I have tested with file transfers from a PC to PC and also with each end router ( Mikrotik Bandwidth test. the CPU gets to about 50% with either way of testing.
Yes, the routes are set up in the mikrotik s on each end, Ie route for lan 192.168.1.x route is 10.10.10.9 which one of the interfaces that connects to the AP on the roof, and all the wireless stuff has ip addresses in 10.10.10.x range. Just a simple route, nothing fancy.
I did find someone else had a post a while back with the similar issue, with link speed fast over just the link, BUT when it ended up going to the next device it slowed down, to ironically 70 mbit…
No I know I am probably not going to get 590 mbit over the wireless, but would be nice. But 75 is a far cry from the air data rate.
My speculation is that there is something going on with what I will call packet bounce. Keep in mind that I used to get at or near 200 mbit on this exact same link, with the exact same hardware when I explain my theory here…
Data gets sent from Side A over wireless, the TX rate is seen to increase on that device, the other end receives that packet, HOWEVER it gets bounced back to the other device, either becuase of bad config, or hardware issue (ROS) The TX and RX rates do not increase here on the far side, becuase any data that only goes through the wireless, never reaches the bridge, so the rates do not increase,. Data was sent back to the original side, again, traffic comes in, then out on wireless only, but becuase it never get to the “bridge” its not shown to increase the traffic rate, TX on the AP side still shows ~75 mbit. The AP sends that data back or “out” on the wireless and the other far client side then recives that packet, sends it to the bridge, the RX rate is shown to go up ~75mbit now.
Why I think this is happening, is becuase if you were to take the 75mbit X3 ( trips across the wireless) you would have ~225 mbit Which is close to where the link was before, and equates to the 3 trips across the wireless.
That’s just a theory, but I have only been upgrading the ROS and Router Board firmware quite often, so I do not know when it exactly happened.
ROS has had an issue in the past on other devices where configuration changes were persistent even after a full device reset,a nd a config restore, where the restore file actually created that same issue again, where as if you were to rebuild the config from scratch, the issue was not to be seen again. This was also experienced by me several years ago, and many people thought I was out to lunch, but then others started to report it.
I will put money on it that there is an issue here with this, and MANY people have experienced it, and they always get the same answers… the CPU on the device, wireless interference, etc. Which are all valid. What I have going on this link, is that it is the same exact hardware, in the same spot. I even asked one of the neighbors to chnage the thier frequency and that actually got me from -60db to -55–56 or so, and a higher signal strength, BUT still limited to the ~75mbit. Occasionally 80 mbit. I have checked for queue rules or anything else that may effect it in this way. The number is not pegged at 75mbit, but that is the most common on tests. I have seen it as low as 64mbit, during daytime when other wireless in the area is being used.
SO I am open to ideas on what to try. If it is my setup, I will change it. But I am at a loss for this one.
What is the expected life of a NetMetal 5? 921UAGS-5SHPacD
