Re: [PATCH v1 net] tcp: Refine SYN handling for PAWS.

[Eric Dumazet <edumazet@google.com>]

On Mon, Mar 27, 2023 at 4:06 PM Kuniyuki Iwashima <kuniyu@amazon.com> wrote:
>
> Our Network Load Balancer (NLB) [0] has multiple nodes with different
> IP addresses, and each node forwards TCP flows from clients to backend
> targets.  NLB has an option to preserve the client's source IP address
> and port when routing packets to backend targets.
>
> When a client connects to two different NLB nodes, they may select the
> same backend target.  Then, if the client has used the same source IP
> and port, the two flows at the backend side will have the same 4-tuple.
>
> While testing around such cases, I saw these sequences on the backend
> target.
>
> IP 10.0.0.215.60000 > 10.0.3.249.10000: Flags [S], seq 2819965599, win 62727, options [mss 8365,sackOK,TS val 1029816180 ecr 0,nop,wscale 7], length 0
> IP 10.0.3.249.10000 > 10.0.0.215.60000: Flags [S.], seq 3040695044, ack 2819965600, win 62643, options [mss 8961,sackOK,TS val 1224784076 ecr 1029816180,nop,wscale 7], length 0
> IP 10.0.0.215.60000 > 10.0.3.249.10000: Flags [.], ack 1, win 491, options [nop,nop,TS val 1029816181 ecr 1224784076], length 0
> IP 10.0.0.215.60000 > 10.0.3.249.10000: Flags [S], seq 2681819307, win 62727, options [mss 8365,sackOK,TS val 572088282 ecr 0,nop,wscale 7], length 0
> IP 10.0.3.249.10000 > 10.0.0.215.60000: Flags [.], ack 1, win 490, options [nop,nop,TS val 1224794914 ecr 1029816181,nop,nop,sack 1 {4156821004:4156821005}], length 0
>
> It seems to be working correctly, but the last ACK was generated by
> tcp_send_dupack() and PAWSEstab was increased.  This is because the
> second connection has a smaller timestamp than the first one.
>
> In this case, we should send a challenge ACK instead of a dup ACK and
> increase the correct counter to rate-limit it properly.

OK, but this seems about the same thing to me. A challenge ACK is a dup ACK ?

It is not clear why it matters, because most probably both ACK make no
sense for the sender ?

>
> Let's check the SYN bit after the PAWS tests to avoid adding unnecessary
> overhead for most packets.
>
> Link: https://docs.aws.amazon.com/elasticloadbalancing/latest/network/introduction.html [0]
> Link: https://docs.aws.amazon.com/elasticloadbalancing/latest/network/load-balancer-target-groups.html#client-ip-preservation [1]
> Fixes: 0c24604b68fc ("tcp: implement RFC 5961 4.2")

The core of the change was to not send an RST anymore.
I did not change part of the code which was not sending an RST :)



> Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
> ---
>  net/ipv4/tcp_input.c | 2 ++
>  1 file changed, 2 insertions(+)
>
> diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
> index cc072d2cfcd8..89fca4c18530 100644
> --- a/net/ipv4/tcp_input.c
> +++ b/net/ipv4/tcp_input.c
> @@ -5714,6 +5714,8 @@ static bool tcp_validate_incoming(struct sock *sk, struct sk_buff *skb,
>             tp->rx_opt.saw_tstamp &&
>             tcp_paws_discard(sk, skb)) {
>                 if (!th->rst) {
> +                       if (unlikely(th->syn))
> +                               goto syn_challenge;
>                         NET_INC_STATS(sock_net(sk), LINUX_MIB_PAWSESTABREJECTED);
>                         if (!tcp_oow_rate_limited(sock_net(sk), skb,
>                                                   LINUX_MIB_TCPACKSKIPPEDPAWS,
> --
> 2.30.2
>