On Mon, 30 Sep 2019 at 02:29, Stuart Henderson <stu@spacehopper.org> wrote:
>
> On 2019/09/29 16:26, Ville Valkonen wrote:
> > Week+ ping? Cc'ing Stuart as he's the maintainer of nfdump port.
>
> Yes you should definitely cc the maintainer - but no diff,
> and the inline parts in the quoted mail are not usable directly (missing
> headers, hardcoded paths that should come from variables)..
Hello Stuart,
please see the original email below.
I came to realize that custom flags will omit the default flags. Not
sure if that's the problem but we could use the similar approach that
unbound has where user flags are concatenated to the default flags.
<snip>
Hello,
please CC me as I'm not subscribed in to the ports mailing list.
Here's [1] an rc script for nfcapd daemon which is part of the nfdump package.
And while here, document nfexpire in some level in README [2].
Comments?
[1]
<snip>
#!/bin/ksh
daemon="/usr/local/bin/nfcapd -D"
daemon_user="_nfcapd"
daemon_flags="-l /var/cache/nfcapd -P /var/run/nfcapd/nfcapd.pid"
rc_reload=NO
. /etc/rc.d/rc.subr
rc_pre() {
/usr/bin/install -d -o ${daemon_user} /var/run/nfcapd
}
rc_cmd $1
</snip>
[2]
It should be noted that nfcapd can fill disk rather quickly in vibrant network
environments. Therefore it's important to flush the flow stats every now and
then. One could use the following root's crontab entry to expire flows older
than 31 days:
@daily /usr/local/bin/nfexpire -t 31d /var/cache/nfcapd >/dev/null 2>&1
--
Kind regards,
Ville Valkonen
</snip>
Regards,
Ville
No comments:
Post a Comment