How to determine vendor-class with OpenWrt

Is there an easy way to find out the vendor class of a device under OpenWrt if there is no documentation for the device?
I Found some documentation how to use a vendor-id / vendor-class in OpenWrt. But nothing about how to find out, what a device is requesting.
Only this: https://chrisjrob.com/2009/04/14/determine-vendor-class-identifier/
A dhcp3 server is used there. But Openwrt uses dnsmasq ...

So, what's the OpenWrt way?

https://openwrt.org/docs/guide-user/network/protocol.dhcp#troubleshooting

Ok, thank you for the link.

I did understand how to use tcpdump to show whats going on.

But I did not understand " DHCP client scripts".

First a script /etc/udhcpc.user is created. The script tells the logger to "tag" logged data with some miracle data. Ok so far.
But by what mechanism this script is called? Why this name? No hint...
After digging in the udhcp doc I found a default script is called by udhcp if no extra parms are given. So lets have a look at the default location.
Oh yeah, its there. And wow it calles /etc/udhcp.user.

Now I ask myself if I have to make the script /etc/udhcp.user executable too.
Since if I like to run a script it has to be executable...

Why not one line in the doc to make this clear for everyone ...?

So lets go on.
If I start then wan interfaces as described I got results. But a question remains.
where comes the "dhcp.script" value from?

At least I restarted the LAN interface. Since I was looking for Info's on the LAN but not the WAN side. But I got nothing :frowning:

No, it's being sourced:

Sounds like you are new to OpenWrt. :smiley:
It's normal here to find and read the sources for advanced use cases.

From the parent script which sources the user script.

Start tcpdump with the same options and reconnect your client.

1 Like

Yes, I'm a old man, but a absolute OpenWrt newbie.

Sounds like a friend of mine. He is retired also, but has worked as mainframe developer for many years.

Why write documentation?
The ony real and up to date doc is the source itself.
It's clear, readable by everyone and without any difference between the doc an the running program.

But for me it's a big difference between reading and understandig :frowning:

1 Like

This topic was automatically closed 10 days after the last reply. New replies are no longer allowed.