Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 10 Feb 2021 15:10:13 +0300
From:      "Andrey V. Elsukov" <bu7cher@yandex.ru>
To:        Dewayne Geraghty <dewayne@heuristicsystems.com.au>, freebsd-stable@freebsd.org
Subject:   Re: 13.0-BETA1: ipfw regression?
Message-ID:  <9e3643cd-7696-f612-9b58-94e08ccc92ef@yandex.ru>
In-Reply-To: <abb69160-9cd4-297b-ef97-e5cc9115f179@heuristicsystems.com.au>
References:  <202102100646.11A6kQGS068916@nuc.oldach.net> <abb69160-9cd4-297b-ef97-e5cc9115f179@heuristicsystems.com.au>

next in thread | previous in thread | raw e-mail | index | archive | help
On 10.02.2021 10:40, Dewayne Geraghty wrote:
> Stefan, Would you check that you have net.inet.tcp.always_keepalive=1,
> and perhaps that net.inet.tcp.keepidle and net.inet.tcp.keepintvl are
> reasonable to ensure that the expected keep alives are running.  I don't
> have a FreeBSD 13 to view the defaults, but Helge might be right that
> "it" is already fixed :)
> Regards, Dewayne.

ipfw does send its own keep-alive packets for TCP connections when
state's lifetime expires.
But if remote host doesn't reply this doesn't help.
To check what happened Stefan needs to capture packet dump and inspect it.

-- 
WBR, Andrey V. Elsukov



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?9e3643cd-7696-f612-9b58-94e08ccc92ef>