I swapped in '-n 1' for '-T 1' and it ran basically the same and got the same incomplete (only 30 of 33) list.
I tried both together i.e. '-n 1 -T 1' and it was quite slow, relatively speaking (somewhere between 100 and 1000 times slower), so it took 20 seconds, but it caught all 33.
I retried the example from the tutorial to get a better feel for the timing difference
The 14945 and 11966 were complete enough to find the duplicate, but the 3274 wasn't. The (probably) exhaustive one is just too slow to be useful.
So -n 1 is the best choice for this setup, but different workloads may behave differently.
I tried both together i.e. '-n 1 -T 1' and it was quite slow, relatively speaking (somewhere between 100 and 1000 times slower), so it took 20 seconds, but it caught all 33.
I retried the example from the tutorial to get a better feel for the timing difference
Code:
Switches | Hits | Time
----------+-------+-----
<neither> | 3274 | 33s
-n 1 | 11966 | 65s
-T 1 | 14945 | 747s
-T 1 -n 1 | 16058 | 2h34mSo -n 1 is the best choice for this setup, but different workloads may behave differently.