Quick and simple way to check the details on a fiber SFP on a Mikrotik router. Replace sfp1_name with the SFP name or leave out the name and select a number.
Note that there are a couple of limitations of using the Cloud Hosted Router (CHR). The main issue is that the default license doesn’t allow for more than 1Mbps on each interface.
We’ll create a tunnel between two Mikrotik RouterOS routers. Once we have the tunnel connected, we can then route traffic between them.
Note: You can add Preshared keys, but we don’t cover that in this post, just to keep things simple. Check out the following post if you want to add Preshared keys.
Here is how we will want our routers set up. The WireGuard PtP IP is the IP addresses used on both ends of the tunnel. The WAN IP is the IP of each Router. Local IP on Host B is setup to distribute DHCP.
Host A
WAN IP: 172.16.0.1 WireGuard PtP IP: 10.1.1.1/30
Host B
WAN IP: 10.0.0.2 WireGuard PtP IP: 10.1.1.2/30 Local IP: 192.168.0.1/24
We need Host A to be able to access Private IP’s (192.168.0.0/24) behind Host B.
We’ll pretend that the 172.16.0.1 address is a public IP, and Host B, is behind some sort of NAT network.
To create the Point-to-point, or PtP, we will create a WireGuard VPN tunnel, and then add routes from Host A to Host B.
For each Mikrotik we need to create a WireGuard interface, and then a peer. One of the peers needs a keep alive if we are behind a NAT.
Wireguard Setup Overview
Here is an overview screenshot of what our WireGuard settings will look like. Host A is on top, and Host B on the bottom. On the left are the WireGuard interfaces, and the right contains the Peers.
We copy the Public Key from the remote WireGuard interface, to the Public Key on the local Peer. I.e. The Host_B Peer contains Host_A’s Interface Public Key and vice verse
Host A
If you want to, you can use the WinBox GUI to setup and configure the router.
In the Allowed Addresses, put 10.1.1.0/30 and 192.168.0.0/24*.
Finally, put in the Public Key from Host B. Note that we can’t do this until we create the WireGuard Interface on Host B, so you’ll need to come back for this step.
*The Allowed Address sets which addresses work on the other side of the tunnel. If we don’t specify 192.168.0.0/24, then we won’t be able to route to those addresses. If we don’t add 10.1.1.0/30, then our tunnel won’t work at all. Since we only need to route to the 192.168.0.0/24 network from the Host A side, we don’t need this IP range on Host B.
That should be it. Verify that there is a connection. From Host A, ping 192.168.0.1 or any other remote device.
Troubleshooting
Unfortunately, there appear to be some wonky bugs with WireGuard on RouterOS. It does appear to be getting better, but here are a couple things to check if the tunnel is not connecting.
Verify that the Firewall is not blocking WireGuard. You can allow the WireGuard port in the Firewall.
Try disabling and re-enabling the Interfaces and/or Peers
Verify that all the routes for the PtP are in /ip/routes. If not, try manually adding the route (10.1.1.0/30) on the WireGuard interface on both routers.
Add a keep alive if a router is behind a firewall/NAT.
Reboot and or Upgrade the RouterOS version and firmware.
Note: running /user/add will prompt you for the rest of the options.
Delete the default admin user with
/user remove admin
We want to delete the default admin user for two reasons. 1. There is no default password for this user. 2. It is a default username which means it will be targeted for brute force attacks.
Consider using the /users/groups for more granular control.
Disable unused services
In the following, we disabled all services except SSH and Winbox. We also limit access to those services only from private “RFC 1918” IP addresses. Customize as needed.
/ip service
set telnet disabled=yes
set ftp disabled=yes
set www disabled=yes
set www-ssl tls-version=only-1.2
set ssh address="set winbox address="192.168.0.0/16,172.16.0.0/12,10.0.0.0/8"
set api disabled=yes
set winbox address="set winbox address="192.168.0.0/16,172.16.0.0/12,10.0.0.0/8"
set api-ssl disabled=yes tls-version=only-1.2
for www-ssl and api-ssl, tls-version is not a required argument, but you may consider using it if you need the API or Webfig.
Mikrotik Recently patched CVE-2023-37099 which was a way someone with an admin account, could escalate to a “super admin”, or jail break a router.
It appears the technique has been around for about a year.
Affected versions: < 6.49.7
The good news is that someone would already have to have an account to elevate permissions. If your routers have been using strong passwords or SSH public/private keys and have internet management disabled, then you are probably fine.
Setting up OSPF between Mikrotik routers is not too difficult. The following commands should work with RouterOS version 7+. Run these commands on each Mikrotik changing out the router-id.
Create a Loop-back interface
First it would be a good idea to create an loopback interface that will stay up. We’ll use this address as the router-id. This should be unique per router.
First we’ll create the instance. Use the address from the above loopback address. Technically you can use whatever id you want as long as it is a 32 bit “address” and is unique.
IMPORTANT NOTE: If this router is also the default gateway, you’ll need to specify the “originate-default=always” option to share the default gateway over OSPF to the other routers. You don’t have to do this if you don’t want to share the default route.
Create OSPF Area
Now we can create an OSPF area. For a simple OSPF setup, we’ll just use the default 0.0.0.0 area.
Now we can add an instance. This is responsible for what networks get shared with OSPF. If you want to do all the addresses on the router, then use 0.0.0.0/0. If you only want to do specific networks, run an entry for every network, changing 0.0.0.0/0 to the network of interest.
After that we can check to make sure things worked.
/routing/ospf/neighbor/print
You should see at a neighbor. It can take a little bit for the neighbors to show up.
You can also check the routes on the router.
/ip/route/print
OSPF has a default distance of 110, so checking the routes is a quick way to verify the routes are getting updated. Do note that if you have a static route in with a lower distance, that will take precedence over OSPF.
The goal for this script is to alert us if a remote site looses power. We can do this using a Mikrotik that has two PSUs. One is plugged into battery backup and the other in the non battery plug.
In this example, we are using PSU2 “number 8” We can find the number using
/system/health/print
We can now create a new scheduler entry with the following. Change out the number 8 to your PSU number, and change the webhook to your Teams webhook.
:local curState [system/health/get value-name=value number=8]
:local name [/system/identity/get value-name=name]
:local webhook "https://teams.webhook.microsoft.com/webhook/more"
if ($curState != $lastState) do={
if ($curState = "ok") do={
/tool fetch http-method=post http-header-field="Content-Type: application/json" http-data="{\"text\": \"$name : Power is on.\"}" url="$webhook"
}
if ($curState != "ok") do={
/tool fetch http-method=post http-header-field="Content-Type: application/json" http-data="{\"text\": \"$name : Power is off. On battery backup\"}" url="$webhook"
}
:global lastState $curState
}
Set to the appropriate interval (i.e. 5 minutes). The script will only alert once when the power state changes. This minimizes receiving an alert every 5 minutes while the power is off.
The purpose of these scripts is to update the local DHCP lease table with a remote IP Address Management (IPAM) system.
Scheduler Script
This little script is added to the scheduler and goes through the entire DHCP lease table and uploads each MAC address and IP address pair to a website.
Change out the top three variables. May also need to change out the URL depending on how the website receives data.
This script is to be used on the DHCP server script. Can add it by going to DHCP Server -> DHCP -> Double Click Server -> Script
Any time a new DHCP lease is obtained, this script is fired. Note that some of the variables like $leaseBound are specific to the script being used by the “DHCP server”
Also helpful to note that the script only runs if a new lease is obtained, or a lease expires and it disappears from the leases page. A DHCP renew does not trigger the script.
:local username "myapiuser"
:global password "myapipassword"
:global url "upload.incredigeek.com"
# The maximum retries
:local max 60
:local attempts 0
:local success 0
:do {
:set attempts ($attempts+1);
:if ($leaseBound = 0) do {
:do {
/tool fetch url="https://$url/api/v1/network/ipam/dynamic_ip_assignment?ip_address=$leaseActIP&mac_address=$leaseActMAC&expired=1" mode=https keep-result=no user=$username password=$password
:set success 1;
} on-error={
:log error "DHCP FAILED to send unassignment to $url on attempt $attempts out of $max for $leaseActMAC / $leaseActIP";
:delay 10s;
}
} else {
:delay 1s;
# see note below
:local remoteID [/ip dhcp-server lease get [find where address=$leaseActIP] agent-remote-id];
:do {
/tool fetch url="https://$url/api/v1/network/ipam/dynamic_ip_assignment?ip_address=$leaseActIP&mac_address=$leaseActMAC&expired=0" mode=https keep-result=no user=$username password=$password
:set success 1;
} on-error={
:log error "DHCP FAILED to send assignment to $url on attempt $attempts out of $max for $leaseActMAC / $leaseActIP";
:delay 10s;
}
}
:if ($success) do {
:log info "DHCP lease message successfully sent $leaseActMAC / $leaseActIP to $url";
:set attempts $max; # break out of the do..while loop
}
} while ( $attempts < $max )
}
