It is useless to test a diode with an Ohm meter, the range does not offer sufficient current to FWD bias it.
This is why there is a diode test range that actually imposes a voltage in order to Fwd bias it and measures the volt drop across the diode.

ALL of my ohm meters give different readings for the different connections to a diode. That includes both analog and digital meters. Andif the device is not a diode they will give a similar reading in both directions.
So an ohm meter test will be able to show if some component is NOT a diode in that manner.

 

On my Fluke and other various meter I have used, the resistance range with a diode shows identical each way, so that theory does not hold up.
If I am in doubt, I use the Diode test range!

 

So, out of curiosity I did some testing on 1N4148 signal diode. All the meters I have close to hand do indicate a measurable resistance when forward biased and an overload when reversed.

All measurements in Ω function, with auto ranging. Note that manual range selection changed the value causing lower readings in lower ranges. Measurement of the voltage used in the ranges shows an substantially higher voltages at lower ranges. For example the Siglent used ~2V in MΩ but ~4.9V in KΩ. I chose to use auto ranging since that is the naïve option.

Meter	Bias	Indicated Value
Fluke 287	forward	273KΩ
	reverse	OL
Fluke 101	forward	.692MΩ
	reverse	O.L
Voltcraft VC-330	forward	1.814MΩ
	reverse	O.L
Siglent SDM 3055 (two wire)	forward	.273MΩ
	reverse	overload
Siglent SDM 3055 (four wire)	forward	.273MΩ
	reverse	overload

 

I chose to use auto ranging since that is the naïve option.

I prefer the more worldly one!

 

So it looks like the Siglent and the Fluke both agree on 273KΩ

 

Looks like a computer fan, or some other basic fan system. The more simple setups use thermistors, the more complicated or finer/accurate controls use diode.

What PCB does it attach to? The components on that PCB will likely tell the story of if diode or thermistor.

 

So it looks like the Siglent and the Fluke both agree on 273KΩ

Haven't measured but I assume they are using the same voltage for the range.

 

But going back to the basic issue, Why would you use a meter on the R range to check a diode??