SD-WAN OMP Graceful Restart timer

[RS19]

In my environment the OMP Graceful Restart timer is set to 7 days.

If the vSmart is down, with the above the cEdge will have the routes till 7 days.

After 7 days the routes will flush and the dataplane will be down.

My question is assuming the vSmart is down 2 days before, is there any command in cEdge which says that the routes will be valid for only 5 days ? Is there any way from CLI I can find till how many days or hours the route is valid ?

[RS19]

any quick help will be useful

[Torbjørn]

You need to check which peer you received the route from with "show omp route [detail]", check the downtime of said peer with "show omp peer (ip) detail" and check the graceful-restart time in configuration with "show run | i graceful-restart-timer".

I don't believe there is a single command that will give you the remaining lifetime of the route.

[RS19]

I have only 1 vSMart. So in this case if the vSmart is down, I have no way to check in cEge the validity of the route.

I need to manually calculate by what time the vSmart is down & to need to calculate manually the validity of the route. Is my  understanding right ?

[Torbjørn]

You should be able to execute the commands in my previous reply on your cEdge to figure out when your route will expire. AFAIK there is no command in IOS-XE that will tell you explicitly the remaining validity of your OMP route.

[totie]

FYI `show sdwan omp peers detail` exist on my CAT8000v IOS-XE 17.11.1a.  Caveat when you issue `show sdwan omp peers ?` the question mark result does not show that `detail` command is non existence but believe me try it anyways and it is there.  Any show commands you do with vEdge you can try on cEdge with the equivalent `show sdwan`.

[RS19]

if my understanding is right, I need to execute the below command right ?

 "show omp route [detail]",

[Torbjørn]

Yes, then find the advertising vsmart, listed as "peer" in the output. Find out how long the specific vsmart has been down with "show omp peer [ip] detail" and compare this with the graceful restart timer which is 7 days by default, if it is not default you can find the configured value with "show run | i graceful-restart-timer".

[RS19]

In othe eniromentI have 2 vsmart. Attahed is the screen shot for reference. 

From the output , routes are installed from v-Smart1 & from the other v-Smart2, there are no routes installed.
Does it means, v-Smart1 is active & the other v-smart2 is standby/passive.

Also let me know what happens when v-Smart1 goes down. 

[RS19]

Due to some reasons, not able to attached the screen shot.

But below are the details

vSmart 1      State: UP  R/I/S 12294/4952/86

vSmart 2      State: UP  R/I/S 7590/0/86
Want to understand why no routes are installeed from vSmart 2. Does it means it is standby/passive ?

Also let me know what happens when vSmart 1 goes down.

[Kanan Huseynli]

Hi,

you have two routes received from both vSmarts. If both send identical information (which normally happens) route with lower peer systemIP is installed. That's why vSmart1's routes are chosen.

But when you lose vSmart1 and its routes become "stale", then different logic is used. You have stale routes from vSmart1 and active routes from vSmart2. Stale VS active , then active routes are installed. In short, when you lose vSmart1, routes quickly reprograms routing to use routes from vSmart2.

https://www.cisco.com/c/en/us/support/docs/routers/sd-wan/217479-omp-best-path-selection-peculiarities-an.html

https://www.cisco.com/c/en/us/td/docs/routers/sdwan/configuration/routing/ios-xe-17/routing-book-xe/m-unicast-routing.html#best-path