[kwlug-disc] Problem reaching server

[B.S.]

I have also been experiencing disappearing connectivity since early this 
year with a remote K/Ubuntu 14.04. eth0 appear up and happy per 
ifconfig, but nothing pings.

So ... possibly this has nothing to do with vm's.

In my case, this started happening on a second machine within the last 
week or so - the only common factor to the two is running openvpn.

I had suspected route munging with vpns going up and down, but have 
since seen ifconfig with ok listing and unable to ping other nodes on 
the net. In this case 'ifconfig eth0 down' ... 'ifconfig eth0 up' 
brought things back to life.

Which is all to say ... it's possible the problems you are experiencing 
lie at a lower hardware / ethernet level than vm (or openvpn). [Mind 
you, the bridge being in/accessible will likely also cause routes to go 
up and down. Perhaps a commonality.]

In the first machine's case I set cron to reboot it  daily and haven't 
had a problem since. Since I can't be in front of it when it gets this 
way to diagnose it, and only being in front of it in such circumstances 
is useful for trying to figure out what's going on. Remotely, if it's 
inaccessible, it's inaccessible, you're done - you have to beat on it 
from the inside (direct keyboard) to examine state, cause, and remediation.

So ... problems may be at a lower level than application software.

FWIW.


On 03/12/2015 07:04 PM, Khalid Baheyeldin wrote:
> That happened to me when I had bridging on the physical host that has
> virtualization. Things work, until you reboot, then server is not reachable.
>
> I didn't figure out why, and ended up abandoning the bridge I added to
> /etc/network/interfaces, and using the default networking that is setup by
> libvirt.
>
> On Thu, Mar 12, 2015 at 6:51 PM, CrankyOldBugger <crankyoldbugger at gmail.com>
> wrote:
>
>> I'm puzzled...  I rebooted my Ubuntu server this morning (as updates
>> needed a reboot) and now I can't reach it.
>>
>> Here's what I determined:
>>
>> It was running fine before the reboot!
>> I can't ping the server from another machine on the same subnet.
>> I can't SSH to the server.
>> SSH is running on the server.
>> I can't telnet to the server
>> I can ssh from the server to itself, so ssh is running and listening on 22.
>> If I try to ssh from another linux box on the same subnet, using either
>> hostname or IP, I get "No route to host".
>> Traceroute should only one hop, that being the client I started traceroute
>> from, but that line ends in !H, which I believe means can't connect.
>> IP address is fine.
>> I can't ping out from the server.
>> Now here's the fun part: I have two virtual machines hosted on this
>> server, and they're both running fine.  I can connect to them and they can
>> connect to the internet.
>>
>> Not sure what to think at this point.  Anyone have any ideas?