[ovirt-devel] unable to add host(s)

[Dan Kenigsberg]

On Fri, Sep 29, 2017 at 12:12 PM, Ondra Machacek <omachace at redhat.com> wrote:
> On Wed, Sep 27, 2017 at 11:34 AM, Martin Mucha <mmucha at redhat.com> wrote:
>> Hi,
>>
>> I have some hosts which are resilient to my effort to add them into engine:
>>
>> 1. produces some error about iptables:
>>
>> 2017-09-27 11:45:29,711+0300 DEBUG otopi.plugins.otopi.services.systemd
>> plugin.executeRaw:813 execute: ('/usr/bin/systemctl', 'enable',
>> u'iptables.service'), executable='None', cwd='None', e
>> nv=None
>> 2017-09-27 11:45:29,717+0300 DEBUG otopi.plugins.otopi.services.systemd
>> plugin.executeRaw:863 execute-result: ('/usr/bin/systemctl', 'enable',
>> u'iptables.service'), rc=1
>> 2017-09-27 11:45:29,717+0300 DEBUG otopi.plugins.otopi.services.systemd
>> plugin.execute:921 execute-output: ('/usr/bin/systemctl', 'enable',
>> u'iptables.service') stdout:
>>
>>
>> 2017-09-27 11:45:29,717+0300 DEBUG otopi.plugins.otopi.services.systemd
>> plugin.execute:926 execute-output: ('/usr/bin/systemctl', 'enable',
>> u'iptables.service') stderr:
>> Failed to execute operation: No such file or directory
>>
>> 2017-09-27 11:45:29,718+0300 DEBUG otopi.context context._executeMethod:143
>> method exception
>> Traceback (most recent call last):
>>   File "/tmp/ovirt-5BawtgI21o/pythonlib/otopi/context.py", line 133, in
>> _executeMethod
>>     method['method']()
>>   File "/tmp/ovirt-5BawtgI21o/otopi-plugins/otopi/network/iptables.py", line
>> 108, in _closeup
>>     self.services.startup('iptables', True)
>>   File "/tmp/ovirt-5BawtgI21o/otopi-plugins/otopi/services/systemd.py", line
>> 122, in startup
>>     service=name,
>> RuntimeError: Failed to enable service 'iptables'
>> 2017-09-27 11:45:29,720+0300 ERROR otopi.context context._executeMethod:152
>> Failed to execute stage 'Closing up': Failed to enable service 'iptables'
>>
>>
>> • see
>> ovirt-host-deploy-20170927104530-vm-18-133.eng.lab.tlv.redhat.com-310be768.log
>> in attachment
>> • but perpetual effort and retries seems to fix this, so I did not get this
>> one again.
>>
>>
>> 2. generally hosts tries to install uninstallable cockpit-whatever and other
>> non existant libraries, which I can overcome with
>> /etc/ovirt-host-deploy.conf.d/50-offline-packager.conf, but it's making
>> whole process kinda very slow. Is there a better way?
>>
>> 3. and finally host vm-17-100.eng.lab.tlv.redhat.com responds to same
>> treatment (disabling package manager, multitude of retries to install host
>> and do yum updates before installing host ) differently, and now it always
>> became nonresponsive upon installation, without any information (apparently
>> to me) in logs. Otherwise host function pretty normal, it only cannot be
>> added to engine.
>>
>> check
>> ovirt-host-deploy-20170927105857-vm-17-100.eng.lab.tlv.redhat.com-e01e0dd4-c0cc-499d-8f1f-6d0b12b64804.log
>> in attachment.
>>
>> ————
>>
>> am I doing something wrong? Or is this a bug?
>
> Do you use firewalld cluster? If yes, please check:
>
> https://github.com/oVirt/ovirt-engine/blob/master/README.adoc#host-deploy-via-ansible

+Eitan who bumped into similar case recently.

Ondra, I believe there should be an immediate, synchronous error
message if something as important as host firewall configuration did
not finish successfully. If the host is not usable, the caller of
add_host API should receive an error message.


>
>>
>> If you need any further information, please ask. I can retry any time,
>> possibly with different host.
>>
>> M.
>>
>> _______________________________________________
>> Devel mailing list
>> Devel at ovirt.org
>> http://lists.ovirt.org/mailman/listinfo/devel
> _______________________________________________
> Devel mailing list
> Devel at ovirt.org
> http://lists.ovirt.org/mailman/listinfo/devel