Index | Thread | Search

From:
Stuart Henderson <stu@spacehopper.org>
Subject:
Re: [patch] wireguard floods dmesg
To:
Lloyd <ng2d68@proton.me>
Cc:
"tech@openbsd.org" <tech@openbsd.org>
Date:
Wed, 11 Dec 2024 12:11:04 +0000

Download raw body.

Thread
  • Stuart Henderson:

    [patch] wireguard floods dmesg

  • On 2024/12/11 07:45, Lloyd wrote:
    > Hello,
    > 
    > Wireguard has a rudimentary logging function accessed by setting the debug flag on the interface. Logging Wireguard activity is desirable e.g. for SIEM purposes. However, the current implementation will flood the system message buffer with the following message if the other side of the tunnel is down:
    > 
    > wg0: No valid endpoint has been configured or discovered for peer 0
    > wg0: No valid endpoint has been configured or discovered for peer 0
    > wg0: No valid endpoint has been configured or discovered for peer 0
    > 
    > (etc... many times per minute)
    
    How many is "many"? If it's really very bad then ratecheck(9) can help.
    
    > The below patch quiets the chatter by mapping the unused link0 flag as a "more verbose debugging" setting. Only the above message is muted, but scoping more in similar fashion may be warranted.
    
    I don't like using link0 for that.
    
    
    
  • Stuart Henderson:

    [patch] wireguard floods dmesg