Ok, I looked at the DHCP setting and "Use Router as DHCP server" option is ticked.

However I have an "Address reservation" option there to assign IPs manually to MAC addresses. So I assigned IPs to my main pc and my server so I won't have to worry about updating port forwarding rules from now on. I am not worried about other IPs that connect wirelessly so I leave them as they are.



So I should be sorted now, right?

yep that looks fine, shouldn't have any worries on the router side.
if the ip change still persists then its a client issue.

The routing table would have changed from adding the switch so it probably got a tad confused.

Um, how exactly?

There is still only one subnet - the router doesn't care, or even know, how many switches are involved.

it still has a routing table...

if the ips have changed on the clients and the routing table has stale entries in it, the router will use the entry on the table before anything else. A hardware restart usually clears it unless the routes are static.

if its a layer 3 switch it has its own table and will further confuse the situation. if its a router with a four port switch, thats essentially a layer 3 switch as unless the routing functions are turned off, thats exactly what its doing.

Well in my opinion the routing tables will still be exactly the same after adding the extra switch.

Nope if they are using RIP they will start talking to each other as neighbours on the same subnet and updating each others routing tables.

*Edit* you aren't just adding another layer 2 switch, you are adding a router that you are using as a switch... subtle difference.

The product he listed was this, a bog standard layer 2 switch. I assume that's what he used - I can't see any reference to anything else.

In which case, there won't be any change.

Have I missed a reference to a different product?

O.k. my fault

think I was thinking of a similar tread on another forum where an old router was used as a switch, which was giving the above issues.
so its 2 clients one non managed switch.

Also I think from the posts, there is no no issue to fix. seems that the IPs have been added to the reserved list and everything is talking to each other as intended.

As to why the IPs would have changed. there are still routing tables on the clients, also its possible the dhcp leases just ran out at that time and renewed to different ones. can happen if there have been iphones or other wireless devices connecting whilst the others were off.

Main thing is OP is up and running and initial queries all resolved.

Yeah, that's what I think happened here - when they came back up, they came up in a different order/there were different devices on the network, so they got given a different address.



Deffo.

Well no, 192.168.x.x would be 255.255.0.0 but I'm sure you know that.

I'm sure you've just confused a few people who were perfectly happy before they read that

I resisted commenting (8-p)

I didn't even notice it!

me either