Date: Thu, 19 Jun 2003 16:49:50 +0300 From: Vlad GALU <vladg@vipnet.ro> To: freebsd-net@freebsd.org Subject: Re: Redundant link configuration Message-ID: <20030619164950.49c942c0.vladg@vipnet.ro> In-Reply-To: <20030619164024.L698@mcd01p59.mrc.alcatel.ro> References: <FE045D4D9F7AED4CBFF1B3B813C8533701AE842A@mail.sandvine.com> <20030619164024.L698@mcd01p59.mrc.alcatel.ro>
next in thread | previous in thread | raw e-mail | index | archive | help
[-- Attachment #1 --]
On Thu, 19 Jun 2003 16:50:17 +0300 (EEST)
Ciprian Badescu <ciprian.badescu@alcatel.ro> wrote:
> Hi,
>
> I'm not very old in networking, and I want to migrate to following
> configuration:
>
>
>
> Local router Remote router
> +---------------+ +---------------+
> | | | |
> | |192.168.0.1 link 1 192.168.0.11 | |
> | |---------------------------------------| |
> | |---------------------------------------| |
> | |192.268.0.2 link 2 192.168.0.12 | |
> +---------------+ +---------------+
>
I tried channel bonding and it worked pretty good. I haven't
yet tested it extensively. Give it a try though. As for the system
interface, all I know about this is that you should use mii-based
NIC's, because they know how to report downlinks. I'm not aware
of any failover software at this very moment, but this can be my
poor memory. I hope someone else can answer to this more properly.
>
> The link 192.168.0.1 <->192.168.0.11 Is the actual one, using RF, and
> I want to move all traffic to the new OF link 192.268.0.2
> <->192.168.0.12.
>
> I don't know how to configura the routers to use the second link, and
> to keep the first link for redundancy, and if something is happening
> with link2, tha traffic must use automatically link1.
>
> It's enough tu use routing metrics, or I must use a routing protocol
> (and if so, which one)?
>
>
>
>
>
> --
> Ciprian Badescu
>
> On Thu, 19 Jun 2003, Scot Loach wrote:
>
> > Date: Thu, 19 Jun 2003 08:20:34 -0400
> > From: Scot Loach <sloach@sandvine.com>
> > To: "'freebsd-net@freebsd.org'" <freebsd-net@freebsd.org>
> > Subject: socket leak in FreeBSD 4.7
> >
> > If I execute the following program on a FreeBSD 4.7 system:
> >
> > int main()
> > {
> > for (int i = 0; i < 70000; i++)
> > {
> > socket(PF_INET, SOCK_STREAM, 0);
> > }
> > return 0;
> > }
> > the sockets never seem to be freed.
> >
> > vmstat tells me the memory is still in use:
> >
> > bash-2.05a$ vmstat -z
> >
> > ITEM SIZE LIMIT USED FREE REQUESTS
> >
> > tcpcb: 544, 106000, 65557, 4474, 70029
> > udpcb: 192, 106000, 7, 35, 481
> > unpcb: 160, 0, 7, 43, 78
> > socket: 192, 106000, 65571, 4487, 70639
> >
> > However, fstat and sockstat do not show a large number of sockets.
> >
> > netstat shows them, and they look like this:
> >
> > tcp4 0 0 *.* *.*
> > CLOSED
> >
> >
> > If I explicitly close the sockets before the program exits, this
> > does not occur. However I'm worried about the case where a
> > high-volume server exits unexpectedly and leaks sockets.
> >
> > Any ideas of what might be wrong?
> >
> >
> > _______________________________________________
> > freebsd-net@freebsd.org mailing list
> > http://lists.freebsd.org/mailman/listinfo/freebsd-net
> > To unsubscribe, send any mail to
> > "freebsd-net-unsubscribe@freebsd.org"
> >
> _______________________________________________
> freebsd-net@freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-net
> To unsubscribe, send any mail to "freebsd-net-unsubscribe@freebsd.org"
>
--
Vlad GALU
Network Administrator VipNET Bucharest
tel: 021/3039940
email: vladg@vipnet.ro
web: http://www.vipnet.ro
PGP: http://mirapoint.vipnet.ro/public_key.pgp
[-- Attachment #2 --]
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (FreeBSD)
iD8DBQE+8b+CBQlxy6GegvARAo/1AKDifIFKRaHE20g2fdg+iDRJ9/KrIQCgyr6k
EIpww7tfN+wppxn42nMLa1I=
=qioA
-----END PGP SIGNATURE-----
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030619164950.49c942c0.vladg>