[ovirt-devel] ovirt 3.5 Test day 1 - vdsm-tool configure libvirt with python code

[ybronhei]

On 07/03/2014 06:11 PM, Yedidyah Bar David wrote:
> ----- Original Message -----
>> From: "ybronhei" <ybronhei at redhat.com>
>> To: "Yedidyah Bar David" <didi at redhat.com>, devel at ovirt.org
>> Sent: Thursday, July 3, 2014 6:05:52 PM
>> Subject: Re: [ovirt-devel] ovirt 3.5 Test day 1 - vdsm-tool configure libvirt with python code
>>
>> On 07/01/2014 05:13 PM, Yedidyah Bar David wrote:
>>> Hi all,
>>>
>>> I was assigned to test [1], which was fixed by [2], which pointed
>>> at [3].
>>>
>>> Most things worked as expected.
>>>
>>> Issues I noticed:
>>>
>>> * the table says that vdsClient with or without '-s' should work against
>>> vdsm with ssl=true or ssl=false. In my tests '-s' worked with true, without
>>> '-s' worked with false, but the other options didn't work.
>>>
>> "vdsClient -s" means to use secure communication, which will work
>> properly if "ssl=true" is configured in vdsm.conf (btw, if ssl not
>> specified there true is the default).
>>
>> bare in mind, that after changing the conf file you need to restart
>> vdsmd service.
>>
>> when you change vdsm.conf like "ssl=true" to "ssl=false", before running
>> vdsmd again, you need to perform "vdsm-tool configure" to configure all
>> related service to work properly with the new vdsm configuration (in
>> that case, not secured which require libvirtd.conf and qemu.conf update)
>>
>> after vdsm-tool configures libvirtd.conf and qemu.conf accordingly , you
>> can start vdsmd and see that "vdsClient" (without -s) works properly.
>>
>> this works as far as I checked in vdsm 3.5
>
> Not sure I got you right - with the above (edit/configure/restart), is it
> intended that '-s' will work with 'ssl=false'? Because iirc it didn't.
-s means secured, and you set ssl=false. so no, when you set ssl to 
false, vdsClient with -s should not work.
>
> I do not think it's important that it works - if a user wants ssl they should
> obviously do 'ssl=true' and '-s', and if not, 'ssl=false' and no '-s'. The
> only reason I tested this and point out above is that it is mentioned in the
> wiki table that it's supposed to work and it didn't for me (again, if I
> understood everything correctly).

mooli, maybe the wiki misleading in that area. consider to update it
>


-- 
Yaniv Bronhaim.