3 ms·
This is what I love about the design and operational philosophy of the Erlang/OTP evironment: the trickier class of harder-to-test/reproduce transient/timing pr
by ReligiousFlames 9y ago
This is what I love about the design and operational philosophy of the Erlang/OTP evironment: the trickier class of harder-to-test/reproduce transient/timing pre-defensive-coding issues can sometimes be resolved automagically under supervision trees restarting said Erlang process. Log the stacktrace as a failure to maybe be fixed and move on. That's in addition to a robust soft-realtime, SMP VM which allows remote, live debugging.
- d6e 9y agoAs a newbie to Erlang, how does Erlang handle this case better for a service than a Python process which logs and restarts on failure? Also, this runtime testing strategy doesn't handle cases where you have other systems depending on that Erlang code path being successful, does it?.