OpenStack with CentOS distro not installed.

Did CentOS actually install or not? I mean, can you login to the CentOS instance?

{"build_id": "2014-03-17_01-18-16", "mirantis": "yes", "build_number": "248", "nailgun_sha": "f58aad317829112913f364347b14f1f0518ad371", "ostf_sha": "dc54d99ddff2f497b131ad1a42362515f2a61afa", "fuelmain_sha": "16637e2ea0ae6fe9a773aceb9d76c6e3a75f6c3b", "astute_sha": "f15f5615249c59c826ea05d26707f062c88db32a", "release": "4.1", "fuellib_sha": "73313007c0914e602246ea41fa5e8ca2dfead9f8"}

CentOS 6.4 distributive,
KVM hipervizor,
Nova Network,
Install Savanna.
{Controller (KVM)} + {Compute-Ceph (Hardware).}

Result of deploy:
Have ssh access to nodes.
OpenStack is installed.
But logs of node have ERRORs.

Did you run OSTF on the cluster? Did it pass? I mean, if there are some errors in the logs, it does not mean that OpenStack cluster is not opertional. Please, provide more information on what is not working.

{"build_id": "2014-04-01_04-17-26", "mirantis": "yes", "build_number": "264", "nailgun_sha": "a2ae66f22dd7fff9cb4cfaae87aca4e6c7aad331", "ostf_sha": "dfb5c8bfdc5fc2d937a05f079cb891e6e89ccd92", "fuelmain_sha": "16637e2ea0ae6fe9a773aceb9d76c6e3a75f6c3b", "astute_sha": "5bcacc84cdaee3b31f1178d92a1c0681dc6ce520", "release": "4.1", "fuellib_sha": "52e7f57695f33bafa5d84d524d77f1bc3a2289b2"}

Cluster settings:
virt_type=kvm
config_mode=multinode
release_name=Havana on CentOS 6.4
net_provider=neutron
net_segment_type=gre
node_count=2
debug=True
settings:{"volumes_lvm": false, "volumes_ceph": true, "images_ceph": true, "ephemeral_ceph": false,
 "objects_ceph": true, "osd_pool_size": 1, "savanna": true}

2 nodes: controller- kvm, compute- hardware machine
controller_role=controller
compute_role=compute ceph-osd

interfaces={"eth0":["admin"], "eth1": ["public", "management", "storage"]}

networks:{"public":{"network_size": 16, "netmask": "255.255.255.240", "ip_ranges":[["172.18.92.84", "172.18.92.86"]],
 "vlan_start": 1080, "cidr": "172.18.92.80/28", "gateway":"172.18.92.81"},
 "management":{"vlan_start": 1076},
 "storage":{"vlan_start": 1077},
 "net04_ext_L3":{"floating": ["172.18.92.87", "172.18.92.94"], "cidr": "172.18.92.80/28", "gateway":"172.18.92.81"}}

When deploying OpenStack environment
 to kvm CentOS distribution is installed with errors
 to hardware machine CentOS distribution not installed(this machine have bootstrap console)

~~~~~~~~~~~~~~~~~~~~~~~~~important~~~~~~~~~~~~~~~~~~~~~~~
If I deploy OpenStack environment on same machines with same parameters, but release_name=Havana on Ubuntu 6.4,
Deployment of environment is done

This bug reproduced only when replacing the subnet and only on CentOS distribution.
Network Settings changed in options window when installing Fuel Master machine.

This last screenshot with controller node. Then screen has not changed.

> This bug reproduced only when replacing the subnet and only on CentOS distribution.
Network Settings changed in options window when installing Fuel Master machine.

Vadim, can you please give some more details, what would be exact steps to reproduce? Now it's hard to understand from many comments in the ticket, what are the actual minimal steps to reproduce it.

Last diagnostic snapshot has exception in anaconda log:
2014-04-01T15:35:04.531220+00:00 info: running "ssh-keygen -q -t rsa1 -f /etc/ssh/ssh_host_key -C -N "
2014-04-01T15:35:05.585403+00:00 info: Running kickstart %%traceback script(s)
2014-04-01T15:35:05.585403+00:00 info: All kickstart %%traceback script(s) have been run
2014-04-01T15:35:05.585473+00:00 crit: anaconda 13.21.195 exception report
2014-04-01T15:35:05.585473+00:00 info: Traceback (most recent call first):
2014-04-01T15:35:05.585480+00:00 info: File "/usr/lib/anaconda/storage/devicelibs/lvm.py", line 414, in lvactivate
2014-04-01T15:35:05.585480+00:00 info: raise LVMError("lvactivate failed for %s: %s" % (lv_name, msg))
2014-04-01T15:35:05.585761+00:00 info: File "/usr/lib/anaconda/storage/devices.py", line 2502, in setup
2014-04-01T15:35:05.585761+00:00 info: lvm.lvactivate(self.vg.name, self._name)
2014-04-01T15:35:05.585768+00:00 info: File "/usr/lib/anaconda/storage/devicetree.py", line 1611, in handleVgLvs
2014-04-01T15:35:05.586099+00:00 info: lv_device.setup()
2014-04-01T15:35:05.586099+00:00 info: File "/usr/lib/anaconda/storage/devicetree.py", line 2068, in _setupLvs
2014-04-01T15:35:05.586107+00:00 info: if self.handleVgLvs(device):
2014-04-01T15:35:05.586319+00:00 info: File "/usr/lib/anaconda/storage/devicetree.py", line 2162, in populate
2014-04-01T15:35:05.586319+00:00 info: if self._setupLvs():
2014-04-01T15:35:05.586324+00:00 info: File "/usr/lib/anaconda/storage/__init__.py", line 451, in reset
2014-04-01T15:35:05.586324+00:00 info: self.devicetree.populate(prog)
2014-04-01T15:35:05.586529+00:00 info: File "/usr/lib/anaconda/storage/__init__.py", line 108, in storageInitialize
2014-04-01T15:35:05.586529+00:00 info: storage.reset(examine_all=examine_all)
2014-04-01T15:35:05.586534+00:00 info: File "/usr/bin/anaconda", line 1114, in <module>
2014-04-01T15:35:05.586890+00:00 info: storage.storageInitialize(anaconda, examine_all=False)
2014-04-01T15:35:05.586890+00:00 info: LVMError: lvactivate failed for nova: Unable to resume vm-nova (253:0)

Any idea why it may happen?

Looks like we really need to reproduce this bug on bare metal. There might be a few things which can cause this exception and failure to happen:
a) hardware issue (disks did not initialize)
b) bug in kernel
c) lvm utility bug

So we need to close as Incomplete for now. If it happens next time, please try to issue lvactivate command manually in the console, also please contact folks in IRC #fuel-dev, so we can group together and resolve it collaboratively.

I couldn't reproduce this.

Please explain exactly what are you doing?
Is it on virtual machines or on hardware?
What settings did you set?
Do you get this error permanently or from time to time?

There were no details / activity for a while. I have to close it as Invalid for now. Please, if you experience an issue again, reopen this bug, and provide all required logs. Ideally, if you could ping folks in #fuel-dev IRC and provide access to your env.