|
On Wed, Oct 08, 2003 at 03:11:46AM +0800, Michael Lee wrote:
> Some more info about my problem. I run tcpdump at my problem node
> (lvs2), and I see some strange entries, such as:
I am not really sure what the problem is.
Hopefully the explanation below will help you to work it out.
> 192.168.1.11 > 192.168.1.12: icmp: net 192.168.0.1 unreachable. [tos 0xc0]
It appears that 192.168.1.12 tried to send a packet to 192.168.0.1
via 192.168.1.11. Persumably this means that 192.168.1.12 routes
192.168.0.1 via that 192.168.1.11 take a look at the routing
table on 192.168.1.12 to verify this (route -n). 192.168.1.11
doesn't know how to get to 192.168.0.1, most likely
because it doesn't have a route to get to there.
The packet could have come from any process. It may or may
not be related to your LVS problem. If it happens repeatedly
it probably is, else you can probably ignore it.
> arp who-has 192.168.1.11 tell 192.168.1.12
This is perfectly normal. 192.168.1.12 wants to send a packet
to 192.168.1.11. As they are on the same network 192.168.1.12 sends an
arp request to find out what the hardware address of 192.168.1.11
is. 192.168.1.12 will store the result for a while, usually
a few minutes, and use the cached results for subsequent packets
during this period.
> 192.168.1.11 > 192.168.1.13: icmp: 192.168.1.11 udp port 1900 unreachable
> [tos 0xc0]
192.168.1.13 has tried to open a connection to 192.168.1.11 port 1900.
This isn't available (i.e. there is no daemon listening on this port)
so this message is sent to notify 192.168.1.13 of this.
What is supposed to be on port 1900?
> 192.168.0.1 is not in my network. lvs1 is 192.168.1.11, lvs2 is 192.168.1.12,
> lvs3 is 192,168.1.13, VIP is 192.168.2.110.
>
> also, there are:
>
> PPPoE [ses 0x405] 218.103.30.69 > 218.102.193.243 icmp: echo request.
>
> I have no clue what it is....
As subsequently discussed, this is to do with your uplink to
the internet. It can be ignored.
--
Horms
|
