Old thread back from the dead!
I don't think SolarWinds has the ability to do a comparrison of 2 UnDP readings yet. (Feature request, anyone?)
However, in the case of the OP's issue (and assuming only Dell servers are covered) there it would be better to alert based on the OI 1.3.6.1.4.1.674.10892.1.700.20.1.5 which is the temperatureProbeStatus and will return one of the following status codes:
other (1),
-- the status of the object is not one of following:
unknown (2),
-- the status of the object is unknown
-- (not known or monitored)
ok (3),
-- the status of the object is ok
nonCriticalUpper (4),
-- the object is at the non-critical upper limit
criticalUpper (5),
-- the object is at the critical upper limit
nonRecoverableUpper (6),
-- the object is at the non-recoverable upper limit
nonCriticalLower (7),
-- the object is at the non-critical lower limit
criticalLower (8),
-- the object is at the critical lower limit
nonRecoverableLower (9),
-- the object is at the non-recoverable lower limit
failed (10)
-- the status of the object is failed
If you have other UnDP queries, it might be best to start a new discussion with a bit more info on things like hardware invovled, the OIDs you're polling, etc.