need help w/ 2 concepts

sinexec

1) I've checked the values for hold down and invalid route timers by show ip proto on my cisco 2500 router which indicated that both of those are the same and equal to 180s. This sort of does not make sense to me because according to the definition of hold down timer it should prevent any changes in routing table regarding some of the routes for a period of time which happens to be 180 sec in my case... It looks like I would always take 180 sec to make changes on that route after invalid route timer expires and flash update has been sent... ?

2) What's the difference between Rxboot and ROM monitor modes?

Thanks.

sikdogg

1) yes, the default hold-down and invalid-route timers for RIP is 180 seconds. If you think about it, this would make sense. The purpose of the hold-down timer is to counter the count-to-infinity problem associated with multiple routes to a given network. When a router finds that a route to a network fails, it waits 180 seconds before it will accept any routing updates to the failed network.

The invalid-route timer counts x number of seconds (I believe it is 6 times the update period) and if it does not receive a route update within that time period for a given network, it will assume that the route to said network is invalid and removes it from its routing table.

Let's see how they work together. The default values for both the hold-down and invalid-route timers work well together. When a route to Network-A fails on a router, it waits for the hold-down period to expire before accepting any route updates to that network. At the same time, the invalid-route timer also starts ticking and when the 180 seconds expires, the invalid-route timer assumes that its route to Network-A is invalid and removes the entry from its routing table. At the same instance that the router removes the route to Network-A from its table, it begins to accept route updates to Network-A again. This ensures that route in its routing table is always fresh and updated.

2)Rxboot is a boot image, a mini IOS w/ limited capabilities, that is used during the boot sequence. Lower end routers (1600s & 2500s) have it in ROM, other routers (usually higher end) have it in bootflash.

ROM monitor is the mode a router goes into when it can't find a valid IOS to run and defaults to the rxboot or boot image. Some router, like the 2500, will show a "rommon>" prompt when in ROM monitor mode. Other routers will show a "rxboot>" prompt. It just depends on the platform but in either case, it signifies that the router is in ROM monitor mode.

sinexec

Thanks for a quick replay. My test it tomorrow...

sikdogg wrote:

Let's see how they work together. The default values for both the hold-down and invalid-route timers work well together. When a route to Network-A fails on a router, it waits for the hold-down period to expire before accepting any route updates to that network. At the same time, the invalid-route timer also starts ticking and when the 180 seconds expires, the invalid-route timer assumes that its route to Network-A is invalid and removes the entry from its routing table. At the same instance that the router removes the route to Network-A from its table, it begins to accept route updates to Network-A again. This ensures that route in its routing table is always fresh and updated.

I do think you might be wrong on this though.
I may be wrong but I appears to me that the invalid route timer should expire first (accoring to the books it is usually set to 3 update period time and do not quite understand why it is different for RIP) then the flash update should be sent out and thats when hold down timer should come in to prevent other routers that are not aware of the changes yet to send out their routing tables with the subnet that is supposed to be removed from all of the router on a network.
The route is only flushed after 270 sec (in case of RIP) and not right after the invalid route timer expires. It is kept in a routing table for additional 190 seconds after it becomes invalid it is just that the router does not advertise it to other routers...

sikdogg

My appologies, I just got home and did some research after reading your comment and found that I was incorrect on a few items.

According to Doyle's text (Routing TCP/IP, Vol. 1), both the hold-down and invalid timers are set for 3 update periods or 180 seconds (I don't know where I got 6 times). This is true for Cisco's implementation of RIP and for RFC 1058 (Cisco's RIP and RFC 1058 differ slightly). The invalid timer does not remove routes from the table as I had incorrectly stated earlier, but instead sets the hop count to 16 (unreachable). You were correct in saying that the routes are removed from the table only after the flush timer has expired, 60 seconds after the invalid timer (240 seconds total, RFC 1058 specifies 120 seconds after the invalid timer (300 seconds total)). The route will be advertised as unreachable until the flush timer expires, at which time the route will be removed.

Again I appologize for not doing sufficient research and replying too hastily, it has been some time since i've worked with RIP. I hope that I haven't mislead you or anyone else for that matter. Good luck on your test, you seem to have a good grasp of the material and should do very well.