Arne Welzel

I haven't been able to recreate the crash using the buffer as an input for a HTTP request. If you're able to capture a pcap that would be amazing. IIUC,...

Thanks @JordanBarnartt - I've managed to produce the following pcaps that trigger the assert in a debug build. IIUC it needs 1) the originator of a connection using a HTTP...

> Is there any way to narrow down, say, the IP addresses and ports involved in the connection and we can monitor for only that (or at least just a...

> I guess the best we can do is monitor all QUIC traffic, but if there are any other suggestions I'm happy to hear them. Could you share your `./configure`...

> That definitely did the trick in getting debug symbols working. We immediately had a couple of crashes, but they seem unrelated so I opened https://github.com/zeek/zeek/issues/3789. FWIW - you can...

I didn't realize back then probably because I just don't use zeekctl like a normal users :-( But I have seen and forgotten this ticket again. > Not sure what...

> This is not just theoretical, I experienced it in production. > > The issue can be fixed by using a different datastructure for the receive buffer, such as std::deque....

Please provide the full output of your zeekctl session, including the `diag` output of a worker, too. Thanks!

> Eventually the container hangs due to pid starvation. Is this a bug, am I doing something wrong, both? You could try running the container with `--init` to insert a...

> at runtime via an exception. If the shift is by a `const`, it would be nice to get a compiler error at build time. Nice to have, but thought...