Calculating "longevity" is complicated.
For example, say you exercise 1 hr / day, 365 hrs per year.
Now, if you divide that 365 by 12 hour days, that means that you are exercising about 1 month a year. That is quite a bit. You had better enjoy it. So, say you do that over a period of 60 years, that would be about 5 years of exercising. So, the break-even point would be to either increase the lifespan by more than 5 years, or significantly improve the quality of life.
Does exercise "increase" risk of death? Perhaps a little, especially for the weekend iron men. Perhaps no exercise for 6 months, then head out to 120% of one's capacity.
The other thing I would caution against is exercising to the point where it leads to injury, or requires regular use of pain medications.