Hi,

We have set up a virtaulAP and meshed it with the master interface. We have noticed a bandwidth drop of 25% when the virtualAP is enabled even though it is inactive. Is this expected? Can we have virtualAP's without a decrease in performance?

Thanks,
Jayen

How do you monitor the "25% drop" of this VirtualAP?

I ran this command: `ssh wired_host cat /dev/zero > /dev/null` (where wired_host is a host connected to the routerboard by wire. the ethernet interface on the routerboard is also meshed). i let it run for ~10 minutes, with the virtualAP enabled and disabled and looked at the graph on the master wireless interface in the web interface. i ran similar tests with rsync in the other direction (sending bytes from a wireless device to the wired_host) with similar results. you can see in the graphs below that when the virtualAP is disabled, the are drops in bandwidth, but with the virtualAP enabled, the bandwidth is constantly fluctuating, resulting in a 25% average decrease.

with virtualAP disabled:


with virtualAP enabled:

tried this with a bridge instead of a mesh. results are a bit more predictable.

i don't know what the drop is every two minutes. maybe some buffering thing with the way i'm stressing the connection. but one can see with the virtualAP that the connection takes longer to recover, and is unable to reach the maximum throughput.

with an enabled inactive virtualAP:


without the virtualAP:

I have the same problem

I decided it. I erased the old configuration and configure a new way. configured identically