[Users] Flush old logs

[Nicolas Ecarnot]

Hi,

This is not a serious issue, just a very small pain in the neck, but I 
can not get rid of this error message :
http://www.zimage.fr/photo.php?id=143945

Amongst 2 other nodes, this node is working very well, it can reboot 
when asked, its network is perfect, it can handle every order I send, it 
has registered perfectly.

I took the time to wait for the monthly flush.
I also reduced the threshold delay and waited for the cleanup to happen.

So, do I have to dig through the database looking for the precise flag 
to update, or is there some more clever way?

Regards,

-- 
Nicolas Ecarnot

Le 22/01/2013 10:41, Nicolas Ecarnot a écrit :
> Hi Eli,
>
> Le 22/01/2013 00:13, Eli Mesika a écrit :
>>> From: "Nicolas Ecarnot" <nicolas at ecarnot.net> To: users at ovirt.org
>>> Here's a simple question : In the "Alerts" message queue at the
>>> bottom of the screen, I see an error message dating from two weeks
>>> (about the failure to verify to restart status of a host). I'm
>>> pretty sure this shouldn't be there anymore, as the power
>>> management of this host has already been tested and approved many
>>> times since.
>>
>> Hi Nicolas The Alert should be deleted from the Alert View once you
>> had succeeded to configure & test it successfully If you only
>> configure it without testing or test failed, only the Alert saying
>> that PM is not configured for the Host will be deleted. The message
>> you had noted " failure to verify to restart status of a host" is not
>> the exact one, can you give please the exact message?
>
> "Failed to verify Host xxxxx Restart status, Please Restart Host xxxxx
> manually."
>
> The tests I did on power management, either via the test button, or via
> some actions, including power up, reboot, are all reacting well.
>
> I'm close to think there is actually NO problem, except this old log msg.
>
>>> I haven't find any trivial way to flush this message queue, and I
>>> was wondering whether it was stored in a database?
>>
>> Yes, it is stored in the database.
>>>
>>> Is there a way to clear these error messages? (Am I right with the
>>> db location?)
>>
>> The table holding all those messages is called audit_log, deleting
>> from this table manually is not recommended. The table has also a
>> auto-clean mechanism controlled by the AuditLogAgingThreashold
>> configuration value that is set by default to 30 days. this value may
>> be modified using the engine-config tool (requires restart of the
>> engine to take place)
>
> Ok, so either I wait 15 more days for this msg to disapear, either I
> find the adequate config command line to shorten it, and restart.
>
> Ok, this is enough for me. Thank you for your explanation.
>
> Regards,
>


-- 
Nicolas Ecarnot