On 05/28/2014 04:42 AM, sarath azad wrote:
> Hi Corey,
>
> Thanks for the inputs. I will check the status of the sensors and will
> update by doing "re-armed" where ever required.
> I got the problem in my system. I ran IPMITool to check the status of
> SEL on the system, and it was full/overflowing. Because of this the
> events were not getting detected. After clearing the SEL using
> IPMITool, now able to get the events nicely. 
> Could you please tell if we could get a event if SEL gets full and is
> there any function which could be called in that case to clear the SEL?
>

Ah, I should have thought about that.  Yes, that's probably the most
likely issue.

You will have to manage the event log; you probably don't want to leave
events in there once you have processed them.  You should add an event
handler to the domain with ipmi_domain_add_event_handler().  Then call
ipmi_event_delete() for each event you get.  I'm not sure if you care
about any other events (lots of stuff happens with events, from
intrusion detection to sensors out of range).

-corey

> thanks and regards
> S Sarath 
>
>
> On Tuesday, May 27, 2014 11:47 PM, Corey Minyard <[email protected]> wrote:
>
>
> On 05/27/2014 06:02 AM, sarath azad wrote:
> > Hi Corey,
> >
> > While testing for the discrete event handler for power supply unit (we
> > have a couple of power supplies running in redundant mode), by
> > disconnecting one of them, the event handler function for the discrete
> > sensors, was got called for the first time. But after that we tried
> > the test for many times but the discrete event handler function was
> > not at all getting called.
> > But the status of the corresponding power supply unit does shows "AC
> > Lost" status state set correctly.
> > Could you please tell how to debug this case? How to check why only
> > for the first time event handler was generated but on subsequent tests
> > (after restarting the system) event handler was not at all getting
> > called? Though event handler was not called, but the power supply
> > status was correctly shown.
>
> I'm not quite sure, but some sensors need to be "re-armed" to trigger an
> event again.  If your sensor returns false for
> ipmi_sensor_get_supports_rearm(), you have to call ipmi_sensor_rearm()
> on it to allow it to generate new events.  You will need to do this
> after the sensor returns to normal (the power supply is back in) or you
> will immediately get a new event.
>
> If that's not the issue, you can look at the event log to see if the
> sensor is actually generating an event.  If it is, then something is
> wrong in OpenIPMI.  If not, then it's your hardware/firmware.
>
> -corey
>
>
> >
> > thanks and regards
> > S Sarath
>
>
>


------------------------------------------------------------------------------
Time is money. Stop wasting it! Get your web API in 5 minutes.
www.restlet.com/download
http://p.sf.net/sfu/restlet
_______________________________________________
Openipmi-developer mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/openipmi-developer

Reply via email to