On Tue, Nov 1, 2016 at 9:01 AM, Yedidyah Bar David <didi@redhat.com> wrote:
On Tue, Nov 1, 2016 at 1:38 AM, Thing <thing.thing@gmail.com> wrote:
> Hi,
>
> I have installed IPA across 3 nodes.  In order to point the ovirt server at
> the new IPA/DNS servers and to clean up I ran engine-cleanup aiming to delte
> the ovirt setup.   However it seems even though I ran this something,
> "vdsm?" is still running and controlling the networking.

engine-cleanup only cleans up the engine. It does nothing to the hosts.
I am not aware of a similar "vdsm-cleanup" tool.
In principle, you could have moved your hosts to maintenance and removed them
from the engine. No idea what this would tell vdsm to do wrt to your question.
It definitely would not undo ifcfg-* changes.

>
> So down under /etc/sysconfig/network-scripts I see,
>
> =========
> /etc/sysconfig/network-scripts
> [root@ovirt1 network-scripts]# ls -l
> total 256
> -rw-rw-r--. 1 root root   130 Nov  1 10:32 ifcfg-enp0s25
> -rw-r--r--. 1 root root   254 Sep 16  2015 ifcfg-lo
> -rw-rw-r--. 1 root root   252 Nov  1 10:32 ifcfg-ovirtmgmt
> 8><-----
> ==========
>
> So my first Q is why when I run engine-cleanup isnt the networking cleaned
> up?  should I file this as a bugzilla?

See above, it's not in the scope for engine-cleanup.

>
> After that I can see that ifcfg-ovirtmgmt is still controlling DNS,
>
> ==========
> 8><----
> [root@ovirt1 network-scripts]# tail ifcfg-ovirtmgmt
> ONBOOT=yes
> IPADDR=192.168.1.34
> NETMASK=255.255.255.0
> GATEWAY=192.168.1.1
> BOOTPROTO=none
> MTU=1500
> DEFROUTE=yes
> NM_CONTROLLED=no
> IPV6INIT=no
> DNS1=192.168.1.240
> [root@ovirt1 network-scripts]#
> ===========
>
> So I tried to set,
>
> DNS1=192.168.1.31
> DNS2=192.168.1.32
> DNS3=192.168.1.33
>
> rebooted and, no I see,
>
> DNS1=192.168.1.240 again
>
> So I also see that the vdsm service is still running,
>
> ==========
> [root@ovirt1 network-scripts]# systemctl status vdsm-network.service
> ● vdsm-network.service - Virtual Desktop Server Manager network restoration
>    Loaded: loaded (/usr/lib/systemd/system/vdsm-network.service; enabled;
> vendor preset: enabled)
>    Active: active (exited) since Tue 2016-11-01 10:32:51 NZDT; 2h 0min ago
>   Process: 2873 ExecStart=/usr/bin/vdsm-tool restore-nets (code=exited,
> status=0/SUCCESS)
>   Process: 2848 ExecStartPre=/usr/bin/vdsm-tool --vvverbose --append
> --logfile=/var/log/vdsm/upgrade.log upgrade-unified-persistence
> (code=exited, status=0/SUCCESS)
>  Main PID: 2873 (code=exited, status=0/SUCCESS)
>    CGroup: /system.slice/vdsm-network.service
>
> Nov 01 10:32:45 ovirt1.ods.graywitch.co.nz python[2887]: DIGEST-MD5 client
> step 2
> Nov 01 10:32:45 ovirt1.ods.graywitch.co.nz python[2887]: DIGEST-MD5
> parse_server_challenge()
> Nov 01 10:32:45 ovirt1.ods.graywitch.co.nz python[2887]: DIGEST-MD5
> ask_user_info()
> Nov 01 10:32:45 ovirt1.ods.graywitch.co.nz python[2887]: DIGEST-MD5 client
> step 2
> Nov 01 10:32:45 ovirt1.ods.graywitch.co.nz python[2887]: DIGEST-MD5
> ask_user_info()
> Nov 01 10:32:45 ovirt1.ods.graywitch.co.nz python[2887]: DIGEST-MD5
> make_client_response()
> Nov 01 10:32:45 ovirt1.ods.graywitch.co.nz python[2887]: DIGEST-MD5 client
> step 3
> Nov 01 10:32:51 ovirt1.ods.graywitch.co.nz python[2887]: DIGEST-MD5 client
> mech dispose
> Nov 01 10:32:51 ovirt1.ods.graywitch.co.nz python[2887]: DIGEST-MD5 common
> mech dispose
> Nov 01 10:32:51 ovirt1.ods.graywitch.co.nz systemd[1]: Started Virtual
> Desktop Server Manager network restoration.
> [root@ovirt1 network-scripts]#
> ===========
>
> why after cleaning up is this still active?
>
> Next, I have grep'd under /etc/ and cannot find where its getting its
> obsolete network DNS info info from.

I think it's somewhere under /var, adding Dan.

If VDSM is running, any network devices it used are owned by it.
Editing manually ifcfg files it owns will result in VDSM restoring them to their persisted state
in vdsm configuration files.

I guess, you can just uninstall/remove vdsm and continue from there: sudo yum remove -y vdsm\*
 

>
> So I need to know where this info is stored? so I can edit is via the CLI?
> database?
>
> There is no web ui running as engine-cleanup has removed that so I cant work
> via the web ui.
>
> Is there anything else I need to manually stop, disable and remove after
> running engine-cleanup?

See above.

See also [1], which is for a hosted-engine host, not exactly your case
but similar. You are welcome to comment there, and/or open another RFE
bz for general hosts (not hosted-engine ones).

I might be wrong, though - perhaps there is something to undo ifcfg*
changes that I am unaware of. Dan?

There is automatic restoration for rollback scenarios, but these are usually overwritten multiple times
for each successful setup. We just save the last good network setup and in case of failure to a new setup
it is restored back to the good point.
I also do not see to much logic in allowing restoration to the 'before vdsm' period. The network connections
of the hosts could have been changed from that time.
 

[1] https://bugzilla.redhat.com/show_bug.cgi?id=1001181

Best,
--
Didi
_______________________________________________
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users