    Currently Working: CCIE R&S
    LinkedIn: http://www.linkedin.com/in/lewislampkin (Please connect: Just say you're from TechExams.Net!)
  • johnifanx98johnifanx98 Posts: 329Member
    I'm looking at IP internal route TLV. Both BW and delay are included.

    EIGRP Packet Format | CCIE Certification Exam Lab,CCIE Security,Routing And Switching(R&S),Service Provider(SP),Voice Lab Workbook
    Mrock4 wrote: »
    It's important to realize that the EIGRP metric (bandwidth/delay by default as Iris said) is not sent with updates. It is a local value, known as the "K values" which is configured using the command "metric weights" command under router config mode. A router receives the metric from it's neighbor (which was derived from the neighbors bandwidth/delay).
  • johnifanx98johnifanx98 Posts: 329Member
    Instant, this does help really. I guess it also falls under the subject of redistribution.
    instant000 wrote: »
    johnifanx98:

    Keep in mind that the routing table that you see from
    "show ip route" only shows the BEST routes for each destination shown.

    Each routing protocol makes its own table of valid routes.
    That is, there is a RIP-specific table. "show ip rip database"
    There is also an EIGRP-specific table: "show ip eigrp topology"

    The FIB that you see from "show ip route" includes the collection of the valid longest-prefix, lowest-AD routes. that come from all the concerned input tables, and includes connected and static routes.

    this is how this table is populated:
    longest-prefix is preferred (most specific match to the route)
    if that ties, then obviously, you want lowest AD (most trustworthy routing protocol for the route)
    if that ties, then you want the lowest metric (most efficient route)
    if you are still tied, load-balance (within routing protocol parameters, some allow you to fudge this, within constraints)

    I am not sure if this was part of your question or not, but I hope it helps.
Sign In or Register to comment.