I don't understand this, why no dual 60/80ghz with 5ghz failover products?
The 60ghz LHG radios are FANTASTIC, I really, really like them due to the beamforming. You just chuck them in and hey presto you have a 2gbit/s link, I don't even really bother to align them they 'just work', right up until it rains then it all goes to shit. Hence we install separate 5ghz radios alongside them with BFD for failover redundancy. But in my opinion MikroTik has the best possible marketshare for 60ghz short range links if they added an integrated 5ghz radio alongside it. That way we don't need extra mast space, we don't need additional equipment. Heck even if the failover isn't automatic, the beauty is that MikroTik radio's are full routers, so we can implement OSPF+BFD directly on that link, with additional scripting to pre-emptively drop the link for even more control and failover

Seriously, where are the dual band high throughput wireless devices? Is MikroTik allergic to making money? Because some of our links we spend 20x as much money on other vendors just to support this 1 feature. We would GLADLY pay triple the price for a device that can combine the 2
MikroTik: Get off your ass and make this product. I don't even care if you have no automatic failover technology. Though if you do come up with a new protocol that can automatically negotiate link changes before the link drops, even better

Exactly my same opinion. MK is like "let's make hyper sport car.. and wait..... don't put wheels on it"

Yep, lots of really good products and really good ideas that are only halfway finished. It's like their R&D department do nothing until 4:30pm on a friday then madly come up with something before 5pm beers

- 36/72 core router, with no support for MetaRouter/Virtualization and almost everything is single threaded..........
- Develop their own MME wireless routing protocol that is entirely useless
- CRS3xx series, missing most of the useful industry standard switch features
- PoE switches.... that only support ModeB (SO STUPID!!!!!)
- Plenty more examples I couldn't be bothered typing, it makes my head hurt

Seriously, where are the dual band high throughput wireless devices?

It is not so easy to make a dual band (5/60 GHz) wireless device that shares the same antenna for both bands. That is also why most other manufacturer's don't have it either.
(when you think they have it, carefully study the datasheet to see if you can find the catch!)
When you need a separate antenna anyways, it is not so much more difficult to have an existing second device instead of the combination.
Now, I could see the usefulness of a 60GHz device with TWO ethernet ports and pass-through PoE, so you can easily hook them together and feed down a single cable.
Maybe even some pre-configured or easy-to-configure (quickset) failover configuration that assumes the second ethernet port connects to the failover link.

2 devices is vastly inferior to 1 combined device
- Need to run extra cables (PoE passthrough does not help if they are different voltages)
- More config on switches/routers with more ports or VLANs, extra IP addresses
- Need to pay a lot more attention to loop prevention
- Need more physical mast space which is often not easy
- Aligning 2 radio's instead of 1
- Can be some agreement issues, i.e. "you are allowed 3 radios on this building"
- 2 devices cannot talk to each other and let each other know what their RSSI/SNR/MCS values are etc. If its on the same device it can read these so you can adjust failover, not just rely on it being completely down

On the last point, I wish MikroTik would hurry up and allow SNMP values to be pulled and stored in variables, that would GREATLY help with this and we could script it in. Or better yet a new industry standard protocol gets developed that is designed specifically for wireless bridging that communicates values, lets them adjust where and how traffic flows by reporting packet loss/latency/bandwidth reporting, unequal cost multi-pathing, failover before the link actually fails, CS5/6/7 priority traffic gets transmitted on both for best reliability and first successful packet is accepted on remote side etc

Antenna does not need to be perfect. If we are only going ~1km we do not need the dish to be tuned perfectly for 5ghz so it can do 8km....

I agree that these should be more capability for link-quality based routing, but I do not restrict that to only failover on a single link.
We are operating a larger network of sites interconnected by various wireless links and sometimes tunnels over internet, and we face the problem that the available routing protocols do not incorporate link quality metrics in the routing decision.
We are using BGP and in practice it only decides on "number of hops". When 3 sites each have links between eachother but e.g. the link from A to C is much longer and lower in quality than A-B and B-C, it may be best to send traffic from A to C via B unless one of those links (A-B or B-C) fails. It can be kludged by setting extra parameters in routing filters, but it is a static solution that does not take variation in the link quality into account.
It amazes me that a company for which this situation is core-business (routed sites with wireless links) does not have a better solution for this problem like a customized link protocol or some automated way to vary a link quality metric based on CCQ etc.
(BGP would not be really suited for this because it does not have such fine-grained link metrics, but OSPF would, and there is no support for it either)

With such a more advanced routing protocol, 60 GHz and 5 GHz functions of the same link could be incorporated as well. It would not be a hard failover.

It is not so easy to make a dual band (5/60 GHz) wireless device that shares the same antenna for both bands. That is also why most other manufacturer's don't have it either.

Actually, they have. The airFiber 60 has 60Ghz with 5 Ghz backup. Works great!

IgniteNet has had one out for years, if you can put up with the software issues.

It is not so easy to make a dual band (5/60 GHz) wireless device that shares the same antenna for both bands. That is also why most other manufacturer's don't have it either.

Actually, they have. The airFiber 60 has 60Ghz with 5 Ghz backup. Works great!

You should not have cut away my remark: (when you think they have it, carefully study the datasheet to see if you can find the catch!)

You should not have cut away my remark: (when you think they have it, carefully study the datasheet to see if you can find the catch!)

I've examined the datasheet, but no 'catch' is jumping out at me. We've deployed about 12 links in Manhattan and they work great. Is there a ticking time bomb in them that fail after 3 months??

You should not have cut away my remark: (when you think they have it, carefully study the datasheet to see if you can find the catch!)

I've examined the datasheet, but no 'catch' is jumping out at me. We've deployed about 12 links in Manhattan and they work great. Is there a ticking time bomb in them that fail after 3 months??

It does not share the same antenna for both bands! It is a parabolic dish for 60 GHz and a small patch for 5 GHz.
When you want the fallback for cases where the link attenuation suddenly is high, it will not work because it has only very little gain on 5 GHz.
Of course MikroTik could make a similar thing by putting a small antenna on the back of the 60 GHz antenna, and maybe they will (the newer PCB has unpopulated space for a 5 GHz radio).
But it is nothing like a built-in-LHG5 would be. It is way less than an SXTsq.

It does not share the same antenna for both bands! It is a parabolic dish for 60 GHz and a small patch for 5 GHz.
When you want the fallback for cases where the link attenuation suddenly is high, it will not work because it has only very little gain on 5 GHz.
Of course MikroTik could make a similar thing by putting a small antenna on the back of the 60 GHz antenna, and maybe they will (the newer PCB has unpopulated space for a 5 GHz radio).
But it is nothing like a built-in-LHG5 would be. It is way less than an SXTsq.

I don't see anything on the AF60 datasheet about a patch antenna, it only reports the 5Ghz antenna gain at 11 dBi. This is much less than single band antennas, but these links are short distances. The built in antenna saves a TON of money when you pay rooftop fees PER antenna. For us the built in backup has worked great while saving a TON of money. We've swapped out many LHG60G links even though they work great.

Yes it doesnt NEED to be high gain

Who would install a 60+5ghz radio and want 30db of gain? So you can make a link go 8km on 5ghz only and never use the 60ghz? that would be stupid
It doesn't need much gain, the PRIMARY purpose is the 60ghz radio, so as long as the 5ghz can do the same distance to serve only as a backup, mission accomplished, it's perfect, doesnt need any more. 1.6km is nothing for 5ghz, very easily attainable with a small or substandard antenna

The reason is that gain is inversely proportional to opening angle of the antenna, so when you use an antenna with low gain you get a lot more interference from other users of the spectrum.
That is especially important on 5 GHz where there are a lot of users, and you want a small opening angle to get only the signal from your peer, and not all the others.
So gain is not only important to cover large signals at good signal strength, it is also important to provide the selectivity you need when receiving your peer.
(and using high gain with proportionally lower tx power is also good citizenship to other users, as you do not disperse your signal over an area that does not need it)

Again you are missing the point that it is a 'backup'
Don't need super long range, or particular high performance. 99.99% of the time it is not being used at all
It's entire purpose in life is to provide connectivity when there otherwise is none, it's not to be an awesome standalone 5ghz product that comes with a 'bonus' 60ghz......... You are way overthinking this