-
Notifications
You must be signed in to change notification settings - Fork 1.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Graceful Restart procedures do not apply when hard-administrative-reset flag is disabled #18238
Comments
Could you show a full configuration? Update: I see the config in your repo. Will test it. |
I'll test it as @dawkopagh is out for a week and I'll let you know |
When the session comes up the router resends the BGP table, hence you see withdraws + updates, that's kinda expected. I don't really understand where is an issue yet. |
The issue is that based on the docs: "When N-bit is set and bgp hard-administrative-reset is turned off Graceful-Restart is activated and routes are retained." This is the case for this setup, however there is no graceful restart - it just does a typical restart, withdrawing the routes and then adding them. Expected behavior based on the docs description would be:
And this is what was happening in FRR8. |
What happens with FRR 10.x (or master)? Because I checked with the master, and it works fine (stale routes are retained when
|
This is me checking it on frr master:
And checking the rtmon logs:
In my opinion, because the routes have a new timer and there was a moment when they weren't there and they appear as deleted and readded in the logs, that it means that the graceful restart is not properly applied, even though the |
Could you show the "show bgp ipv4 detail" after clear? And please show a full log before/after the clear (with "debug bgp updates", "debug bgp neighbor"). |
1-st scenario logs (for 10.1.1, exactly the same thing happening for master)With logs:
Output of
|
So.. is this happening only for evpn imported routes? |
The same happens for ipv4 unicast (I assume you are asking because evpn imported routes are expected to be withdrawn due to this snippet?) Tested on master:
show bgp ipv4 detail (before clear):
logs:
kernel withdrawals:
negotiated graceful-restart settings between peers:
|
Description
When sending CEASE/Administrative Reset (clear bgp) with N-bit is set and bgp hard-administrative-reset disabled routes are being withdrawn from kernel. This should not be the case per this description
Version
How to reproduce
I have provided an instruction in my public repo here, just follow README.md
Expected behavior
Graceful Restart procedures should apply and routes should be retained in kernel
Actual behavior
Routes are being withdrawn from kernel when the session goes down and re-added after session establishes again.
Additional context
No response
Checklist
The text was updated successfully, but these errors were encountered: