Have done some lab testing on RC13 and RB333… Nstreme2 is not really working, it’s never stable and jumps up and down in bw tests. Max result i get is around 20Mbit FD. If i however use the same 2 radios and create 2 seperate WDS links (same boxes with swivel) and do bw tests i get around 60/60 FD and this is really awsome because one box is both recieving and sending in the bwtest, so there’s no dought that RB333 can deliver, its just Nstreme2 that is messing up.
Was it working fine in older versions?
What frequencies are you using?
What is your nstreme configuration?
Also, are you using Nstreme1 when you get 60FD with two single links?
Standard setup, terminate Nstreme2 in Bridge. Bridge holds IP address. Ether1 is also terminated in same Bridge
Have good pings on nstreme2 and can easily transfer 50+ mbit both ways, but whenever i make use of both radios at the same time (trying FD tests) performance drops to almost nothing.
4. Yes, using Nstreme1, best fit Max 4000, polling and disable CSMA. RB1 Terminate both WDS in 2 Bridges holding each one IP adress. Same frequencies as above. The second box terminates both WDS links in same bridge. using 2 different sorces for bwtests to and from RB1
Make sure you send supouts from both ends to support@mikrotik.com. Just a FYI, RC10 using nstream dual and XR5 cards, are getting 35 meg across two ntream2 links at almost 41 miles right now with one of my customers.
I finally after several month i fixed my nstream2 issue with rb333.
The problem, in my opinion is the mini-pci socket. Infact the cards are to close each others. To avoide interference when using nstream2 and rb333 you have to drastically lower the power output of trasmitting card.
Infact after i used a very high separation, 5170 e 5825 and then lowered the power output of the trasmitting card to 10 the link went stable and at the maximum of the modulation.