[hipl-users] Re: hipsetup failes
- From: Mika Kousa <mkousa@xxxxxxxxx>
- To: HIPL Users <hipl-users@xxxxxxxxxxxxx>
- Date: Fri, 6 Feb 2004 20:42:49 +0200 (EET)
On Fri, 6 Feb 2004, Simon Schuetz wrote:
-> HI,
-> I tried to test my HIP machines using hipsetup.
-> The base exchange failes with the messages below.
-> Host Mario plays the role of "oops".
-> Ethereal shows a complete base exchange (I1,R1,I2,R2).
->
-> What's going wrong?
Let's see.
-> Initiator side:
->
-> =======================================================
-> Initiator mode
-> Connecting to peer host mario
-> test hit:421d6d7e6e2e0d0dde05799b2a48e45d is HIT -> map
-> test ipv6:421d6d7e6e2e0d0dde05799b2a48e45d: skip, is hit or same
-> test ipv6:3ffe0000000000000000000000000002: MAP
-> test hit:3ffe0000000000000000000000000002 is not HIT
-> got gai addresses:
-> GAI: ai_flags=2048 ai_family=10 ai_socktype=1 ai_protocol=6
-> ai_addrlen=28 ai_canonname=(null)
-> AF_INET6: ship6_port=12345
-> in6_addr=0x421d6d7e6e2e0d0dde05799b2a48e45d
->
->
-> connect took 3.0058450000 sec
-> OK
-> grep: mario: No such file or directory
This error from grep looks weird. hipsetup shouldn't use the given host
name anywhere as a file name. Have you added host mario's HIT and name to
/etc/hip/hosts ? Something like
421d:6d7e:6e2e:0d0d:de05:799b:2a48:e45d mario
-> ### HIP BASE EXCHANGE FAILED !!! ####
-> ========================================================
->
-> The responder side shows:
->
-> ========================================================
-> Responder mode
-> nc6: listening on :: 12345 ...
-> nc6: listening on 0.0.0.0 12345 ...
-> nc6: connect to 421d:6d7e:6e2e:d0d:de05:799b:2a48:e45d 12345 from
-> 534a:5fa4:78e:ad51:cfeb:fabd:c52d:bb0 1121
Determining from this the TCP SYN packet from the initiator got through,
so the HIP base exchange was successful. I guess that the previously
mentioned error from grep causes the test in hipsetup script's function
check_hip_state_established to fail, so hipsetup tells that the base
exchange failed. If you look at file /proc/net/hip/sdb_state on both hosts
and see that the peer states are established, then the base exchange was
successful and the only problem is some bug in hipsetup script. See if the
problem continues if you use original host name "oops" where you now have
changed "mario".
Other related posts:
