[BUG] Another memory leak gets nuclei killed

[mastercho]

Is there an existing issue for this?

• I have searched the existing issues.

Current Behavior

nuclei spikes CPU and RAM at 100% and gets killed if you scan critical severity list with 30 lines

Expected Behavior

To finish scan without being nuclei killed

Steps To Reproduce

nuclei -l list.txt -silent -t /root/nuclei-templates/ -retries 2 -nmhe -severity critical -rl 150 -o nuclei_output/critical.txt

Relevant log output

I collected profiles with PPROF, i can send in discord or you can download here: https://gofile.io/d/JIFH1S

Environment

- OS: Ubuntu 24.04
- Nuclei: 3.4.10
- Go: 1.25.3

Anything else?

No response

[Mzack9999]

This is reproducible with an http server with very long http responses as all of them are hold in memory

[mastercho]

Its 16GB RAM and AMD EPYC limited to 6 CPU-Cores , 320 NVME SSD Server, if 16GB RAM is not enough then i think issue is somewhere in code

[Mzack9999]

@mastercho would it be possible to test if the PR #6571 fixes the issue?

[mastercho]

go install github.com/projectdiscovery/nuclei/v3/cmd/nuclei@feat-6567-memory-leak
go: github.com/projectdiscovery/nuclei/v3/cmd/nuclei@feat-6567-memory-leak (in github.com/projectdiscovery/nuclei/v3@v3.4.8-0.20251102102607-853687477df3):
        The go.mod file for the module providing named packages contains one or
        more replace directives. It must not contain directives that would cause
        it to be interpreted differently than if it were the main module.

GO wont allow me to install PR and dont want to download and compile from scratch

[dwisiswant0]

GO wont allow me to install PR and dont want to download and compile from scratch

$ git clone https://github.com/projectdiscovery/nuclei && cd nuclei/
$ gh pr checkout 6571
$ make build
$ ./bin/nuclei [...]

[mastercho]

Still being killed, new profiles with PR https://gofile.io/d/SLKN3A

[mastercho]

any updates? As we cant run lists right now and waiting for this fix

[mastercho]

@dwisiswant0 do you need anything from my side to solve that?

[dwisiswant0]

Could you try reproducing the issue on the latest release (v3.6.0)? And run it with the -profile-mem flag so we can get the memory profiling details. That'll help a lot with troubleshooting.