Date: Wed, 16 May 2012 01:51:59 +0000 From: "Li, Qing" <qing.li@bluecoat.com> To: David DeSimone <fox@verio.net>, "freebsd-net@freebsd.org" <freebsd-net@freebsd.org> Cc: "qingli@freebsd.org" <qingli@freebsd.org> Subject: RE: [stable-9] Message-ID: <B143A8975061C446AD5E29742C531723CA3D27@pwsvl-excmbx-05.internal.cacheflow.com> In-Reply-To: <20120515231432.GH28776@verio.net> References: <1337105271.3434.8.camel@powernoodle-l7.corp.yahoo.com> <B143A8975061C446AD5E29742C531723CA33D1@pwsvl-excmbx-05.internal.cacheflow.com>, <20120515231432.GH28776@verio.net>
next in thread | previous in thread | raw e-mail | index | archive | help
Yes, you are right, I should have paid closer attention to the output, howe= ver, what=0A= I replied still apply to RADIX_MPATH in general.=0A= =0A= Now, when I first introduced RADIX_MPATH, to get the (not always even) gene= ral load =0A= balancing among interfaces attached to the same prefix, you only need to ad= d the following=0A= additional host routes, as in:=0A= =0A= route add -host gateway-ip -interface igb0=0A= route add -host gateway-ip -interface igb1=0A= =0A= However, I just realized that strategy no longer work after the ARP entries= are maintained =0A= separately from the general routing table.=0A= =0A= Let me think it through and try to come up with a patch.=0A= =0A= --Qing=0A= =0A= ________________________________________=0A= From: owner-freebsd-net@freebsd.org [owner-freebsd-net@freebsd.org] on beha= lf of David DeSimone [fox@verio.net]=0A= Sent: Tuesday, May 15, 2012 4:14 PM=0A= To: freebsd-net@freebsd.org=0A= Subject: Re: [stable-9]=0A= =0A= Li, Qing <qing.li@bluecoat.com> wrote:=0A= >=0A= > The route selection is based on a hash function of source-ip and=0A= > destination-ip when RADIX_MPATH is enabled. You do not need to perform=0A= > specific actions, other than perhaps setting varying weights on each=0A= > entry as an option. So depends on the traffic destination the chosen=0A= > route may always be the same one.=0A= =0A= These comments suggest that there are multiple routes to be selected=0A= from. However, this output:=0A= =0A= > Internet:=0A= > Destination Gateway Flags Refs Use Netif=0A= > Expire=0A= > default <redacted> UGS 0 33320033 igb0=0A= > XXX.XXX.XXX.0/25 link#1 U 0 0 igb0 =3D>= =0A= > XXX.XXX.XXX.0/25 link#2 U 0 0 igb1=0A= =0A= suggests that there is only ONE default route, pointing to ONE=0A= interface, igb0. Without an extra default route that is also pointing=0A= to igb1, I can't see how the system woudl ever forward traffic out igb1,=0A= unless it was directed to the local /25.=0A= =0A= --=0A= David DeSimone =3D=3D Network Admin =3D=3D fox@verio.net=0A= "I don't like spinach, and I'm glad I don't, because if I=0A= liked it I'd eat it, and I just hate it." -- Clarence Darrow=0A= =0A= =0A= This email message is intended for the use of the person to whom it has bee= n sent, and may contain information that is confidential or legally protect= ed. If you are not the intended recipient or have received this message in = error, you are not authorized to copy, distribute, or otherwise use this me= ssage or its attachments. Please notify the sender immediately by return e-= mail and permanently delete this message and any attachments. Verio, Inc. m= akes no warranty that this email is error or virus free. Thank you.=0A= _______________________________________________=0A= freebsd-net@freebsd.org mailing list=0A= http://lists.freebsd.org/mailman/listinfo/freebsd-net=0A= To unsubscribe, send any mail to "freebsd-net-unsubscribe@freebsd.org"=0A=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?B143A8975061C446AD5E29742C531723CA3D27>