Basic repeater questions

Hello,

I want to setup a repeater (RB941), connected to my main AP (RB951).
So the clients can hop between the main AP and the repeater without losing connection and the range is extended.

Both devices are factory reset with default config, so according to the tutorial I must do:

Main AP:

/ ip address add address=192.168.0.1/24 interface=wlan1

192.168.0.1/24 changed to 192.168.88.2 (start of dhcp range is increased to .10).

/ interface wireless set wlan1 disabled=no mode=ap-bridge band=5ghz frequency=5180 ssid=Main_gw wds-mode=static

Set my AP as I want, wds-mode=static is important here.

/ interface wireless wds add disabled=no wds-address=XX:XX:XX:XX:XX:X2  master-interface=wlan1

no wds-address= MAC of the repeater AP

/ interface bridge add
/ interface bridge port add interface=wlan1 bridge=bridge1
/ interface bridge port add interface=wds1 bridge=bridge1

I can not add wlan1 to bridge-local and bridge1, so I suppose only wds1 needs to be added to the bridge.

Repeater AP:

/ interface wireless set wlan1 disabled=no mode=ap-bridge band=5ghz frequency=5805 ssid=To_clients wds-mode=static

Set my AP as I want and is the same as the main ap, this is working fine, wds-mode=static is important here.
I want to have the same SSID as the main gateway, so that can be the same?

/ interface wireless wds add disabled=no wds-address=XX:XX:XX:XX:XX:X3  master-interface=wlan1

no wds-address= MAC of the main AP

/ interface bridge add
/ interface bridge port add interface=wlan1 bridge=bridge1
/ interface bridge port add interface=wds1 bridge=bridge1

I can not add wlan1 to bridge-local and bridge1, so I suppose only wds1 needs to be added to the bridge.

/ ip adress add address=192.168.0.3/24 interface=wlan1

192.168.0.1/24 changed to 192.168.88.3 (start of dhcp range is increased to .10).

On the second router I can not ping the main router, neither external IP’s and vice versa.
It is not clear now how it can work, isn’t there a second DHCP server running on the repeater AP?
Do I need to start here with an empty configuration?
I’ve read quite a lot about mesh but it is not clear what to choose and how it works.


Thanks

Forget wlan1 as interface in second device. Take wds and virtual Ap interfaces and bridge them. Why you think that 941 and 951 will work on 5g band? Set different mac address for virtual Ap.

Put addresses on the bridges not on the individual interfaces if they are bridge members.

Remember that the wire is the best wifi you can use. Especially if you are going to make such inefficient repeater.

I know the wire is better and that it is not that efficient as repeater but with 2 mbit downstream… I don’t care.
If you want to drill through 120 cm walls built with rocks, wireless is a beter solution (but only penetrating one wall is possible with Wifi)

In the main AP I added the WPS to bridge-local.
In the repeater AP I added WPS and wlan1 to a seperate bridge.

When I connect, the computer seems to connect always to the repeater (inssider) even is the main AP next to the computer. Is this due the WDS Static setting?

Also no DHCP and internet is available on the repeater so what am I missing?


I just need a working setup for a repeater so users can roam between AP’s.

Did you virtual APs with different mac than their master interfaces?

Virtual AP? Do you mean WDS?

I want to have the same SSID so why do I need to use those VAP’s?

Because I experienced that physical wlan doesn’t work like Ap bridge correctly when hosting wds interface. Don’t use wlan in any bridge and use virtual Ap instead but with different mac address. Ssid and security profile can be the same.

Can we get back to the topic of how to set up the repeater?
I too am trying to set up a repeater to improve signals two floors up in the house (the AP is in the basement). I have a couple of 951s to work with other than the 951 (high power) used as the AP.

The instructions in the online manual are ancient, and don’t line up well with today’s Winbox or Webfig, and I hesitate to try command lines I don’t quite trust. So here’s the simple situation.

  1. There is an AP that works as a regular “home” AP.
  2. There is another 951 that I want to make into a repeater and preferably use as a station bridge too for a local Ethernet port. And I think I once had this working a few years ago but can’t recall the exact settings, which took a long time to get right.
  3. I’ve seen a bunch of instructions, but they never quite all add up. I know I have to create WDS interfaces at both ends and bridge them to the wireless. (I’ve seen conflicting instructions as to whether both should have WDS “MAC” address :00, or should have different addresses.) I got as far as getting the remote unit to be a wds-slave AP that I could connect to, but I couldn’t get it to connect to the real AP even though it had the same frequency, SSID and password.

It is not clear when I should use wds-slave mode or station-wds mode, or something else, on the remote unit.
It is not clear if I need to do anything special on the main AP other than create wds1 and put it on the bridge. wds1 is “not active” “not running” “not slave” on the AP with no WDS remote running.

So can anyone provide “cookbook” instructions for making this work in either GUI? Thanks.

I am on mobile now, will come back later, but the mode you are looking for is Ap bridge. The same frequency and following “radio” settings and the same security profile on both devices. Then the wds will work. When creating virtual Ap, change its mac. And finally bridges are simple.

I am back. Check these rsc files from two rb941-2nd that work in bridged wds mode together with virtual ap on both sides. There are also more things around, like vlans, but you can deal with it, I am sure. If you would like to use it in your setup, take care about changing mac addresses accordingly to your situation.
wds1.txt.rsc
wds2.txt.rsc

Thanks Jarda. Those configurations are fairly complex, though. And the WDS pieces seem inconsistent with some of the instructions I’ve seen. You have a bunch of different made-up MAC-type addresses, for instance, while some (but not all) WDS instructions say that both sides should use the same (00). Which seems odd to me, to be sure. Also, you have static-mesh WDS; others say to use static. I wish there were better documentation on every setting, as well as on how to set things up.

Again, if anybody has “simple” GUI how-to instructions to make WDS work for a repeater, I would probably not be the only person to appreciate it.

At least my configuration works and I am able to repeat it anytime again. Reading what others say but thinking on your own together with making your own tryouts will bring you to the solution.

Jarda, I’m sure it works, but it’s hard to identify what does the WDS and what needs to be copied. Hence my request for a “cookbook” approach, which MikroTik doesn’t seem to have updated in far too long. Looking at your code, with some questions inserted behind #,

add admin-mac=1C:5E:0C:EF:D7:B1 auto-mac=no name=bridge1
add admin-mac=2C:5E:0C:EF:D7:B1 auto-mac=no name=bridge2

Do I need to do that for WDS, or is that specific to something else?

/interface wireless
set [ find default-name=wlan1 ] …
mode=ap-bridge # Both ends use that in your example, not wds-slave or other?
wds-default-bridge=bridge2
wds-ignore-ssid=yes # Is this significant? I thought both ends would just match SSID
wds-mode=static-mesh # Again, why not just static, with only two devices?

/interface wireless wds
add disabled=no l2mtu=1600 master-interface=wlan1 name=wds2 wds-address=
4C:5E:0C:F9:99:30

Is that a made-up MAC address?

/interface wireless
add disabled=no l2mtu=1600 mac-address=4C:5E:0C:EF:D7:B6 master-interface=
wlan1 name=mywifi security-profile=mywifi ssid=mywifi
wds-cost-range=0 wds-default-cost=0 wps-mode=disabled

So that’s a different wireless interface, or what’s its name?

/interface bridge port
add bridge=bridge1 interface=ether1
add bridge=bridge1 interface=ether2
add bridge=bridge1 interface=ether3
add bridge=bridge1 interface=ether4
add bridge=bridge2 interface=vlan2.1
add bridge=bridge1 interface=mywifi
add bridge=bridge2 interface=wds2
add bridge=bridge1 interface=vlan1.wds2

I noticed on my 951 AP that adding the first Ethernet port (upstream) to the bridge ports list caused it to stop working.

And your other end says,

set [ find default-name=wlan1 ] …
mode=ap-bridge # both ends are ap-bridge?
wds-default-bridge=bridge2 # does WDS need its own bridge?
wds-ignore-ssid=yes wds-mode=static-mesh…

Any clarifications are of course welcome. Color me confused. Thanks.

I am on phone mostly so therefore my posts are short. If you send me your Skype name via private message I can get in touch with you directly. We can set the wds link over the teamviewer session and you can create nice explanatory manual with pictures for others after that.

Now to your recent questions:

  1. Administrative mac addresses for bridges: This is quite different topic that has nothing to do with wds itself. If you just create a bridge without administrative mac address, it inherits the lowest address from the belonging running interfaces. It means, it could change as the running interfaces can change their state. As it is not so handy to have changing mac address of the bridge on thy fly, especially if you use dhcp client on the bridge to get IP address from elsewhere, you have an option to set the permanent own mac address to the bridge itself.

When joining wds “uplink” and virtual ap interfaces together, it is the most easy way to use a bridge. Sure, you can route on those interfaces freely, but it is not your goal. Therefore yes, you need a bridge to join the wds and virtual ap together. You can involve other interfaces in it too, if necessary.

  1. Both sides need to be in ap bridge mode, otherwise you cannot make virtual ap interface on such wlan interface.
    Wds-default-bridge is not so important, if you enslave the interface in the bridge manually. If you did not that, the wds interface would enslave itself dynamically to the bridge when it goes to running state.
    wds-ignore-ssid is significant, because the wds link is established according to bssid (mac) of the other side. It is necessary when you have (or will have) more APs and do not want the wds link to jump where they decide to. When I consider wds as wire, I do not want the wire to be connected once here and once there. I rather have my physical topology static. And anyway, as the physical wlan interface is not virtually used (it does not transmitts any ssid, nor is used in any bridge, so it is not possible to connect to it and get any communication with it, there must be additional virtual APs for that, it is not possible to join the wds link according the ssid at all.
    wds-mode is something I am not sure about the explanation. I lost many hours of finding the correctly working combination of all those parameters. Static mash is just a result of these experiments.

  2. Mac addresses on wds links: On both sides put the mac address of opposite wlan interface (the one I told before is not used as ssid transmitting interface :slight_smile:. If the physical wlan interfaces have the same securit profiles and radio parameters (frequency, protocol and so…) they will join by the wds interfaces.

4)4C:5E:0C:EF:D7:B6 is just virtual ap interface created on 4C:5E:0C:EF:D7:B5 physical interface. This one needs to have a ssid and can have different security profile to connect the normal clients to the ap. If you bridge this one with the wds uplink, and on the opposite side the corresponding wds interface will be bridged to the network, this is finally what makes the “repeater” to repeat. Actually, the word “repeater” is not correctly used, because it means “to hear and say exactly what I heard”. You see now, that it does not happen here, as logically it works in the same way like you would connect the distant ap via the cable and bridged the cable with wifi. Just the cable is over the wifi too (wds).
According to my experiments, the virtual ap should have different mac address than its hosting physical wlan.

  1. Adding interface to bridge makes it to stop working. Well, hard to say what is your current situation. Maybe the port is equipped with its own address or with dhcp client that looses the ip assignment when added to the bridge, because the bridge has different mac address… Establish the bridge first, make the ip settings related to the bridge and then add interfaces to it. Slave interfaces (of bridges or switches) should not have their own addresses.

  2. Yes, both ends are AP bridges, as I already explained. Wds does not need its own bridge. Two bridges are there just because of vlans. Each vlan needs its own bridge, but it does not apply to you until you want to have more wifi networks with splitted addressing and not talking together, but physically working on the same wlan interface. Read somewhere something about vlans if you are interested in this topic.

Hope it is clear now and hope you can manage to set your configuration on your own now.

Keeping fingers crossed.

Thank you for those explanations. They look helpful. It’s too bad MikroTik hasn’t created better documentation. I guess that comes in part from the Linux background – there are a gazillion features in Linux, put there by many contributors, few of whom bother with documentation. RouterOS includes them but it doesn’t mean that anyone really understands some of them.

At least that’s how it looks.

You just need more practice. It is not so much complicated how it looks. Even the documentation is quite good if you are able to understand it.