Mathew Heard

What do you want @zhuizhuhaomeng?

@dndx I 100% agree with your points. I think the root of the actual problem is that client_body_timeout is between reads. Which the nginx folks think is OK (I agree)...

If you look at benchmarks of tinc you will quickly find that for many real world workloads the largest user of CPU time is TUN/TAP. I did some work on...

I don't know if io_uring is really worth the effort tbh. I don't have strong data to back this up however. Packet mmap appears to be the fastest way to...

It looks like its related to #2

Basically when the lua context shuts down due to a timeout etc the file-io op is still executing and using lua memory.

Correct, although I'm less keen on re-implementing libsystemd. I'd rather link against the system lib if support is enabled. Who knows what will change in the future and what hackery...

Here is a test which you can replicate easily. It can also be used to replicate the issue on Armbian if the systemd watchdog timeout is set to 10s as...

It's worth noting that the Allwinner H3 has a max watchdog timeout value of 16s. Current systemd main loops are far in excess of this value.

@gsliepen @hg Any idea why the proxy test fails? I can't see why myself. I purposefully did not change the proxy.c dns side of things as it should be a...