We have spurious overtemp traps on 5140 - hot spot 6.
%Aug 8 13:20:03:301 2024 sw5140 DEV/4/TEMPERATURE_ALARM: -Slot=2; Temperature is greater than the high-temperature alarming threshold on slot 2 sensor hotspot 6.
%Aug 8 13:20:43:300 2024 sw5140 DEV/5/TEMPERATURE_NORMAL: -Slot=2; Temperature changed to normal on slot 2 sensor hotspot 6.
I happened to be able to login at 13:20:55 and disp environment showed the temperature went down already to 51 degrees from 96 or more.
I wonder whether these alarms are useful/realistic or just noise (EDIT: in the latter case it may be a hw issue perhaps).
This short timeframe suggests I would have to take snmp samples quite often if I wanted to get a clearer picture.
Or may be I should have the trap trigger some intensive polling for a minute or so.
This IRF stack of two has this may be once or twice a week. The room is airconditioned at 18°C and there is nothing blocking the fans or so. Only hotspot 6 is affected on each but it does not happen at the same time. The device are still fairly new.
There are no features enabled that would suggest a high cpu load. I have no idea whether hotspot 6 is cpu related.
And we do not see this in the same way elsewhere on on 5140.
Any comments / opinions ?