# dig
**dig** can be used to make [DNS](./Networking/DNS.md) queries.

🔹️ ``dig [options] @server name type``

``server`` is the nameserver to query *(1.1.1.1)* ``name`` is what it should query *(costiser.ro)* ``type`` is how it should query *(A, NS, MX, …)*

For example, ``dig costiser.ro A`` returns ``costiser.ro. 300 IN A 104.21.51.8``, which means ``costiser.ro``: - has the IP address of ``104.21.51.8`` - it has a cache time of 300s *(the DNS resolver remembers it for 300s, then asks a higher-up nameserver)*

Most of the time you’re only interested in the answer portion of dig’s output, so you can make a ``.digrc`` file in your home folder and include ``+noall +answer`` in it. To view all the output just run dig with ``+all``.

You can use ``+trace`` to simulate a recursive DNS resolver’s requests. It starts from the root nameservers and goes down the DNS hierarchy until it reaches the SOA for that domain.



*****
Fun stuff
---------

jvns made a dns query-er in Ruby [here](https://jvns.ca/blog/2022/11/06/making-a-dns-query-in-ruby-from-scratch/). Logic and code could be adapted to C# or python maybe, will have to see. The gist of it is this: 
- open a UDP socket
- learn what a DNS query looks one *(header and question)*
- make the header *(query ID, the flags, number of questions, answer records, authoritative records, and additional records)*
- make the question *(domain name, query type, query class)* 
- encode domain name to DNS format
- make the query-sending function

And that’s it! You can (should) also make a DNS **parser**, which takes a DNS response and processes it into something actually useful (i.e. domain name -> IP address) It needs to: 
- parse the DNS header
- parse the domain name
- then implement DNS compression
- parse the DNS query
- parse the DNS record