VLANs on VM as routers and posibility of ssh problems?

[syd]

Hi,

I have two routers working as VMs on Proxmox. Routers (CentOS5) got VLANs ethernets which are set up on VE like: eth0.11, eth0.22, eth1.33 etc. Main job is going well (it seems like that), but I have a problem with get in to VM which is after other VM.

Idea schema is easy:
host1---VM1---VM2---Internet

VM1 and VM2 are running on Proxmox. On VM1 is NAT, VM2 is configured only on public IPs.
Problem:
I can't get in via SSH to the VM2 from host1, but I can from VM2 and from Internet. Only trail which I get is from VM2 log:
sshd [xxxx]: Did not receive identification string from <IP nated 1:1 to host1>

After some googling I haven't found any sensible reason for that. So I though that maybe is knowing some nuance about hops through virtual routers on the same VE and SSH troubles here.

Regards

 

[udo]

Hi,
schould work! I guess a wrong routing entry (VM2 or VM1), or wrong NAT. (look with tcpdump)
Do you use for the vlans bridges like

auto eth0
iface eth0 inet static
        address  0.0.0.0
        netmask  0.0.0.0

auto eth0.11
iface eth0.11 inet static
        address  0.0.0.0
        netmask  0.0.0.0

auto eth0.22
iface eth0.22 inet static
        address  0.0.0.0
        netmask  0.0.0.0

auto eth0.33
iface eth0.33 inet static
        address  0.0.0.0
        netmask  0.0.0.0

auto vmbr11
iface vmbr11 inet manual
        bridge_ports eth0.11
        bridge_stp off
        bridge_fd 0

auto vmbr22
iface vmbr22 inet manual
        bridge_ports eth0.22
        bridge_stp off
        bridge_fd 0

auto vmbr33
iface vmbr33 inet manual
        bridge_ports eth0.33
        bridge_stp off
        bridge_fd 0

Udo

 

[syd]

Hi,
how can I get that print? I don't want to paste here result of "ip address show". Also I have to go to data centre to log in to Proxmox now :/ Second also, Debian is not my favourite distribution

Thanks!

 

[udo]

Hi,
how can I get that print?

Thats the (part of) content of /etc/network/interfaces

I don't want to paste here result of "ip address show". Also I have to go to data centre to log in to Proxmox now :/

sure,
but you can use something like this:
Original: 74.125.230.82/29
Write as 11.22.230.82/29
so it's possible to find issues with routing/netmasks and so on.

Second also, Debian is not my favourite distribution

Thanks!

long time ago i used rpm-based distros (suse), but after a short familiarization with deb-based distros debian win.

Udo

 

[syd]

After long time, but just saw this older post, while writing on other problems. But... the upper problem still exist. I just left it and I'm logging from my LAN to VM2 not directly for its IP, but after log in to VM1 and then do "ssh <ip_of_VM2>". It's annoying, but I have no idea what's is wrong till today. Also I started to doubt that Proxmox can handle more than one router on VE, when traffic is going through router1 via router2 via routerX, which are all on the same PVE :/ (I want to have only two and have it separated by VLANs and its even working, but there are some issues like this and others things)

So, now I'm more sapient than months ago and have my config with me It's the same, like when I was making this thread:

# network interface settings
auto lo
iface lo inet loopback

iface eth0 inet manual

iface eth1 inet manual

auto vmbr0
iface vmbr0 inet static
        address  172.16.1.10
        netmask  255.255.255.0
        bridge_ports eth0
        bridge_stp off
        bridge_fd 0

auto vmbr1
iface vmbr1 inet manual
        bridge_ports eth1
        bridge_stp off
        bridge_fd 0

auto vmbr11
iface vmbr11 inet static
        address  172.21.1.10
        netmask  255.255.0.0
        gateway  172.21.1.1
        bridge_ports eth0.11
        bridge_stp off
        bridge_fd 0

auto vmbr22
iface vmbr22 inet manual
        bridge_ports eth1.22
        bridge_stp off
        bridge_fd 0

auto vmbr33
iface vmbr33 inet manual
        bridge_ports eth0.33
        bridge_stp off
        bridge_fd 0

auto vmbr44
iface vmbr44 inet manual
        bridge_ports eth1.44
        bridge_stp off
        bridge_fd 0

auto vmbr88
iface vmbr88 inet manual
        bridge_ports eth0.88
        bridge_stp off
        bridge_fd 0

Ok, so if anybody have some ideas, please share. The problem is that SSH is hanging, when I try to connect directly to VM via other VM on the same PVE. Pings go well to twos. Internet traffic is going through this two routers as KVM on PVE (but with some exceptions, bcoz I have a problem with routinh public classes to my LAN, via two VM routers, but this is bigger issue on other thread, I think). I have Proxmox v1.7 (can't be easily updated do 1.9, bcoz of this problem: http://forum.proxmox.com/threads/7063-quot-bad-udp-cksum-quot-on-DNS-requests-from-PVE -no answers, btw )

Best Regards!
syd