proxmox 3.2 - internet down after a few seconds

[trustbyte]

i'm trying the latest proxmox and having problems with the network configuration
bridged network, below my configuration

# network interface settings
auto lo
iface lo inet loopback


iface eth0 inet manual


iface eth1 inet manual


auto vmbr0
iface vmbr0 inet static
address 10.100.100.251
netmask 255.255.255.0
gateway 10.100.100.2
bridge_ports eth0
bridge_stp off
bridge_fd 0


route -n
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
10.100.100.0 0.0.0.0 255.255.255.0 U 0 0 0 vmbr0
0.0.0.0 10.100.100.2 0.0.0.0 UG 0 0 0 vmbr0


as soon as it boots up ping to 8.8.8.8 works for a few seconds that it stops working but ping to 10.100.100.2 my firewall/router is still responding.
service networking restart does the trick
again, a few seconds working ping to internet than once again, 100% packet lost.

anyone has some clue?

could it be my firewall?!? or something locally on proxmox that i missed?

 

[Mr.Holmes]

Did you verify the routing table also in the moment when fails?

If the routing table is still ok try (from another console)

tcpdump -i eth0

if packets are sent to the router.

If yes:
- probably a router problem

If no:
- check interfaces with ifconfig -a (maybe a problem with eth1?)

 

[trustbyte]

My routing table is the same when it's working and when it's not working, as in my first post.
My ifconfig -a is the same when it's working and when it's not working.

I have only eth0 and vmbr0. eth1 is not cable connected.

tcpdump on eth0 show almost nothing when internet is not working. when i restart the networking full of things pop up, my ping works.
the interesting part is this I believe, just after the last ping working:

1. proxmox.demo.com > google-public-dns-a.google.com: IC MP echo request, id 3253, seq 250, length 64
2. ARP, Request who-has proxmox.demo.com tell 10.100.100.0, length 46
3. 2c:75:6a:ac:ab:a3 (oui Unknown) > 04:00:c5:00:00:ee (oui Unknown) SNAP Unnumbered, test, Flags [Poll], length 66
4. 2c:75:6a:ac:ab:a3 (oui Unknown) > 04:00:c5:00:00:ee (oui Unknown) SNAP Unnumbered, test, Flags [Poll], length 66






Why is it the my IP ending in 0, 10.100.100.0??
Shouldn't it be 10.100.100.251?

 

[Mr.Holmes]

Hello trustbyte

Yes, if ARP request came from you PROXMOX server it should be 10.100.100.251 - and if your subnet is a /24 (resp. 255.255.255.0) one it is definitely wrong!

Probably there are some more machines connected with the only one network where PROXMOX server is connected to. My impression is something in that network is wrong.

To get clear conditions for further investigations I would set up a second network (with eth1) to which also the client you use for the console is connected to. If it´s possible remove temporarily other parts from 10.100.100.0/24
(btw.: is the netmask correct? Sometimes I forget to specify it and for 10.x.x.x the default is /8 - in such a case network is not working properly any more)
and leave only the router connected.

Then also tcpdump will not bring such a flood of messages any more, recommended also to activate the option -n in order to see the address behind names as e.g. proxmox.demo.com . If not possible use capturing filters (see man tcpdump and man pcap-filter). If you put the output via tcpdump .... > tcpdumpresult into file "tcpdumpresult" it´s easier investigating the results.
Good luck!

Mr.Holmes

P.S.: Do have already VMs running? Maybe one of them is misconfigured.

 

[trustbyte]

thanks Mr. Holmes.

It was something regarding the netmask, as i changed it from /24 to /8 it´s stable, works without glitches, permanently.

I´m not going to go hunt why and where, as other people are involved, some working, etc.