Hi Mikrotik Experts,
I’ve an issue makes me crazy for almost 4 months now! , I’ll write down the details of my network plus the codes I’m using now.. I really appreciate your help!
{CCR1009-8G-1S-1S+} Connected to 2 internet lines: 20MB & 30MB, 3 Ubiquiti (UNIFI APs) and 1 normal TP-Link access point they are covering decent area of the building without interferance. We have almost 70 devices are connected all the time
Once I implemented the codes, everything was working great, load balancing, hotspot and everything… after few days people start to have issues like internet comes and go every 10 mins in some part of the day!
I was programming the the routerboard to be shutdown everyday night then i have timer cut the electricity for the modems same time, then they start again in the morning through the timer.
Some people told me its not good to shutdown the system everyday, these equipment has been made to work 24 hours!, So i took the timer off then I’ve rested the routerboard and they were working 24/7 just fine
The same issue happened again.. People doesn’t have stable internet
Do you have anything unusual in the logs? Like port flapping or pppoe disconnects? Are you running the latest RouterOS and RouterBOOT versions?
You will have to repeat the problem yourself and use troubleshooting tools like ping and traceroute to detect which device is the culprit. It may be the CCR, wan connections, APs or even the client device.
Thank you so much for your fast reply, so basically only a few out of of 70 connected users has been disconnected for few mins then they connect again but some times it happens many times a day.
So i just restart both the Access points & Server, then it will be oki for some time. But i am not sure where’s the problem first of all? is it the Access points or the Routerband?
For instance, just now 4 users out of the 70 has been disconnected, and i see that everything in the server is just fine except the attached log pic. , then i only restarted the Ubiquiti APs, the it worked just fine!
What RouterOS and RouterBOOT versions are you running?
Judging by the last few lines in your image, it could be a DHCP issue. Is it possible that you have another DHCP server in your network? Perhaps one of the access points may have it and probably giving wrong addresses to some clients before the CCR.
The DHCP Server log lines are normal behavior, based on what we see on our v6.x MikroTik equipment.
First, and foremost, you have two URGENT security problems.
Your router is open to the world for remote login attempts. Hence, all those “login failure for user…” messages in the log. You need to modify your INPUT firewall rules to restrict login access to only trusted IP addresses. It’s very important that you use Safe Mode when making this change, so that you don’t inadvertently lock yourself out of the unit.
The configuration file you posted has your PPPoE settings in plain-text, including the passwords. You need to contact your provider(s) and get those passwords changed ASAP.
I’m inferring from your original post that the Ubiquiti and TP-Link access points are accepting connections from regular computers – laptops & such? Are the logs on the access points showing any wireless disconnections around the time that the connected device loses Internet access? Does the MikroTik show any of the routes as disabled when connected devices are losing Internet access?
Your configuration shows “add bridge=Local-Bridge interface=ether8” twice. While I would expect that the MikroTik should ignore/reject the duplicated entry, that’s probably worth cleaning up.
On our network, we have had problems with DNS caching with a max-udp-packet-size of only 512. We changed our /ip dns settings to max-udp-packet-size of 4096 and have had better performance after that. Prior to the change, we would intermittently have some pages not load (seemed to be concentrated around sites that have DNSSEC implemented).
Judging by the last few lines in your image, it could be a DHCP issue. Is it possible that you have another DHCP server in your network? Perhaps one of the access points may have it and probably giving wrong addresses to some clients before the CCR.
Thanx for replying.. But Actually no its not the problem. it happened before via DHCP in the TP-Link A.. But i’ve disabled it.
Thank you so much DLNoah for your deep observation for the Log!
Your router is open to the world for remote login attempts. Hence, all those “login failure for user…” messages in the log. You need to modify your INPUT firewall rules to restrict login access to only trusted IP addresses. It’s very important that you use Safe Mode when making this change, so that you don’t inadvertently lock yourself out of the unit.
I wanna know more details how to do that!, Ccuz I notice that issue all the time. Iam accessing the network via Computer connected directly to RCC in port 3, also some time I access remotely via http://121.121.10.5/
Your configuration shows “add bridge=Local-Bridge interface=ether8” twice. While I would expect that the MikroTik should ignore/reject the duplicated entry, that’s probably worth cleaning up.
Noted!. Anyhow i wasn’t using this port.
I’m inferring from your original post that the Ubiquiti and TP-Link access points are accepting connections from regular computers – laptops & such? Are the logs on the access points showing any wireless disconnections around the time that the connected device loses Internet access? Does the MikroTik show any of the routes as disabled when connected devices are losing Internet access?
Every user in the network has access to two devices only.. So most of them use Mobile and Laptop/Desktop only.
I also always notice that the number of the devices connected to APs is higher than the active Users whom are using the network, as you can see in the attached pics
One More thing, i dont Understand why there are so many users are in the Active Session but their bandwidth is 0/0 for long time. Do you think is it something related to keepalive timeout As I’ve posted in the original post?
The wiki has a number of articles, such as Firewall Filter Rules, that can help for securing your router.
As referenced by emils, we really can’t tell just from your configuration what’s going on. You’re going to need to replicate the problem and gather troubleshooting information:
What exact error messages or other symptoms are showing up for the client when the Internet is “down”?
When the Internet drops, can the client ping the router? Can the client ping an Internet location by IP (such as 8.8.8.8, Google’s DNS)? If you can’t ping, are you getting “Request timed out”, or a specific error?
When the Internet drops, how much total traffic is going out your WAN connections (are you overloaded)? Are your PPPoE sessions disconnecting? Is one or more ports on your router going down during the disconnects?
Is the connection more reliable if you have the clients get their DNS from a public DNS (such as 8.8.8.8, 8.8.4.4) instead of the router’s DNS cache?
It seems that something is misconfigured on the hotspot interface.
Do you require login by mac? Are you using internal database or radius for mac authentication?
Please export your hotspot settings with
/ip hotspot export hide-sensitive
It seems that something is misconfigured on the hotspot interface.
Do you require login by mac? Are you using internal database or radius for mac authentication?
Please export your hotspot settings with
/ip hotspot export hide-sensitive
Yes, login by mac is enabled for few devices and users.
here’s the print of hotspot, I didn’t copy all the users.
It would be helpful you help me to troubleshoot the problem.
I faced the same issue here.
The same scenario. Like I am using CCR and Ubiquiti. At some time end user is not able to access internet. Even we are not able to ping the user ip address until we remove them from host or reboot the ubiquiti device.
Are you using a Ubiquiti switch for the APs also? In some recent testing I found the UniFi switch didn’t like me having a hotspot bridge connected to it with RSTP enabled, it would administratively block the port until I toggled it physically. So it would show connected but not pass any traffic.
With that said my suggestion would be to try disabling rstp on the hotspot bridge if you only have 1 Port connected back to the rest of the network
