MX BGP session no longer established

[ARocha]

Hi all, 

I''m having a BGP issue for a couple of sites. Every morning we have to reboot the MX devices and checking the logs I can see the BPG hold timer expired.

Unsure how to check the cause of this issue or more information about it.

Any suggestions? Thanks

Jul 19 07:49:30 BGP BGP session no longer established peer_ip: x.x.x.x, local_as: xxxx, remote_as: xxxx, new_state: Close

Jul 19 07:49:30 BGP BGP received notification peer_ip: x.x.x.x , local_as: xxxx, remote_as: xxxx  
error_code 4 
error_subcode 0 
desc Hold timer expired

[Brash]

Take a look at the hold timers for each device, along with the documentation on how the timers are negotiated.

Meraki information can be found here:
https://documentation.meraki.com/MX/Networks_and_Routing/Border_Gateway_Protocol_(BGP)#Default_Updat...

A few things to think about:

 - Is it happening regularly at a specific/predictable time?

 - Is it only specific sites or random?

Beyond that, I'd suggest engaging support from both vendors to review additional logging.

[ARocha]

Thanks for your response, Brash.

Yes, it is happening every morning. Also, the same template is applied for more than 100 sites and just a couple of sites are experiencing the same issue.

I'm a bit resistant to changing the timers, causing issues for those sites that are working fine.

[PhilipDAth]

Has it always happened or is the issue more recent?  If more recent, what changed around that date?

[ARocha]

The site started experiencing issues when we migrated to the MX device. Since the MX was deployed in some sites, it started happening.

I also noticed a lot of events dropped

Jul 21 17:43:19 Events dropped Events dropped 841 events were not logged.
Jul 21 17:40:09 Events dropped Events dropped 826 events were not logged.
Jul 21 17:36:59 Events dropped Events dropped 854 events were not logged.
Jul 21 17:33:49 Events dropped Events dropped 773 events were not logged.
Jul 21 17:30:49 Events dropped Events dropped 827 events were not logged.
Jul 21 17:27:39 Events dropped Events dropped 854 events were not logged.
Jul 21 17:24:29 Events dropped Events dropped 775 events were not logged.
Jul 21 17:21:29 Events dropped Events dropped 827 events were not logged.
Jul 21 17:18:19 Events dropped Events dropped 840 events were not logged.
Jul 21 17:15:09 Events dropped Events dropped 826 events were not logged.
Jul 21 17:11:59 Events dropped Events dropped 840 events were not logged.
Jul 21 17:08:49 Events dropped Events dropped 822 events were not logged.
Jul 21 17:05:39 Events dropped Events dropped 839 events were not logged.
Jul 21 17:02:29 Events dropped Events dropped 826 events were not logged

[GIdenJoe]

It looks like the remote end is killing your session because the MX has stopped sending keepalives.  Or perhaps these keepalives are no longer making it across.  If it is a direct link then I would assume the MX stops sending them due to some other issue.
I believe a support case is in order here.