M. Ángel Jimeno

hey pd team @Ice3man543 @forgedhallpass what is the intended behavior? Thanks!

If one of you shares the `dmesg` output to understand why the kernel killed the process and a pprof profile for the process, I could look into this issue. https://stackoverflow.com/a/20672691

Having a memory profile would be super helpful. If you are kind enough to share one + the corresponding dmesg output, ping me and I'll be happy to contribute a...

If only one of you shared the dmesg log and a memory profile of the process, we could build a fix for this bug.

@IfrahIman a memory profile + the exact commit your nuclei was built from would be super helpful. Will have a look at this during the next week but it looks...

Hey @IfrahIman - what @Ice3man543 said plus the output of `$ go version -m `, it should look like this: ```console go1.19 path github.com/projectdiscovery/nuclei/v2/cmd/nuclei mod github.com/projectdiscovery/nuclei/v2 (devel) dep git.mills.io/prologic/smtpd v0.0.0-20210710122116-a525b76c287a...

Okay, there's definitely a goroutine leakage on fileutil package `TimeoutReader` type which is used when nuclei is configured to read from `stdin`: https://sourcegraph.com/github.com/projectdiscovery/nuclei@dev/-/blob/v2/pkg/core/inputs/hybrid/hmap.go?L75-78&subtree=true Here is a small reproducer for the...

You can run nuclei with the `-ep` flag to enable the pprof handlers and then run `$ go tool pprof -png http://localhost:8086/debug/pprof/heap > out_n.png` (8086 port is hardcoded here https://github.com/projectdiscovery/nuclei/blob/1010cca84e62e04cd675debfce20ce96d2e9cd3c/v2/internal/runner/runner.go#L74...

@forgedhallpass - note `httputil` package prints HTTP/1.1-like requests, it doesn't print HTTP/2 requests: > HTTP/2 requests are dumped in HTTP/1.x form, not in their original binary representations. Source: https://cs.opensource.google/go/go/+/refs/tags/go1.19.1:src/net/http/httputil/dump.go;l=208-209 Make...

Same as #2586, in case you want to close it as a dupe @forgedhallpass @ehsandeep