Replies: 2 comments · 1 reply
-
Historically the performance to resume used to be really bad - the fuzzer would have spent a considerable amount of the 10 minutes just going through the original queue. |
Beta Was this translation helpful? Give feedback.
-
if you run the most current stable branch then there is the fast resume feature (automatically active) - however this only works if the target is not recompiled. |
Beta Was this translation helpful? Give feedback.
-
I've currently a setup where I'd like to continuously fuzz about 20 targets or so (will likely increase over time)... Currently I basically have a script that has an infinite loop over all targets and fuzzes each with
AFL_EXIT_ON_TIME
set to ~10 minutes. So when nothing is found after 10 minutes, the next target is fuzzed. This seems to work well as it gives targets where more is found more time too.What I'm wondering though, is that advisable to stop & start the fuzzers so often? startup seems to be fast for my use cases (takes a few seconds at most), but I notice the "pending" number is always quite high that way.
I have also set
AFL_AUTORESUME
andAFL_FINAL_SYNC
.Beta Was this translation helpful? Give feedback.
All reactions