[Swan] libreswan/racoon interoperability problem with NAT-T

Paul Wouters paul at nohats.ca
Sun Apr 2 22:12:44 UTC 2017


On Fri, 31 Mar 2017, Xinwei Hong wrote:

You can try ikepad=no

Other then that, I don't understand what's wrong with racoon.

Paul

> Date: Fri, 31 Mar 2017 18:59:18
> From: Xinwei Hong <xhong at skytap.com>
> Cc: swan at lists.libreswan.org
> To: Paul Wouters <paul at nohats.ca>
> Subject: Re: [Swan] libreswan/racoon interoperability problem with NAT-T
> 
> Thank you Paul.
> I added "compress=yes", seems not help. Compression_algorithm is only used in phase 2, right? The
> failure here is phase 1. Firewall does not like the issue either because packet on UDP 4500 can be seen
> on both end.
> 
> I have attached racoon log file. I see:
> 
> invalid length of payload
> 
> ERROR: phase1 negotiation failed due to time up. 8c6688cabaaf90ca:a8d8758b59948609
> 
> ERROR: phase2 negotiation failed due to time up waiting for phase1. ESP 10.2.128.240[0]->10.0.0.1[0]
> 
> 
> Thanks,
> Xinwei
> 
> 
> On Fri, Mar 31, 2017 at 1:53 PM, Paul Wouters <paul at nohats.ca> wrote:
>       On Thu, 30 Mar 2017, Xinwei Hong wrote:
>
>       [moderator note: please next time post a pointer to a "pastebin" of the
>       log instead of included large log files]
>
>             I have a VPN setup between libreswan (pluto+netkey) and a racoon
>             (racoon+netkey), the racoon is behind a NAT device. The negotiation somehow
>             failed saying that "NAT-D payload #0 doesn't
>             match"
> 
>
>       There is not a NAT payload error. Your device has more then one IP
>       address, and it is trying to calculate the payload for both your
>       IP addresses. It notes that 10.0.0.1 does not match. But it also
>       notes that 10.2.128.240 works fine.
>
>       Your actual problem is:
>
>               Mar 30 19:47:02 vvr-10 pluto[24271]: vvr-0:
>               "conn_vvr-0-ipsectunnel-0-remote-0" #1246: max number of retransmissions
>               (8) reached STATE_MAIN_I3.  Possible authentication failure: no
>               acceptable response to our first encrypted message
>
>       Your racoon logs will likely have more information in it then your
>       libreswan log, as racoon is the party rejecting the packet.
>
>             conn conn_vvr-0-ipsectunnel-0
>                     authby=secret
>                     left=10.2.128.240
>                     right=10.2.128.241
>                     ike=3des-sha1;modp1024
>                     phase2alg=3des-sha1;modp1024
>                     leftsubnet=10.100.0.0/24
>                     rightsubnet=10.100.1.0/24
> 
>
>             on racoon, we have racoon.conf
>             # Phase 1 (Main Mode) Configuration
>             remote 10.2.128.240 {
>              exchange_mode main;
>              proposal_check obey;
>              lifetime time 28800 seconds;
>              nat_traversal on;
>              #script "phase1-up.sh" phase1_up;
>              #script "phase1-down.sh" phase1_down;
>              dpd_delay 15; dpd_retry 5; dpd_maxfail 5;
>              proposal {
>               encryption_algorithm 3des;
>               hash_algorithm sha1;
>               dh_group modp1024;
>               authentication_method pre_shared_key;
>              }
>             }
>
>             # Phase 2 (Quick Mode) Configuration/Proposal (for IPsec SA).
>             sainfo anonymous {
>              encryption_algorithm 3des;
>              authentication_algorithm hmac_sha1;
>              pfs_group modp1024;
>              lifetime time 3600 seconds;
>              compression_algorithm deflate;
>             }
> 
>
>       The only mismatch I see is for compression. You can try and removing the
>       compression line from racoon, or adding compress=yes to libreswan.
>
>             Mar 30 19:47:52 testhost-601-1 racoon: ERROR: phase1 negotiation failed due to
>             time up. 80b77211a2f1ddba:141872152ca7772f
> 
>
>       This is odd. Another possibility is that you have firewalled UDP 4500
>       and so the NAT switching from UDP 500 to UDP 4500 is failing?
>
>       Paul
> 
> 
> 
>


More information about the Swan mailing list