Fabric8 Vagrant & DNS issue with container

[Charles Moulliard]

Hi,

Since last Friday, I have a DNS issue with the docker container Fabric8 ElasticSearch Ipaas. We can ping a host within the container but can't curl/wget a http link

Example 

# wget https://www.github.com/lukas-vlcek/origin-aggregated-logging/releases/download/v0.1/search-guard-0.5.1.zip

--2016-06-13 07:02:29--  https://www.github.com/lukas-vlcek/origin-aggregated-logging/releases/download/v0.1/search-guard-0.5.1.zip

Resolving www.github.com (www.github.com)... failed: Name or service not known.

wget: unable to resolve host address ‘www.github.com’

# ping www.github.com

PING github.com (192.30.252.123): 56 data bytes

64 bytes from 192.30.252.123: icmp_seq=0 ttl=61 time=123.815 ms

64 bytes from 192.30.252.123: icmp_seq=1 ttl=61 time=130.109 ms

The resolv.conf file is defined as the docker container of Fabric8

# more /etc/resolv.conf

search default.svc.cluster.local svc.cluster.local cluster.local

nameserver 172.30.0.1

nameserver 10.0.2.3

options ndots:5

How can I resolve that ?

Regards,
--

Charles Moulliard

Apache Committer & PMC / Architect @RedHat

Twitter : @cmoulliard | Blog :  http://cmoulliard.github.io

[Charles Moulliard]

Additional info: 

Within the docker ES container, I can ping the DNS Server 10.0.2.3 but not 172.30.0.1

[James Strachan]

Does DNS work OK on the VM? I've seen this fail before in the past. 

Sometimes this fixes it:

  vagrant reload

Or sometimes I've had to edit docker to add explicit DNS servers. e.g. on the VM do this:

  sudo bash

  vi /etc/resolv.conf

then add:

nameserver 8.8.4.4

nameserver 8.8.8.8

save then run:

  sudo service docker restart

--
You received this message because you are subscribed to the Google Groups "fabric8" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fabric8+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

James
-------
Red Hat

Twitter: @jstrachan
Email: james.s...@gmail.com

Blog: https://medium.com/@jstrachan/

fabric8: http://fabric8.io/

open source microservices platform

[Charles Moulliard]

I think that this problem occurs when we switch from a fixed IP network (LAN) to Wifi network

Question: 

The Virualbox created from Vagrant file contains 3 Networks -->

- 172.17.0.1 : Docker Network

- 10.0.2.0 : DNS Server ?

- 172.28.128.0: OpenShift Network

The network 10.0.2.0 contains a gateway address 10.0.2.2 and DNS Server : 10.0.2.3. Is the DNS Server created by OpenShift (= SkyDNS Server) ? How can we configure it ?

[James Strachan]

On 13 June 2016 at 08:58, Charles Moulliard <ch0...@gmail.com> wrote:

I think that this problem occurs when we switch from a fixed IP network (LAN) to Wifi network

Yeah - whenever I move networks I always do a 'vagrant reload' to be sure. That usually fixes it

[Charles Moulliard]

Hi,

Have we changed something with our Vagrant VM box (DNS, Landrush, ...) as since last Friday I must systematically add these nameservers to the /etc/resolv.conf file of the box created using this file (https://github.com/fabric8io/fabric8-installer/blob/master/vagrant/openshift/Vagrantfile) and restart docker/OSE ?

vi /etc/resolv.conf
nameserver 8.8.4.4
nameserver 8.8.8.8

service docker restart
systemctl start openshift.service

Regards,

Charles

[James Strachan]

I've created / recreated / reloaded the Vagrant VM for openshift and kubernetes a few times since Saturday and never had to touch resolv.conf once

[James Strachan]

as we talked about - when changing networks - did you do a 'vagrant reload'?

[Charles Moulliard]

Since this Monday, I'm working with a new VM Machine and my Mac is connected using an Ethernet Cable (RJ45) to Internet. I haven't switched since yesterday to the WIFI network. The IP address of my MAcOS machine is fix (= NO DHCP Server is used)

Doing multiple vagrant reload doesn't resolve the issue

HA router is working : https://www.dropbox.com/s/di9hnijr4scc1y3/Screenshot%202016-06-14%2010.02.28.png?dl=0

Here is the config of the Vagrant VM Box

[root@vagrant vagrant]# more /etc/resolv.conf 

; generated by /usr/sbin/dhclient-script

nameserver 10.0.2.3

nameserver 8.8.4.4

nameserver 8.8.8.8

[root@vagrant vagrant]# more /etc/hosts

127.0.0.1   vagrant.f8 vagrant localhost localhost.localdomain localhost4 localhost4.localdomain4

::1         localhost localhost.localdomain localhost6 localhost6.localdomain6

and Landrush

cat /Users/chmoulli/.vagrant.d/data/landrush/log/landrush.log

I, [2016-06-13T19:31:33.370835 #95339]  INFO -- : Starting RubyDNS server (v0.8.5)...

I, [2016-06-13T19:31:33.370989 #95339]  INFO -- : Listening on udp:0.0.0.0:10053

I, [2016-06-13T19:31:33.371241 #95339]  INFO -- : Listening on tcp:0.0.0.0:10053

I, [2016-06-14T08:06:41.841452 #95339]  INFO -- : Resource class: Resolv::DNS::Resource::IN::A

I, [2016-06-14T08:06:41.843289 #95339]  INFO -- : Resource: #<Resolv::DNS::Resource::IN::A:0x000001008a6ed8 @address=#<Resolv::IPv4 172.28.128.4>>

[Charles Moulliard]

I think that I know the problem

When we create the VirtualBox using Fabric8-Openshift Vagrantfile with the Version 1.1.6 of origin, the DNS Servers defined within the container are correct

sh-4.2$ more /etc/resolv.conf 

search default.svc.cluster.local svc.cluster.local cluster.local localdomain

nameserver 172.28.128.4

nameserver 10.0.2.3

But if now, we use the version 1.3.0.alpha of Origin, then we get a "wrong" DNS IP Address

nameserver 172.30.0.1,

nameserver 10.0.2.3

Question: Can we change the IP Address of the DNS Server to use 172.28.128.4 and not  172.30.0.1?

[Charles Moulliard]

The workaround for OSE Origin 1.3.1 is perhaps to prepend the ip address of the nameserver 172.28.128.4 as described here :http://developers.redhat.com/blog/2015/11/19/dns-your-openshift-v3-cluster/

[Charles Moulliard]

https://github.com/fabric8io/fabric8-installer/pull/90  submitted to resolve the issue. If someone can have a look and validate it, that should be super !

To unsubscribe from this group and stop receiving emails from it, send an email to fabric8+unsubscribe@googlegroups.com.

To unsubscribe from this group and stop receiving emails from it, send an email to fabric8+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

Charles Moulliard

Apache Committer & PMC / Architect @RedHat

Twitter : @cmoulliard | Blog :  http://cmoulliard.github.io

--

Charles Moulliard

Apache Committer & PMC / Architect @RedHat

Twitter : @cmoulliard | Blog :  http://cmoulliard.github.io