From: Art -kwaak- van Breemen Subject: Re: Regression in pmtud lvs-nat with ipv6 since at least 3.10 Date: Tue, 18 Feb 2014 11:05:33 +0100 Message-ID: <20140218100533.GE18029@telegraafnet.nl> References: <20140217153206.GC18029@telegraafnet.nl> <20140217161601.GD18029@telegraafnet.nl> Mime-Version: 1.0 Return-path: Content-Disposition: inline In-Reply-To: Sender: lvs-devel-owner@vger.kernel.org List-ID: Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Julian Anastasov Cc: lvs-devel@vger.kernel.org Hi Julian, On Mon, Feb 17, 2014 at 11:34:43PM +0200, Julian Anastasov wrote: > May be ipv6_find_hdr() can not be called with > exact protocol number, the offset is not updated, > it remains 0 and -ENOENT is returned. As result, > icmp_offset is not changed to 40. I was thinking along that line too, since sizeof(struct icmp6hdr) would exactly skip that part, hence pointing ciph at the second half of the first ip address. I try to make a test setup, since the real lvs servers are actually production ;-). But I can live with a little downtime on my desktop 8-D. Regards, Ard