[ovirt-users] Installation of oVirt 4.1, Gluster Storage and Hosted Engine

Thu Jul 13 12:23:40 UTC 2017

On 07/13/2017 04:30 PM, Simone Marchioni wrote:
> Il 12/07/2017 10:59, knarra ha scritto:
>> On 07/12/2017 01:43 PM, Simone Marchioni wrote:
>>> Il 11/07/2017 11:23, knarra ha scritto:
>>>
>>> Hi,
>>>
>>> reply here to both Gianluca and Kasturi.
>>>
>>> Gianluca: I had ovirt-4.1-dependencies.repo enabled, and gluster 3.8 
>>> packages, but glusterfs-server was missing in my "yum install" 
>>> command, so added glusterfs-server to my installation.
>>>
>>> Kasturi: packages ovirt-hosted-engine-setup, gdeploy and 
>>> cockpit-ovirt-dashboard already installed and updated. vdsm-gluster 
>>> was missing, so added to my installation.
>> okay, cool.
>
> :-)
>
>>>
>>> Rerun deployment and IT WORKED! I can read the message "Succesfully 
>>> deployed Gluster" with the blue button "Continue to Hosted Engine 
>>> Deployment". There's a minor glitch in the window: the green "V" in 
>>> the circle is missing, like there's a missing image (or a wrong 
>>> path, as I had to remove "ansible" from the grafton-sanity-check.sh 
>>> path...)
>> There is a bug for this and it will be fixed soon. Here is the bug id 
>> for your reference. https://bugzilla.redhat.com/show_bug.cgi?id=1462082
>
> Ok, thank you!
>
>>>
>>> Although the deployment worked, and the firewalld and gluterfs 
>>> errors are gone, a couple of errors remains:
>>>
>>>
>>> AFTER VG/LV CREATION, START/STOP/RELOAD/GLUSTER AND FIREWALLD HANDLING:
>>>
>>> PLAY [gluster_servers] 
>>> *********************************************************
>>>
>>> TASK [Run a shell script] 
>>> ******************************************************
>>> fatal: [ha1.domain.it]: FAILED! => {"failed": true, "msg": "The 
>>> conditional check 'result.rc != 0' failed. The error was: error 
>>> while evaluating conditional (result.rc != 0): 'dict object' has no 
>>> attribute 'rc'"}
>>> fatal: [ha2.domain.it]: FAILED! => {"failed": true, "msg": "The 
>>> conditional check 'result.rc != 0' failed. The error was: error 
>>> while evaluating conditional (result.rc != 0): 'dict object' has no 
>>> attribute 'rc'"}
>>> fatal: [ha3.domain.it]: FAILED! => {"failed": true, "msg": "The 
>>> conditional check 'result.rc != 0' failed. The error was: error 
>>> while evaluating conditional (result.rc != 0): 'dict object' has no 
>>> attribute 'rc'"}
>>>     to retry, use: --limit @/tmp/tmpJnz4g3/run-script.retry
>> May be you missed to change the path of the script 
>> "/usr/share/ansible/gdeploy/scripts/disable-gluster-hooks.sh" . That 
>> is why this failure.
>
> You're right: changed the path and now it's ok.
>
>>>
>>> PLAY RECAP 
>>> *********************************************************************
>>> ha1.domain.it            : ok=0    changed=0    unreachable=0 failed=1
>>> ha2.domain.it            : ok=0    changed=0    unreachable=0 failed=1
>>> ha3.domain.it            : ok=0    changed=0    unreachable=0 failed=1
>>>
>>>
>>> PLAY [gluster_servers] 
>>> *********************************************************
>>>
>>> TASK [Run a command in the shell] 
>>> **********************************************
>>> failed: [ha1.domain.it] (item=usermod -a -G gluster qemu) => 
>>> {"changed": true, "cmd": "usermod -a -G gluster qemu", "delta": 
>>> "0:00:00.003144", "end": "2017-07-12 00:22:46.836832", "failed": 
>>> true, "item": "usermod -a -G gluster qemu", "rc": 6, "start": 
>>> "2017-07-12 00:22:46.833688", "stderr": "usermod: group 'gluster' 
>>> does not exist", "stderr_lines": ["usermod: group 'gluster' does not 
>>> exist"], "stdout": "", "stdout_lines": []}
>>> failed: [ha2.domain.it] (item=usermod -a -G gluster qemu) => 
>>> {"changed": true, "cmd": "usermod -a -G gluster qemu", "delta": 
>>> "0:00:00.003647", "end": "2017-07-12 00:22:46.895964", "failed": 
>>> true, "item": "usermod -a -G gluster qemu", "rc": 6, "start": 
>>> "2017-07-12 00:22:46.892317", "stderr": "usermod: group 'gluster' 
>>> does not exist", "stderr_lines": ["usermod: group 'gluster' does not 
>>> exist"], "stdout": "", "stdout_lines": []}
>>> failed: [ha3.domain.it] (item=usermod -a -G gluster qemu) => 
>>> {"changed": true, "cmd": "usermod -a -G gluster qemu", "delta": 
>>> "0:00:00.007008", "end": "2017-07-12 00:22:47.016600", "failed": 
>>> true, "item": "usermod -a -G gluster qemu", "rc": 6, "start": 
>>> "2017-07-12 00:22:47.009592", "stderr": "usermod: group 'gluster' 
>>> does not exist", "stderr_lines": ["usermod: group 'gluster' does not 
>>> exist"], "stdout": "", "stdout_lines": []}
>>>     to retry, use: --limit @/tmp/tmpJnz4g3/shell_cmd.retry
>>>
>>> PLAY RECAP 
>>> *********************************************************************
>>> ha1.domain.it            : ok=0    changed=0    unreachable=0 failed=1
>>> ha2.domain.it            : ok=0    changed=0    unreachable=0 failed=1
>>> ha3.domain.it            : ok=0    changed=0    unreachable=0 failed=1
>> This error can be safely ignored.
>
> Ok
>
>>>
>>>
>>> These are a problem for my installation or can I ignore them?
>> You can just manually run the script to disable hooks on all the 
>> nodes. Other error you can ignore.
>
> Done it
>
>>>
>>> By the way, I'm writing and documenting this process and can prepare 
>>> a tutorial if someone is interested.
>>>
>>> Thank you again for your support: now I'll proceed with the Hosted 
>>> Engine Deployment.
>> Good to know that you can now start with Hosted Engine Deployment.
>
> Started the Hosted Engine Deployment, but I have a different problem now.
>
> As the installer asked, I specified some parameters, in particular a 
> pingable gateway address. Specified the host1 gateway.
> Proceeding with the installer, it requires The Engine VM IP address 
> (DHCP or Static). I selected static and specified an IP Address, but 
> the IP *IS NOT* in the same subnet as the host1. The VMs IP addresses 
> are all on a different subnet.
> The installer shows a red message:
>
> The Engine VM (aa.bb.cc.dd/SM) and the default gateway (ww.xx.yy.zz) 
> will not be in the same IP subnet. Static routing configuration are 
> not supported on automatic VM configuration.
>
> I'm starting to think that BOTH the hosts IPs and the VM IPs MUST BE 
> ON THE SAME SUBNET.
>
> Is this a requirement or there's a way to deal with this configuration?
> It's related only to "automatic VM configuration" or to oVirt in 
> general? Once installed oVirt Engine can I have VMs on different subnet?
Here are some Requirements :

1) Engine VM and the managed hosts should be on management subnet
2) Managed host should be able to resolve EngineVM hostname and 
vice-versa. We configure the engine appliance with DHCP or static 
addressing.
3) Before proceeding with hosted-engine deployment we recommend the user 
to have an FQDN assigned to a mac address  to go via DHCP way.
4) If you are using static then IP address/ netmask /gateway based on 
host ones , and add entries to /etc/hosts if you do not have local DNS.

Hope this helps !!!

Thanks
kasturi
>
> Hi
> Simone
> _______________________________________________
> Users mailing list
> Users at ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users