Horizon dashboard cannot visit after deploy 1 controller+1 compute (plus cinder)with mos-6.1 centos env

[lidiwen]

Steps to reproduce:
1. Fuel 6.1 Master Node, deploy as VM on vsphere.
2. Create MOS env and select Cent OS deployment.
3. Add 1 nodes with controller role created by vSphere VMs
4. Add 1 nodes with compute ( plus cinder )role created by vSphere VMs
5. Select Neutron GRE network topology
6. Run network verification
7. Deploy the cluster

It can deploy successfully and show message on Fuel UI wiht Open stack dashboard access at http://172.16.0.2.

But http://172.16.0.2 cannot visit, only the controller node can ping and get to the 172.16.0.2.

The compute node also cannot visit keystone host (192.168.0.2).
[root@node-14 ~]# cinder list
ERROR: Unable to establish connection to http://192.168.0.2:35357/v2.0/tokens

[root@fuel network-scripts]# fuel --fuel-version
DEPRECATION WARNING: /etc/fuel/client/config.yaml exists and will be used as the source for settings. This behavior is deprecated. Please specify the path to your custom settings file in the FUELCLIENT_CUSTOM_SETTINGS environment variable.
api: '1.0'
astute_sha: 1ea8017fe8889413706d543a5b9f557f5414beae
auth_required: true
build_id: 2015-06-19_13-02-31
build_number: '525'
feature_groups:
- mirantis
fuel-library_sha: 2e7a08ad9792c700ebf08ce87f4867df36aa9fab
fuel-ostf_sha: 8fefcf7c4649370f00847cc309c24f0b62de718d
fuelmain_sha: a3998372183468f56019c8ce21aa8bb81fee0c2f
nailgun_sha: dbd54158812033dd8cfd7e60c3f6650f18013a37
openstack_version: 2014.2.2-6.1
production: docker
python-fuelclient_sha: 4fc55db0265bbf39c369df398b9dc7d6469ba13b
release: '6.1'
release_versions:
  2014.2.2-6.1:
    VERSION:
      api: '1.0'
      astute_sha: 1ea8017fe8889413706d543a5b9f557f5414beae
      build_id: 2015-06-19_13-02-31
      build_number: '525'
      feature_groups:
      - mirantis
      fuel-library_sha: 2e7a08ad9792c700ebf08ce87f4867df36aa9fab
      fuel-ostf_sha: 8fefcf7c4649370f00847cc309c24f0b62de718d
      fuelmain_sha: a3998372183468f56019c8ce21aa8bb81fee0c2f
      nailgun_sha: dbd54158812033dd8cfd7e60c3f6650f18013a37
      openstack_version: 2014.2.2-6.1
      production: docker
      python-fuelclient_sha: 4fc55db0265bbf39c369df398b9dc7d6469ba13b
      release: '6.1'

[lidiwen (lidiwencs)]

diagnostic snapshot log can download from https://bugs.launchpad.net/fuel/+bug/1512574/

[Timur Sufiev (tsufiev-x)]

@lidiwen,

judging by the info you provided here:

"It can deploy successfully and show message on Fuel UI wiht Open stack dashboard access at http://172.16.0.2.

But http://172.16.0.2 cannot visit, only the controller node can ping and get to the 172.16.0.2.

The compute node also cannot visit keystone host (192.168.0.2).
[root@node-14 ~]# cinder list
ERROR: Unable to establish connection to http://192.168.0.2:35357/v2.0/tokens"

I'm assuming that you've been able to ssh from Fuel master node to controller node (denoted as node-14 in your output), but didn't check if Horizon or Keystone services are available on this controller node. Also, judging by the fact that there are Keystone accessibility problems, this is either a problem with starting Apache, or some network connectivity/routing problem. Anyways, from the Horizon developer's POV the existing information is incomplete, so I'm changing the bug status and reassigning it to fuel-dev team.

[Timur Sufiev (tsufiev-x)]

Well, fuel-dev seems to deprecated, the problem domain is IMO closer to fuel-library than to fuel-python.

[Kyrylo Galanov (kgalanov)]

Hello,

According to the attached snapshot all services were up and running. Moreover, a test image was uploaded to glance which would be impossible without keystone running properly.

I would suppose that networking was not completely configured in vcenter.

I can't tell more without having access to the environment.

Best regards,
Kyrylo

[lidiwen (lidiwencs)]

Hello,
I got answers from https://answers.launchpad.net/fuel/+question/273618,
per okosse (okosse) said on 2015-11-04: #7

Hi, lidiwen.
What hypervisor do you plan to use? If I understand right, you want to use KVM into VMs on ESXi nodes. Unfortunately, we didn't test this case.

I just use vsphere vm to deploy the MOS env, didn't use vCenter integration, this issue didn't occur after I change to physical node to deploy.

and also, I setup a gateway in my lab for MOS public network, this issue didn't occur any more.

Thanks,
lidiwen