Reliability & Worth?

[DatSure]

I've been reading a lot of the posts on here and I'm starting to become worried about the reliability of UBNT products and whether or not it's worth moving to.

Can someone take a moment of their time to answer the following questions for me and assuage my fears in regards to worth of these systems?

1. Reliability - I've read posts on here stating that downtime for these systems can be averaged out to about 10 minutes per day throughout the year. Why is that true and is that what my customers could really expect out of a system I setup in my area?

2. Worth - I have a 50' tower that I can install on my back porch in my neighborhood. In my area that would easily clear a two story house (which is the average) but would not clear most of our trees which can be between around 100 feet tall. Yet the trees aren't thick like a wooded area and are more dispersed (maybe a tree per yard for example).

So my question is... if I put up this tower and purchase an omnidirectional antennae (let's say a 15 dBi antennae) and use a Bullet2 and then used NanoStation Loco's for client CPE devices with window mounts what kind of performance (reliability, speed, & distance) could I expect to receive?

DSL and Cable service exist in the area but they are expensive and the customer support is terrible. I believe if I can beat them on price (not speed) and customer service I can have a legitimate business model. But, I can only do that if the reliability and expandability of the service is equal to or greater than the competitors' products.

Would a different configuration be better for such an urban area in terms of UBNT products and/or another antennae and how many customers could I support with the omnidirectional? I've read on here that 32 is the magic number, is that the case? I think Canopy offers many more per AP. (for the price though)

Thanks to everyone in advance for their support on these questions. I eagerly look forward to seeing what you all have to say.

- Joshua

[UBNT-Mike.Ford]

Hello DatSure,

First and foremost, please be aware that this is a support problem. This is not a forum where people come to post how well the gear works. The only time you hear something is generally when they are having poblems. The amount of posts here is a very minute number when you consider the sheer volume of units that we ship.

Next:

1) This comes down to the nature of 802.11 wireless and has nothing to do with specific UBNT products. Downtime is average from things such as service calls, inclement weather, interference issues, network downtime and the such, once again pretty much separate from hardware failure of UBNT.

2) This is hard to gauage, as RF does not travel well thorugh trees at all in the 2.4Ghz and 5Ghz area. This will add to downtime especiall when the trees become wet as it will refract the RF signal. Once again separate from the UBNT hardware.

3) For this type of Non-Line of site configuration you may be better off going with 900Mhz gear, as the penetration through trees is much, much less then that of 2.4Ghz and 5Ghz gear. However you also become prone to more interference from unlicensed 900Mhz equipment such as phones, microwave, baby monitors, ect.

4) The magic number for user's per AP comes down to a combinatino of factors. How many clients you have, how much bandwidth you have. 35 is the magic number as we are processor limited on the enclosed units. However you can build yourself a far more powerfull AP with things such as a MikroTik Router board and our Xtreme Range series of MiniPCI cards.

There are a lot of factors that going into making a stable WISP. It can be done but take major planning of every aspect of the game. There is no blanket answer to "Will this work? or Is UBNT gear crap" The hardware is only as reliable as the person operating it.

Thanks,

Mike

[DatSure]

Sorry, I felt a bit of negativity in your reply. I didn't mean anything by my comments and only wanted to state what I read and what could be done about it.

Why do you not create these MicroTik boards and package them yourselves as an overall solution for a WISP? Am I wrong that I thought your products were for WISPs? What market are you targeting if I must buy and build my own products from another company?

Again, no disrespect meant by any of this but I am confused on this point? Thanks for the quick reply.

[UBNT-Mike.Ford]

Hello DatSure,

Our market is the WISP market. 90% of the WISP's out there do not use more then 30-40 clients per AP. In this case our equipment works perfect at a fraction of the cost other companies solutions.

Even with MikroTik based hardware you are not going to get more then 60ish user's on a node without running into hidden node, interference ect.

If you need more then 60, odds are you are going to need a Motorola Canopy system for it to work reliably, but then you are going to spend $10k on the AP and 300 for subscribers.

If you have an AP site with 4 Bullet Radios, Each with 90 degree sectors, you can handle about 140 subscribers. Now you add a dual band setup, running 4 Bullet 2's and 4 Bullets 5's all on the same tower, you increase your subscriber count possibility to 280 ect.

This is the reason we do not produce a MikroTik box, as with proper planning you will save yourself a lot of money and headache.

We simplify the radio side of things and allow you to do all your user management and bandwidth management on the backend.

No negativity was meant in my post, im just passionate about UBNT gear :)

Thanks,

Mike

[WHT]

DatSure...

Question 1. Reliability - I've read posts on here stating that downtime for these systems can be averaged out to about 10 minutes per day throughout the year

Most likely what you are reading is the fade margin factor reliability. A WISP may strive for a 20 dB fade margin, which equates to a worst case daily average of ten minutes per day down time. This is a simple reality of any wireless system, it is not unique to a UBNT product.

Question 2. Worth - what kind of performance (reliability, speed, & distance) could I expect to receive?

You could expect the same performance from any other manufactures' radio of similar specifications.

You said "DSL and Cable service exist in the area but they are expensive ...I believe if I can beat them on price"

Check the prices for low speed DSL. Do you really think you can beat under $30 per month? How will you offer support that would be better than what they have?

"Would a different configuration be better for such an urban area in terms of UBNT products and/or another antennae and how many customers could I support with the omnidirectional? I've read on here that 32 is the magic number, is that the case? I think Canopy offers many more per AP. (for the price though) "

I would never consider a wireless deployment that would compete against an incumbent DSL or cable provider. 30 to 35 users is about the same for just about any other access point, some with more horsepower (and three times the price) might support twice that number.

Since you asked "what kind of performance (reliability, speed, & distance) could I expect to receive?", that tells me you are still about two years away from building out your first profitable WISP deployment.

Some of us could deploy an entire WISP project in less than one month and know it will have the performance predicted. But then...we've been to that rodeo before.

[MaximumISP]

Bottom line to me is reliabilty and the level of service provided
is a direct reflection of the skill of the person setting it up
with the proper planing and knowledge it can be quite impressive

My network which is all Ubiquiti gear has a much higher reliabilty
than another local wisp competitor who is all Canopy

I know because he calls me regularly for asistance and
also recently hired me to replace 4 of his backhaul links
with ubnt gear his main trunk is now UBNT lol!

I am also going head to head with DSL in some locations
( not by choice mind you... I was already there first )
either way I have only lost 3 clients to DSL so far
and continue to pick up the odd new one

Its not due to luck or buying the most expensive equipment

[rodneal]

Usually that is only as good as the guy on the tower and the installer on the ground.
You have to be able to do both well to truly appreciate the eccentricities of this business.
We have used almost every piece of equipment out there and it's an expensive lesson.
I really use a lot of MT and all my radio cards are exclusively UB XRs and SRs. (although I'm phasing out all SR this yr)
We have been rolling the new NS2 and 5 as fast as we can but we have a ton of other brands in the field.
We bring in a few to test and then we take a risk and commit to a product and roll with it. We minimize risk by doing business only with those producers that have been doing business for a while in the field and their RMA policies fit with our expectations.
UB has always stood behind their products.
Also we look at failure rates - 1 out of 20 = ok
1 out of 10 = problem
1 out of 5 = move on or work hard with the producer to fix issues if software - if hardware then they need a different manufacturing partner.
Every producer has had problems - only some - very few - shine in support areas and protect you the consumer by providing replacements quickly and with giving you a lot of grief!!
UB being one of those few.
Remember - these are forums are for mutual education and problem solving - a lot of producers do not have forums or do not interact with them if they do.
Rod

[DatSure]

Thanks to everyone is has replied. I think we got off on the wrong foot as this thread ended up taking on a different meaning than I intended.

The only questions I wanted answered were the uptime problems that other posts have mentioned here. 10 minutes of downtime each day on average all year long seems like a lot (regardless of how well the installer/designer of the system was). So my question is... those of you who are using UBNT equipment have you experienced this much of a problem and if so what can be done to prevent it?

Secondly, it sounds like the only places to really install these wireless networks are in rural locations and even those are starting be overtaken by cable/dsl. Is this a dying business (because 3G and WiMax services are coming out and they truly aren't half bad even at these early stages).

So my question then is... what is UBNT and the WISP industry doing to combat these encroaching technologies? How can we compete in the long run?

[MaximumISP]

The only questions I wanted answered were the uptime problems that other posts have mentioned here. 10 minutes of downtime each day on average all year long seems like a lot (regardless of how well the installer/designer of the system was). So my question is... those of you who are using UBNT equipment have you experienced this much of a problem and if so what can be done to prevent it?

I have had a few problems and they were mainly caused from
new interfearance sources popping up in 2.4ghz
Since I am a fairly well trained RF guy I understand the limitations and nuances of it pretty well
(one cannot undervalue this kind of knowledge in this business)
Without these fundimentals there will be reliability and performance problems
failure to plan is in essence.....planning to fail

If you want a better fundamental understanding of rf and the things I am refering to read this book
http://www.ciscopress.com/authors/bio.asp?a=843C05B7-9AE6-440E-8A32-4AF9FE516CD1
Jack is the one who trained me personally

I guess though to truly compair down time one needs to know how reliable
DSL services actually are a quick google turns up this
http://answers.google.com/answers/threadview/id/707325.html
from that link Id say their averages are not exactly what I would call overly impressive

Some of the things I make use of in the UBNT gear are :
smaller channel widths 5-10mhz
the channel shifting feature ( non standard channels )
Sectorizing of APs
Using dual polaity on my APs and clients ( this is a big advantage )
well planned channel usage and reusage

I am also experimenting with third party firmware to provide a polling mechanism
on the ubnt gear which will help greatly to overcome some of the limitations
Search for my posts on OsWave firmware
But this tesing has not been as positive as I would like yet

Unlicensed 2.4 spectrum is an ever changing enviourment so monitoring
and continual adjustments are a necessary part of providing anything
resembling a reliable service in this band

Lisenced spectrum offers big advantages from interfearance but is also expensive
and difficult to acquire and requires special equipment so its not good for say hotspot use

Secondly, it sounds like the only places to really install these wireless networks are in rural locations and even those are starting be overtaken by cable/dsl. Is this a dying business (because 3G and WiMax services are coming out and they truly aren't half bad even at these early stages).

Rural would definately be the target market but hotspots are also very popular
and in many cases they are pretty much expected to be available now

I dont think its a dying market per say but it may be shrinking
the technologies are also improving at a tremendious pace
so who knows what can happen
I dont think standard wifi is going to disappear anytime soon though


How to compete...provide
Great cusomer service
Be Fast and reliable
Have as Low an install cost as possible
Avoid bandwidth caps or blocking of applications (p2p ect..)

Do not connect clients who cannot achieve reliable signal strengths

[rodneal]

A lot there - this business has a lot of potential.
You have to keep up with the changes!!
Remember - money talks!

[DatSure]

Great post MD! Thanks for the answers and recommendations. Has anyone here tried to compete head to head with cable/dsl services and if so what success rate are you having?

I want to provide the service out of spite of our terrible providers. Thanks again to everyone who has responded.

[WHT]

Secondly, it sounds like the only places to really install these wireless networks are in rural locations and even those are starting be overtaken by cable/dsl. Is this a dying business (because 3G and WiMax services are coming out and they truly aren't half bad even at these early stages).
Wireless was a temporary three or four year solution for areas before DSL came out. $120 per month for 2 Mbps wireless or $30 per month for 3 Mbps DSL or $40 for 15 Mbps FIOS.

Thirty mile range claims for WiMax is disingenuous a company saying their wireless gear will thirty miles.

Trying to compete against an incumbent DSL or cable provider is a failed model..unless said provider suxz so bad.

No, WISPing is far from dead when its properly planned

[pcampbell]

customer service - for cable and verizon, around here non existant

Speed of getting you up and running - could potentially be much faster with wireless

cost - i feel like you could provide service cheaper, although DSL here could be as cheap as 16 bucks. :shock:

my fios is $40. how cheap could you provide service?

[WHT]

So my question is... those of you who are using UBNT equipment have you experienced this much of a problem and if so what can be done to prevent it?
Regardless if you are using UBNT products or $400 (?$) Motorola Canopy stations or $2,500 (?$) Sky Pilot access points, you will still have an average annual estimated downtime of ten minutes per day, where...

* It would cost five times as much to bring that downtime to one minute per day.
* Ten minutes per day is a mathematical assumption based on studies done over the past fifty years.
* Realistically, the downtime is less and usually for a few weeks in the spring and fall, and in the mornings and evenings.
* Proper link planning is essential.

[mike95826]

Serve the unserved, or under served. Find your niche. Marinas, Campgrounds, Hotels, Resorts, Small rural towns they all are possibilities. A small resort town of a couple of thousand on the other side of the mountain would generally not be worth it to the big boys but could put some change in your pocket. I work full time in telecommunications for a utility but I have noticed a few places not far from a large city that are not being served by the normal methods (some are using satellite but that is expensive).
I may put up some equipment on a temporary basis to see if it is possible to get a usable signal into the location. Maybe even donate the service to a school, fire department or community center to get the word out.