On a Dell PE 1950 I have a BMC configured for IPMI that is not
responding to IPMI, and which responds to pings and HTTP connects by
sending an ICMP host unreachable (from itself!), e.g.:
16:40:34.809694 IP (tos 0x0, ttl 58, id 17184, offset 0, flags [none],
proto: ICMP (1), length: 112) 10.139.1.72 > 10.139.139.203: ICMP host
10.139.1.72 unreachable, length 92
IP (tos 0x0, ttl 57, id 0, offset 0, flags [DF], proto: ICMP
(1), length: 84) 10.139.139.203 > 10.139.1.72: ICMP echo request, id
6503, seq 2, length 64
16:48:54.359616 IP (tos 0x0, ttl 58, id 17193, offset 0, flags [none],
proto: ICMP (1), length: 88) 10.139.1.72 > 10.139.139.203: ICMP host
10.139.1.72 unreachable, length 68
IP (tos 0x0, ttl 6, id 52976, offset 0, flags [none], proto:
TCP (6), length: 60) 10.139.139.203.43390 > 10.139.1.72.http: S
3287306515:3287306515(0) win 5840 <mss 1460,sackOK,timestamp
4175788272[|tcp]>
I'm pretty sure that this is not being mocked up by an intervening
firewall or anything, since the TTL hopcount coming back is exactly what
I would expect (64 - 58 = 6), but the initial TTL is a bit odd
(successful ping coming back from another BMC is 128 - hops TTL) - the
64 initial TTL makes me wonder if the host (or its iptables instance)
isn't generating the host unreachable.
Has anyone else seen anything like this? A reboot hasn't seemed to
clear it, and per syscfg these two systems have the same configurations.
@alex
--
mailto:[email protected]
_______________________________________________
Linux-PowerEdge mailing list
[email protected]
https://lists.us.dell.com/mailman/listinfo/linux-poweredge
Please read the FAQ at http://lists.us.dell.com/faq
