Re: [Qemu-discuss] QEMU with pre-created TAP in user mode

[anatoly techtonik]

On Mon, Jul 23, 2012 at 10:50 PM, Mike Lovell <address@hidden> wrote:
> On 07/20/2012 03:23 PM, anatoly techtonik wrote:
>>
>> Thanks. Now QEMU starts ok, but I still unable to connect to guest
>> using SSH after giving both guest eth1 and host tap0 IPs from the same
>> network. Pings work ok both sides. SSH to other virtual machines on
>> the same host (under libvirt) works ok. The error:
>>
>> $ ssh -vvv address@hidden
>> OpenSSH_5.9p1, OpenSSL 1.0.0j-fips 10 May 2012
>> debug1: Reading configuration data /etc/ssh/ssh_config
>> debug1: /etc/ssh/ssh_config line 50: Applying options for *
>> debug2: ssh_connect: needpriv 0
>> debug1: Connecting to 192.168.16.15 [192.168.16.15] port 22.
>> debug1: Connection established.
>> debug3: Incorrect RSA1 identifier
>> debug3: Could not load "/home/user/.ssh/id_rsa" as a RSA1 public key
>> debug1: identity file /home/user/.ssh/id_rsa type 1
>> debug1: identity file /home/user/.ssh/id_rsa-cert type -1
>> debug1: identity file /home/user/.ssh/id_dsa type -1
>> debug1: identity file /home/user/.ssh/id_dsa-cert type -1
>> ssh_exchange_identification: read: Connection reset by peer
>>
>> /var/logs/auth.log on guest shows this:
>> ... sshd[1446]: Did not receive identification string from 192.168.16.1
>>
>> For other machines output is different:
>> ...
>> debug1: identity file /home/user/.ssh/id_dsa-cert type -1
>> debug1: Remote protocol version 2.0, remote software version
>> OpenSSH_5.5p1 Debian-6+squeeze2
>> debug1: match: OpenSSH_5.5p1 Debian-6+squeeze2 pat OpenSSH*
>> debug1: Enabling compatibility mode for protocol 2.0
>> ...
>
>
> it sound like you are using multiple guests. are you specifying the macaddr=
> option on the guest network devices for each one? if you don't specify it,
> each qemu process will use the same mac address for each guest,
> 52:54:00:12:34:56. using multiple guests with the same mac address can lead
> to very weird behaviors. "-net nic,macaddr=52:54:00:aa:bb:cc" will tell the
> guest to use 52:54:00:aa:bb:cc instead of the default.

Not the cause. I have several virtual machines in virt-manager (`other
machines` referenced above), but this time I ensured they are all shut
them down. Only one qemu process is running.

$ ps aux | grep qemu
user  21221  6.4  1.4 251424 120936 pts/0   Sl+  23:29   0:15 qemu-kvm
-hda linux.img -net user -net nic -net nic -net
tap,ifname=tap0,script=no
user  21320  0.0  0.0   4744   796 pts/1    S+   23:33   0:00 grep
--color=auto qemu

Host (Fedora 17) interface:
tap0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet 192.168.16.1  netmask 255.255.255.0  broadcast 192.168.16.255
        ether 22:04:19:e7:bb:0e  txqueuelen 500  (Ethernet)

Guest (Debian Squeeze):
eth2: HWaddr 52:54:00:12:34:57
eth3: HWaddr 52:54:00:12:34:56  inet addr:192.168.16.15/24

$ ssh address@hidden
ssh_exchange_identification: read: Connection reset by peer


I've tried `-net nic,macaddr=52:54:00:aa:bb:cc` and qemu-kvm
segfaulted when I tried to assign IP address with:
$ ifconfig eth4 192.168.16.15/24