[Nut-upsdev] SNMP on HP UPS

[Arjen de Korte]

Citeren Philip Ward <p.g.ward op stir.ac.uk>:

>> In that case, I fear that you're out of luck. We can't add a modified
>> IETF MIB, since it would break the MIB auto detection (since both the
>> original and modified IETF MIB would use the same OID to check if they
>> are supported).
> I guess I'm stuck building my own modified version for now. We have a
> meeting with HP representatives tomorrow so I'll enjoy berating them
> regarding standards compliance.

In retrospect, HP is probably right here and our current IETF  
implementation is probably wrong. According to RFC 1628, indexed  
objects use a PositiveInteger value, which means it can't be zero:

->     upsInputLineIndex   PositiveInteger,
        upsInputFrequency   NonNegativeInteger,
        upsInputVoltage     NonNegativeInteger,
        upsInputCurrent     NonNegativeInteger,
        upsInputTruePower   NonNegativeInteger

and

->     upsOutputLineIndex   PositiveInteger,
        upsOutputVoltage     NonNegativeInteger,
        upsOutputCurrent     NonNegativeInteger,
        upsOutputPower       NonNegativeInteger,
        upsOutputPercentLoad INTEGER

So most likely, we need to update the mib2nut mapping to reflect this.  
I noticed a similar problem in the mgemib.h (and I wouldn't be  
surprised if this aren't the only ones).

Best regards, Arjen
-- 
Please keep list traffic on the list