Raspberry Pi listed with 169.25.xx address on 192.168.xx network

[bozolino]

My router's DHCP address space is 192.168.12.xx

Lately, my Raspberry Pi is listed as 169.254.145.95 on my network map, even though it is actually connected at 192.168.12.108. The 169.xx address makes no sense whatsoever, and I wonder where it comes from. Additionally, the Pi's Icon in the network map has a "2" on it, saying "2 clients are connecting thru this device".

Router is RT-AC86U with 384.13 firmware

Can somebody please explain to me what is happening here, and how I can fix it?

 

[yk101]

That looks like a self assigned address. At some point RP wasn’t able to get an IP from DHCP server and self assigned 169 address. Why does it appear in your map? Who knows, but probably a bug in reporting.

 

[htismaqe]

Not exactly sure how the Asus device discovery works but the 169.254.145.95 address is definitely self-assigned. Devices assign themselves an address like that when they can't get a valid DHCP lease. No idea how that would end up in Network Map though since it's local to the host.

As for the "2" it's because that device has two entries in the DHCP lease table. I've seen it many times when I have a live device and change them from DHCP pool to DHCP reservation and their IP address changes. I haven't found a way to delete a lease from the table like I could with DD-WRT, so my solution is to just reboot the router and flush the table.

 

[sbsnb]

Not exactly sure how the Asus device discovery works but the 169.254.145.95 address is definitely self-assigned. Devices assign themselves an address like that when they can't get a valid DHCP lease. No idea how that would end up in Network Map though since it's local to the host.

As far as I can tell, networkmap constructs a list of IP addresses based upon the router's IP, and then proceeds to send an ARP request to every single one of those IPs. The responses are mapped into the table that you see. The only way OP's issue could happen that I can figure is that the Raspberry Pi is responding to ARP requests with the autoconfiguration IP, but that seems far-fetched.

It looks like the starting IP of the scan is hard coded to be the first three bytes of the router's IP, then the last byte starts at 0 and is incremented by 1 with each ARP request until the last byte is 254. I'm not sure why the actual size of the subnet is not calculated based upon the router's LAN configuration.

As for the "2" it's because that device has two entries in the DHCP lease table. I've seen it many times when I have a live device and change them from DHCP pool to DHCP reservation and their IP address changes. I haven't found a way to delete a lease from the table like I could with DD-WRT, so my solution is to just reboot the router and flush the table.

You can also just killall networkmap and then /usr/sbin/networkmap &. The sources also suggest that maybe setting nvram variable networkmap_fullscan=1 might do the trick, but I haven't tried it.