dnsx occasionally omits JSON output entries for some domains when querying ~160 domains, resulting in fewer host records than input

[Vulner-6]

When running dnsx against a list of approximately 160 domains, the resulting result.json file sometimes contains fewer entries than the number of input domains—for example, only 155 host records instead of 160. This suggests that 5 domains are missing from the JSON output entirely, even though they were included in the input.

I am using the following command (simplified for clarity; actual domain list contains sensitive customer data and cannot be shared):

dnsx -threads 10 -no-color -a -aaaa -cname -txt -mx -ns -resp -json -l nodns.txt -o ttt.json

[Vulner-6]

┌──(root㉿kali)-[~/Desktop]
└─# dnsx -version

  _             __  __

_| | _ __ ___ \ / /
/ ' || ' \ / | \ /
| (| || | | |_ \ / \
_,||| |||//_/_\

            projectdiscovery.io

[INF] Current Version: 1.2.2

[dogancanbakir]

Thanks for reporting! Is this happening randomly, and for the same inputs?

[Vulner-6]

I conducted approximately five attempts, inputting a total of 159 domains. In the output results, 151 domains appeared twice. When I reduced the number of concurrent threads, the output consistently contained results for 156 domains across three separate runs.

I extracted the 8 missing domains from the output and scanned them individually; in this case, all 8 domains were successfully included in the results.

This indicates that when dnsx performs large-scale DNS scanning and generates output, it occasionally drops certain entries. Reducing the thread count mitigates the issue to some extent but does not fully resolve it.

My output format is JSON. In theory, the result file should retain an entry for every input domain—regardless of whether the domain resolves successfully or not.

[Vulner-6]

Thanks for reporting! Is this happening randomly, and for the same inputs?

I conducted approximately five attempts, inputting a total of 159 domains. In the output results, 151 domains appeared twice. When I reduced the number of concurrent threads, the output consistently contained results for 156 domains across three separate runs.

I extracted the 8 missing domains from the output and scanned them individually; in this case, all 8 domains were successfully included in the results.

This indicates that when dnsx performs large-scale DNS scanning and generates output, it occasionally drops certain entries. Reducing the thread count mitigates the issue to some extent but does not fully resolve it.

My output format is JSON. In theory, the result file should retain an entry for every input domain—regardless of whether the domain resolves successfully or not.