Client ID and Hostname
-
The DHCP option 12 says client ID but the clients say host name. Regardless, I would expect whatever is provided by the client with option 12 to be in the Client ID column.
-
Where are you seeing these values mixed up? I checked a few systems here and mine are all correct. Client ID column is empty and hostname is filled in. Same for static and dynamic leases.
-
First off, I never noticed this as I always set the host name on the computer the same as the name I assign in the DHCP static mapping. However, a friend mentioned that an ISPs modem can use that host name and wondered if pfSense could do the same. That got me looking into it when I discovered that the host name from the computer was used in the Hostname column. What's the purpose of the Client ID column, if it doesn't show the name in option 12?
-
It's for the DHCP client ID, which is a completely different thing than the hostname. It's not used nearly as often.
I'm still not sure why we added it so prominently in the GUI when few people need it.
Some ISPs require the client ID to be set a specific way. Some clients can send a client ID but maybe not a hostname, or maybe someone wants to set the client ID different than the hostname but match on the client ID. Hard to say really what people might want or need.
The client ID also has nothing to do with DNS, whereas the hostname can depending on the DNS integration settings.
-
It sounds like maybe you've been filling in the "Client ID" in DHCP static mappings unnecessarily when you meant to use that in the description and/or hostname.
-
Then what is the client ID?
Here's what Cisco says:
DHCP Client Option 12 Overview
Configuration parameters and other control information are carried in tagged data items that are stored in the
options field of a Dynamic Host Configuration Protocol {DHCP) message. The DHCP client provides flexibility
by allowing Option 12 to be configured for a DHCP client.
Option 12 specifies the name of the client. The name might or might not be qualified with the local domain. -
No, I put a real description in the description column. Also, I still have to add the host name for the IPv6 address in DNS.
-
It's the Client ID. I may sound like I'm talking in circles but that's literally what it is. It's a specific ID string the client can send and that the server can match.
Say you don't want to match based on MAC address, maybe so the user gets the same address on wireless and wifi. Then you can match based on client ID instead if it can send one.
-
Well, option 12 matches what's in the /etc/hostname file on Linux and the computer name in Windows. I verified that with Wireshark. Is there some other ID that can be provided?
There's also option 61, but I didn't see that in the captures.
-
What's the problem with that? I fail to see what any problem is here. What are you trying to do that is not working? Or what are you hoping to accomplish that you cannot?
Client ID is optional and most people don't need or want it. It's a value sent by the client for matching by the server. It's not like the hostname which is a property either sent by the client or set by the server for use in DNS.
If you don't know what you're doing with it, leave it blank and ignore it.
-
No problem. I just wonder what that client ID column is for, when the contents of option 12, which is supposedly client ID, is placed in the host name column. Maybe some info could be provided on that page to clarify. The pfSense docs are a bit thin on that.
BTW, I'm the kind of person who likes to really dig into something, to understand it fully.