Hey Noel, 

I had this issue too. This workaround did the job for me:

- stop cloud Agent service on KVM host
- execute iptables -F, ebtables -F on KVM host
- uninstall ebtables package on KVM host (I had to use --nodeps) 
- start cloud Agent service on KVM host



On 8 Feb 2013, at 11:56, Noel King <noelk...@gmail.com> wrote:

> Hi Basha
> 
> Again thank you for your reply I had previously gone through that
> documentation/chapter and verified my configuration and am still having
> this problem.
> 
> What I am seeing is as follows,
> 
>   - For investigation only purposes I stop ip tables on the host effective
>   and connectivity issue is resolved.
>   - Some time later iptables state is being brought back up and blocking
>   again
> 
> I feel my current option is
> 
>   - Remove all iptables configuration and reset using the scripts outlined
>   in that chapter.
> 
> This is very odd as I cant find this secondary iptables process that gets
> created.
> 
> Cheers
> 
> Noel
> 
> On 8 February 2013 17:11, S.M Basha <smbash...@gmail.com> wrote:
> 
>> Noe,
>> 
>> i hope this will be help you.
>> 
>> 
>> http://incubator.apache.org/cloudstack/docs/en-US/Apache_CloudStack/4.0.0-incubating/html/Installation_Guide/hypervisor-kvm-install-flow.html
>> 
>> Regards
>> Basha
>> 
>> On Fri, Feb 8, 2013 at 9:31 PM, Noel King <noelk...@gmail.com> wrote:
>> 
>>> Hi
>>> 
>>> I have upgraded to Cloudstack 4.0 on multiple hosts on CentOS 6.3 with
>> KVM
>>> hosts and I are currently experiencing intermittent iptables related
>> issues
>>> connecting to hosts which gets fixed once I temporarily stop iptables .
>>> 
>>> Would somebody mind sharing their test and generic
>> /etc/sysconfig/iptables
>>> configuration so I can compare it too.   I am currently happy our config
>>> matches the install guide any assistance here would be greatly
>> appreciated.
>>> 
>>> 
>>> Kind regards
>>> 
>>> Noel
>>> 
>> 
>> 
>> 
>> --
>> Thanks & Regards
>> S.M.Basha
>> 
>> "If opportunity doesn't knock the door; open the door and drag opportunity
>> inside."
>> 

Reply via email to