All of lore.kernel.org
 help / color / mirror / Atom feed
* Redfish HealthRollup/LED question
@ 2020-07-14 11:44 Deepak Kodihalli
  2020-07-14 21:10 ` James Feist
  0 siblings, 1 reply; 4+ messages in thread
From: Deepak Kodihalli @ 2020-07-14 11:44 UTC (permalink / raw)
  To: James Feist, Gunnar Mills, vishwa; +Cc: openbmc

Hi James/Gunnar,

On IBM platforms, a fault indicator LED associated with a FRU can 
indicate the "Health Rollup" status of the FRU. For eg a CPU can be 
healthy, but something contained within (eg a core) may be broken, so 
the LED will be turned on. I think the Redfish HealthRollup property is 
conceptually similar.

Question is - do you see a problem if we add code in bmcweb that sets 
the HealthRollup property for a FRU if an associated LED is turned on? I 
believe this is aligned with 
https://github.com/openbmc/docs/blob/master/designs/redfish-health-rolllup.md 
(no new D-Bus objects/interfaces being added, etc). On IBM platforms, a 
D-Bus association might exist between the FRU and the LED, and setting 
the HealthRollup property will be based on this. If the said association 
is missing on other platforms, the code will not do anything.

Thanks,
Deepak

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Redfish HealthRollup/LED question
  2020-07-14 11:44 Redfish HealthRollup/LED question Deepak Kodihalli
@ 2020-07-14 21:10 ` James Feist
  2020-07-15  8:14   ` Deepak Kodihalli
  0 siblings, 1 reply; 4+ messages in thread
From: James Feist @ 2020-07-14 21:10 UTC (permalink / raw)
  To: Deepak Kodihalli, Gunnar Mills, vishwa; +Cc: openbmc

On 7/14/2020 4:44 AM, Deepak Kodihalli wrote:
> Hi James/Gunnar,
> 
> On IBM platforms, a fault indicator LED associated with a FRU can 
> indicate the "Health Rollup" status of the FRU. For eg a CPU can be 
> healthy, but something contained within (eg a core) may be broken, so 
> the LED will be turned on. I think the Redfish HealthRollup property is 
> conceptually similar.
> 
> Question is - do you see a problem if we add code in bmcweb that sets 
> the HealthRollup property for a FRU if an associated LED is turned on? I 
> believe this is aligned with 
> https://github.com/openbmc/docs/blob/master/designs/redfish-health-rolllup.md 
> (no new D-Bus objects/interfaces being added, etc). On IBM platforms, a 
> D-Bus association might exist between the FRU and the LED, and setting 
> the HealthRollup property will be based on this. If the said association 
> is missing on other platforms, the code will not do anything.

What changes does this require? I believe this is how the health roll up 
functions today.

> 
> Thanks,
> Deepak

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Redfish HealthRollup/LED question
  2020-07-14 21:10 ` James Feist
@ 2020-07-15  8:14   ` Deepak Kodihalli
  2020-07-15 16:20     ` James Feist
  0 siblings, 1 reply; 4+ messages in thread
From: Deepak Kodihalli @ 2020-07-15  8:14 UTC (permalink / raw)
  To: James Feist, Gunnar Mills, vishwa; +Cc: openbmc

On 15/07/20 2:40 am, James Feist wrote:
> On 7/14/2020 4:44 AM, Deepak Kodihalli wrote:
>> Hi James/Gunnar,
>>
>> On IBM platforms, a fault indicator LED associated with a FRU can 
>> indicate the "Health Rollup" status of the FRU. For eg a CPU can be 
>> healthy, but something contained within (eg a core) may be broken, so 
>> the LED will be turned on. I think the Redfish HealthRollup property 
>> is conceptually similar.
>>
>> Question is - do you see a problem if we add code in bmcweb that sets 
>> the HealthRollup property for a FRU if an associated LED is turned on? 
>> I believe this is aligned with 
>> https://github.com/openbmc/docs/blob/master/designs/redfish-health-rolllup.md 
>> (no new D-Bus objects/interfaces being added, etc). On IBM platforms, 
>> a D-Bus association might exist between the FRU and the LED, and 
>> setting the HealthRollup property will be based on this. If the said 
>> association is missing on other platforms, the code will not do anything.
> 
> What changes does this require? I believe this is how the health roll up 
> functions today.

Yes, the bmcweb code around HealthRollup does look up association 
endpoints today already. What I mentioned above might need code changes 
to look at additional association types.

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Redfish HealthRollup/LED question
  2020-07-15  8:14   ` Deepak Kodihalli
@ 2020-07-15 16:20     ` James Feist
  0 siblings, 0 replies; 4+ messages in thread
From: James Feist @ 2020-07-15 16:20 UTC (permalink / raw)
  To: Deepak Kodihalli, Gunnar Mills, vishwa; +Cc: openbmc

On 7/15/2020 1:14 AM, Deepak Kodihalli wrote:
> On 15/07/20 2:40 am, James Feist wrote:
>> On 7/14/2020 4:44 AM, Deepak Kodihalli wrote:
>>> Hi James/Gunnar,
>>>
>>> On IBM platforms, a fault indicator LED associated with a FRU can 
>>> indicate the "Health Rollup" status of the FRU. For eg a CPU can be 
>>> healthy, but something contained within (eg a core) may be broken, so 
>>> the LED will be turned on. I think the Redfish HealthRollup property 
>>> is conceptually similar.
>>>
>>> Question is - do you see a problem if we add code in bmcweb that sets 
>>> the HealthRollup property for a FRU if an associated LED is turned 
>>> on? I believe this is aligned with 
>>> https://github.com/openbmc/docs/blob/master/designs/redfish-health-rolllup.md 
>>> (no new D-Bus objects/interfaces being added, etc). On IBM platforms, 
>>> a D-Bus association might exist between the FRU and the LED, and 
>>> setting the HealthRollup property will be based on this. If the said 
>>> association is missing on other platforms, the code will not do 
>>> anything.
>>
>> What changes does this require? I believe this is how the health roll 
>> up functions today.
> 
> Yes, the bmcweb code around HealthRollup does look up association 
> endpoints today already. What I mentioned above might need code changes 
> to look at additional association types.
> 

I'd prefer we try to be consistent, but if that is not possible 
additional associations are probably fine, as long as the documentation 
is updated.

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2020-07-15 16:20 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-07-14 11:44 Redfish HealthRollup/LED question Deepak Kodihalli
2020-07-14 21:10 ` James Feist
2020-07-15  8:14   ` Deepak Kodihalli
2020-07-15 16:20     ` James Feist

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.