Date: Tue, 27 Mar 2007 15:52:59 +0100 From: "Ross Draper" <Ross.Draper@gcapmedia.com> To: "Jordan Gordeev" <jgordeev@dir.bg> Cc: freebsd-net@freebsd.org Subject: RE: Vrrp/CARP/ucarp Problems Message-ID: <3DDDCC38D00FA545A6C012475EF2DC0302AF8FD5@LQEVS1.gcapmedia.com>
next in thread | raw e-mail | index | archive | help
Ah... I see your point, thankyou for clarifying. =20
=20
Sorry, this is down to me only explaining half the story. The actual
loadbalancing will be performed by HAProxy, I require carp or freevrrp
to enable "a clustered pair" of loadbalancers.=20
=20
I did toy with the idea arp balancing to enable greater throughput in an
active/active scenario, but as you say mention this is not suitable for
certain deployments.
=20
Kind Regards
=20
Ross
________________________________
From: Jordan Gordeev [mailto:jgordeev@dir.bg]=20
Sent: Tuesday, March 27, 2007 3:47 PM
To: Ross Draper
Cc: freebsd-net@freebsd.org
Subject: Re: Vrrp/CARP/ucarp Problems
The only load balancing that CARP supports, to my knowledge, is ARP
level load balancing. From carp(4):
The ARP load balancing has some limitations. First, ARP balancing only
works on the local network segment. It cannot balance traffic that
crosses a router, because the router itself will always be balanced
to
the same virtual host.
Ross Draper wrote:=20
Hi
=09
Firstly, many thanks to Stefan (who provided me a diff of the
CURRENT update) and Bruce for advising me that the "multiple CARP
interface destroy" bug is fixed in CURRENT.
=09
Jordan, thanks for your reply, but I cant find a reference to
CARP being unsuitable for this purpose in the CARP man page, have I
perhaps misunderstood you? I also did a quick search of the freebsd.org
site and cant find any mention of it being an issue - If you could
provide me with a link I'd be grateful.
=09
Further to this, I have been in the office today performing
local testing as opposed to remote testing and have noticed that when
both machines in the cluster are using xl network cards, failover etc
seems fine. However, when having one node using xl and the other using
em/bge I can see the em or bge card physically go down after it reverts
to backup mode, then come back up and goto master. (this wasnt
displaying in the messages log, but was obvious on the console). I
believe that this down period is sufficient for it to miss the remaining
advertisements and believe it is the master again. For some reason it
doesnt seem to ever remove the mac address or recover after this point,
but I'm not particularly suprised. Not sure how to take this further,
but I'll continue fiddling.
=09
Many thanks
=09
Ross
=09
=09
=09
All correspondence, attachments and agreements remain strictly
subject to fully executed contract.
(c) GCap Media plc 2006. All rights remain reserved.
This e-mail (and any attachments) contains information which may
be confidential, subject to intellectual property protection and may be
legally privileged and protected from disclosure and unauthorised use.
It is intended solely for the use of the individual(s) or entity to whom
it is addressed and others specifically authorised to receive it. If you
are not the intended recipient of this e-mail or any parts of it please
telephone 020 7054 8000 immediately upon receipt.
No other person is authorised to copy, adapt, forward, disclose,
distribute or retain this e-mail in any form without prior specific
permission in writing from an authorised representative of GCap Media
plc.
We will not accept liability for any claims arising as a result
of the use of the internet to transmit information by or to GCap Media
plc.
GCap Media plc. Registered address: 30 Leicester Square, London
WC2H 7LA. Registered in England & Wales with No. 923454
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3DDDCC38D00FA545A6C012475EF2DC0302AF8FD5>