At the risk of oversimplification, the "life" of a bottleneck rifle cartridge will be indicated by case length (after firing and sizing but before reloading) and neck wall thickness.
Bottleneck cases tend to stretch, and the material that moves forward migrates to the case neck. In the process, of course, this weakens the parts of the case that are aft of the shoulder.
Most bottlenecks should be case length measured after every firing or every other firing, and my rule of thumb is two trimmings and then discard.
Case necks that are too thick can be thinned, either by reaming or case neck turning. The former is easier; this step is required if you make cases for wildcats out of something else. (Such as .308 x 1-1/2 Barnes, made out of .308 or .300 Savage, or .33 WCF from .45-70.) In this situation, the neck is too thick because it is now made up of a part of the case that used to be the body, and neck reaming doesn't hurt anything. However, if I found that case necks of factory brass in their factory caliber needed reaming, I'd toss them.
With straight wall cases such as .38 Special and .45 ACP, case failure usually takes the form of cracks in the mouth. Usually, these are the result of working the metal to brittleness through the sizing-belling-crimping cycle. Mouth splits are no big deal (though, obviously, you don't reload split cases).
With bottleneck cases, case failure tends to be either longitudinal splits along the body or head separations. Both can be far more serious than split mouths. So we are a bit more conservative with respect to case life.