What factor base are you using for Pollard's p - 1? It will work, though you need a fairly large factor base for such a small input number (primes up to 31) unless you are using the two-step variant.
As for the quadratic sieve, it pretty much always works for sufficiently large numbers (including yours) so obviously the implementation you are using is broken. How are you carrying out the quadratic sieve?
To be fair though for a number of this size I'd start by carrying out trial division up to the first ten thousand primes (which would factor it) and then directly moving on to ECM.