Here is my setup:
Node: MikroTik hAP
Node: MikroTik LHG-5 Light Head
The LHG is connected to the MikroTik hAP on port 5
My home network is connected to the hAP on the Internet port (1).
I also have a RaspberryPi on port 2.
My hAP is also a tunnel server for 3 other nodes (until we can get our network built out more).
The hAP has a DHCP reservation from my home network.
The RaspberryPi has a DHCP reservation from the hAP.
I upgraded to 3.22.6.0 from 3.19.3.0. After a few days, I upgraded to 3.22.8.0. After the 3.22.8.0, I started to notice some problems with the network the hAP was providing. The RaspberryPi was having problems connecting to the hAP node. I also noticed the tunnels would drop off and the only way to get them to come back is to reboot the hAP. The problem doesn't happen right away, but over a few hours/a couple of days. I would have to say that because I don't access it every day, I don't know the interval....yet.
When I try to ping the hAP node from the RaspberryPi, about 25% of the packets are lost.
When I try to ping the hAP node from my local network, I also see about 25% of the packets are lost.
When I use another 5G node and connect via remote, the LHG-5 also has problems communicating with the hAP. When I telnet into it and ping the hAP, I also see the 25% packet loss.
I have downgraded both nodes firmware to 3.22.6.0. Since the downgrade, I have not noticed any problems (yet).
I have seen some information about the network changing for the Ubiquiti (Rocket, I think) network port. Not sure if this would have any relevance for the hAP.
If I missed any information, it was an oversight; I am trying to give as much of information as possible.
Has anyone seen similar symptoms with their 3.22.8.0?
Thanks and 73,
Matt
NA7RF
Node: MikroTik hAP
Node: MikroTik LHG-5 Light Head
The LHG is connected to the MikroTik hAP on port 5
My home network is connected to the hAP on the Internet port (1).
I also have a RaspberryPi on port 2.
My hAP is also a tunnel server for 3 other nodes (until we can get our network built out more).
The hAP has a DHCP reservation from my home network.
The RaspberryPi has a DHCP reservation from the hAP.
I upgraded to 3.22.6.0 from 3.19.3.0. After a few days, I upgraded to 3.22.8.0. After the 3.22.8.0, I started to notice some problems with the network the hAP was providing. The RaspberryPi was having problems connecting to the hAP node. I also noticed the tunnels would drop off and the only way to get them to come back is to reboot the hAP. The problem doesn't happen right away, but over a few hours/a couple of days. I would have to say that because I don't access it every day, I don't know the interval....yet.
When I try to ping the hAP node from the RaspberryPi, about 25% of the packets are lost.
When I try to ping the hAP node from my local network, I also see about 25% of the packets are lost.
When I use another 5G node and connect via remote, the LHG-5 also has problems communicating with the hAP. When I telnet into it and ping the hAP, I also see the 25% packet loss.
I have downgraded both nodes firmware to 3.22.6.0. Since the downgrade, I have not noticed any problems (yet).
I have seen some information about the network changing for the Ubiquiti (Rocket, I think) network port. Not sure if this would have any relevance for the hAP.
If I missed any information, it was an oversight; I am trying to give as much of information as possible.
Has anyone seen similar symptoms with their 3.22.8.0?
Thanks and 73,
Matt
NA7RF
I have not noticed this.
My setup:
Mikrotik hAP with 3.22.8.0
Port 1 connected to my home network.
2.4 GHz RF AREDN connected to a GL iNet AR300M16 connected to a Grandstream GXP 1630.
Port 5 is DtD connected to a VLAN switch which provides a DtD connection to a
Ubiquiti: two loco M2s, a NanoBridge M9, a Bullet M2, a PowerBeam M5 400, and
a Mikrotik LHG HP5 XL, and
7 active tunnels.
My load average was 0.58, 0.44, 0.39
What is the load average on your hAP when you are losing packets?
73, Chuck
Thanks for the reply. It has taken me some time to perform some tests. Thanks for the insight for the mtr command. I hadn't used that one before.
Before the upgrade, I was getting thing the following. Note, this is mtr to a node that is connected via DTD.
After the upgrade, I got similar numbers but then waited 24 hours. Then I got the follwoing:
I would say 3158 and 88.9% loss is really bad.
The DTD connection was so unstable that for the most part, packets couldn't get to the dish on the hours.
I have since downgraded to 3.22.6, which is the most stable for my hAP. I will give another report after at least 24 hours.
73,
Matt
NA7RF
Thanks for all the help/suggestions. I have applied 3.22.12.0 and everything is working great.
Thanks for the good work on the updates!