incubator-tashi-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Kushal Dalmia" <kdal...@andrew.cmu.edu>
Subject Re: Tashi Networking Issue
Date Sat, 17 Dec 2011 00:11:05 GMT
Hi Micheal,

We figured that out and now the VMs are sshable from the cluster manager
node.
However, we have a small discrepancy which we want to confirm. When we
create a new VM, we see a log at the primitive scheduler which specifies
the IP address associated with the new VM. However, when we see our DHCP
logs we see that the DHCP IP address request received for the VM is
different from the one seen in the log (both are however from the IP range
specified in the config). Also, the VM gets the latter IP address and is
sshable using that. Is there something wrong with our setup or is this the
expected behavior?

Thanks,
Kushal

On Fri, December 16, 2011 5:38 pm, Michael Stroucken wrote:
> Kushal Dalmia wrote:
>
>> Hi Micheal,
>>
>>
>> The VMs cannot ping the other VMs.Also the bridge has been setup with
>> the default delay etc. I think what is wrong is the routing on host B. I
>> ran a tcpdump on B and tried to ping A from the VM. I can see the
>> packets on B coming from the bridge but nowhere to be forwarded. Another
>> point to note is that the interface with the required subnet (i.e. the
>> subnet of the VMs and the bridge IP addresses) is a virtual interface
>> and not a physical card. The hosts A,B and C actually belong to a diff
>> subnet physically. But we want a new subnet for the VMs and the virtual
>> bridges. Any pointers towards how to setup such a config would be
>> greatly helpful. The end result we want is just that the VMs belong to
>> a different subnet than the physical hosts and still be able to ping
>> each other and all the hosts.
>>
>
> Hi Kushal,
>
>
> If you are moving data between bridges or interfaces; have you turned ip
> routing on?
>
> Greetings,
> Michael.
>
>
>



Mime
View raw message