One computer. many Pups

[davids45]

G'day,

I'm curious about DHCP and how it divvies out its addresses to 'devices'.

We've just switched the home network to wifi-via-cable from adsl-ethernet (plus some wifi for phone and tablets). The home network has four pre-wifi HP desktops, each of which now has a USB wifi dongle.

Presently I'm not using a static IP for my main desktop which has more than thirty Pups per hard-drive, any one of which I may boot up.

Is the IP address provided via DHCP when I log in to each Pup likely to be the same?

Does DHCP recognise it is the same computer but with a different OS?

Is the name of the computer given by the OS (e.g. puppypcxxxx) relevant to the given-out IP address?

With wifi, is the IP assigned to the wifi dongle (MAC code?), the computer hardware, or the OS version?

On the new router, I've set the DHCP 'Lease time' to 500 hours but is that worthwhile if the DHCP address keeps changing?

Is there a way to have the computer always receive the same IP address using DHCP without going the static IP route? I am not worried about giving the computers static IPs, just curious about DHCP.

I assume a static IP address for my computer would be that same address for each Pup I start with (so it somewhere recognises the dongle or the computer hardware), but how is DHCP affected, Pup to Pup?

David S.

[rufwoof]

I suspect its MAC based. In my cable modem (Virgin Hub) the admin section includes setting a fixed IP to a particular device (MAC).

Rather than relying upon a fixed IP, a alternative is to use a dynamic domain name service. For a domain name I use ddnsfree, which enables you to associate a fixed domain name such as xxx.ddnsfree.com to a variable IP i.e. my desktop/laptop. For that to work the desktop/laptop reports its IP to the ddnsfree server periodically (you can define the time between each refresh). Under those circumstances it doesn't matter what your IP is, providing whichever ports you use (http or ssh server for instance), anyone using your fixed domain name will see your box - assuming wherever you've connected hasn't blocked inbound traffic towards those/that port. To better ensure against being blocked when out and about I use 443 (usually associated as being https traffic) for my ssh port. Conveniently you can run both a https server and ssh server on the same port 443 as one is inbound originated the other outbound originated, so they don't clash. Even then though some may filter the actual ssh protocol, so not guaranteed. What I've done is set my home server to look for me when/wherever I 'appear' and sets up a reverse sshfs, which is handy for security as the home server can be set to have no inbound originated ssh (more secure), but at the same time still acting as a file server.

I route all browser traffic via ssh (data and dns), so if out and about if after connecting via wifi the ssh keys/host throws out a warning, then that is suggestive of a possible man in middle attack i.e. just disconnect and move on. I use a US based ssh server (hashbang.sh) for that (I'm in the UK). So my ISP (and UK state) can't see what sites I visit, they just see a encrypted data flow (ssh). Sites visited also don't see my IP, but the US ssh servers IP instead.

ddnsfree for dynamic domain name an hashbang.sh for ssh have been excellent from my perspective. Both are free however I do however donate.