On 02/04/2015 02:32 PM, Kevin Fenzi wrote:

On Tue, 03 Feb 2015 17:54:27 +0100 Miroslav Suchy msuchy@redhat.com wrote:

...

[root@fed-cloud10 etc(keystone_admin)]# telnet 209.132.184.9 443 Trying 209.132.184.9... telnet: connect to address 209.132.184.9: No route to host

I am able to connect using 172.24.0.0 network, but it would be nice (and required to have new OS) to be able to communicate even over public IP.

Can someone (nirik, ssmoogen) fix this please? And I would need to do that for fed-cloud11 too. It seems that without this fixed I am unable to add compute node to OS.

It seems the interfaces on 10 are backwards from the ones on 09... but I am not sure how it even works that we can access 10. ;)

(ie, it's trying to use eth1 to talk to the external on 09 instead of eth0).

Will try and figure out whats going on...

kevin

---

infrastructure mailing list infrastructure@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/infrastructure

I could not get to it now so I can not validate, but... I have in roles/cloud_compute/tasks/main.yml - lineinfile: dest=/etc/sysconfig/network-scripts/ifcfg-eth1 regexp="^ONBOOT=" line="ONBOOT=yes" notify: - restart network - lineinfile: dest=/etc/sysconfig/network-scripts/ifcfg-eth1 regexp="^NETMASK=" line="NETMASK=255.255.255.0" notify: - restart network - lineinfile: dest=/etc/sysconfig/network-scripts/ifcfg-eth1 regexp="^IPADDR=" line="IPADDR={{compute_private_ip}}" notify: - restart network - lineinfile: dest=/etc/sysconfig/network-scripts/ifcfg-eth1 regexp="BOOTPROTO=" line="BOOTPROTO=none" notify: - restart network Can it be that there is no MAC binding in udev so eth1 can become eth0 and vice versa? Or something else?

BTW it does not block me any more. I managed to change endpoints to use internal IP. But still nice to have.