Re: [WISPA] mikrotik vs ubiquiti

[Vyacheslav Vasilyev]

 I talk about max achievable throughput of 802.11n ,that may be got in ideal
conditions.
It is obvious  that noask gives more throughput then with  ask in ideal (
for example  in lab via coax cable ptp  connection without collisions and
interference ) conditions.
I agree that in noisy environment using of proprietary protocol like Nstreme
may give more stable link and higher throughput  then standard protocol
802.11a/b/g/n  due to it's some useful features, like packet aggregation and
link adaptation. Does Airmax have simular features? Is it only
noask ?  Noask is acceptable when packet losses rate  due to interference
 are very small.
 Throughput is not only one link parameter. Also packets losses rate(BER) ,
delays, jitter,  MOS ( VoIP applications) and others are also important.
Using of  proprieatry protocol in  802.11 a/g/n  units very often help to
improve these  link parameters.
  We carried out field test of  standard and proprietary 802.11n systems in
various LOS and nearLOS conditions.
One of the conclusion thatNstreme ON in 802.11n is not able to improve
link in case multipath fading interference, but Nstreme On in  802.11a
in the same link  really makes connection more stable then in  standard
802.11a/n mode.

Vyacheslav Vasilyev
Unidata

2010/9/6 Scott Carullo 

> I'm not sure your assessment of UBNT not recommending to use airmax on PTP
> as a general statement hold true.  It is unlikely that they would have built
> in a specific PTP noack mode into airmax configuration if it was their
> suggestion not to use it.  I use it on lots of links and it works very well.
>
> You are incorrect in saying airmax and nstream cannot increase throughput
> in comparison with 802.11n.  In the real world in the wild a lot of times it
> is only airmax or nstream that will even let a link perform reliably,
> regardless of what the textbook says. Not to mention it allows our
> throughput to increase in contrast to your statement.
>
>
> Scott Carullo
> Technical Operations
> 877-804-3001 x102
>
>
>
> --
> *From*: "Vyacheslav Vasilyev" 
> *Sent*: Sunday, September 05, 2010 3:50 PM
> *To*: "WISPA General List" 
>
> *Subject*: Re: [WISPA] mikrotik vs ubiquiti
>
>
>
>
> 2010/9/5 Jeromie Reeves 
>
>> On Sun, Sep 5, 2010 at 2:10 AM, Vyacheslav Vasilyev 
>> wrote:
>> >
>> What is a Rockets PPS with airmax on?
>>
> Ubnt does not recomend to use Airmax On in ptp due to  lower performance.
> We did not test  it.
>  Ubnt ,  MT  and any other atheros 802.11n based products  have  aprox
> equal max throughput in  standard 802.11n mode.
> But when airmax and nstreme are ON  they have  different performance.
> Tecnically Airmax is  polling ( round robin algorithm ) like nstreme or
> turbocell/ outdoor router Proxim .   Ubnt polling uses latest Atheros
> chipset clock timing  ( ubnt  calls it "tdma" ) , that may be usefull only
>  in ptmp .
>  Nstreme 2 is  Nstreme 1, that also  uses clock of atheros chipset
> So both  airmax and Nstreme 1,2 can not increase max throughput in ptp in
> comparison with standard 802.11n (hardware atheros aggregation On, 2 chains)
> in ptp in ideal conditions( no interference).
> Nstreme 1,2  is able to improve  802.11n link in comparison with standard
> 802.11n  ( Nstreme Off)  mode in  presence of interference  or/and multipath
> fading  due to it's feature of link parameter adaptation  according packets
> losses rate.
> I  do not  know is  Airmax support link adaptation ( modulation ) or
> not . I suppose -not yet.
>
>   > We tested TDMA freebsd Sam Lefler MAC 802.11a implementation .at the
> simular
> > platform ( Alix, CM9) . It  also has poor  throughput at small packet
> size
> > ( but much better then standard 802,11a)  and it is may be improved  by
> > using more powerfull h/w.
>
>> Ive read up on Sams work and have been very impressed. I looked at the BSD
>> TDMA
>> driver back in late 08 or early 09 (been a while), when I did it was
>> missing glue and needed
>> a bit of polish. It looked like adding GPS sync to it would have only
>> been a matter of getting
>> 2 or more AP's to hold then start on the same signal and keep the
>> timing windows synced.
>>
>
>  Software TDMA  Linux/freebsd  implementation  based on 802.11 chipset
> hardware is separate issue . I  think it may be useful  in ptp and our test
> showed promising results. With regards to ptmp  IMHO it is not viable. There
> is standard fixed TDMA BWA  techhology called fixed wimax 802.16-2004/2009.
> There is  802.16-2004 miniPCI cards - ASIC hardware TDMA implementation .
> There is TDMA 802.16-2004  BS/CPE  Linux based software. For what  a lot of
> people want full software TDMA implemenation?
>  Vyacheslav Vasilyev
>   UNIDATA
>   Fixed  BWA solution
>
>
>
> 
> WISPA Wants You! Join today!
> http://signup.wispa.org/
>
> 
>
> WISPA Wirel

Re: [WISPA] Bullet5M disassembly?

[Robert West]

Anyone wanna see the inside of an AirGrid?!  HA!

Crazy how just 3 little copper pins make up the antenna and can go so
flippin' far!

Joe-


-Original Message-
From: wireless-boun...@wispa.org [mailto:wireless-boun...@wispa.org] On
Behalf Of Robert West
Sent: Saturday, September 04, 2010 9:15 PM
To: 'WISPA General List'
Subject: Re: [WISPA] Bullet5M disassembly?

Not a problem.  I take EVERYTHING apart.  It's a problem, I know..

Bob-



-Original Message-
From: wireless-boun...@wispa.org [mailto:wireless-boun...@wispa.org] On
Behalf Of Greg Ihnen
Sent: Saturday, September 04, 2010 4:13 PM
To: WISPA General List
Subject: Re: [WISPA] Bullet5M disassembly?

Thank you very much!

Greg

On Sep 4, 2010, at 3:16 PM, Robert West wrote:

> Correction, you have to push the N connector IN and pull the RJ45 end OUT!
> I said that backwards...
> 
> 
> 
> -Original Message-
> From: wireless-boun...@wispa.org [mailto:wireless-boun...@wispa.org]
> On Behalf Of Robert West
> Sent: Saturday, September 04, 2010 3:45 PM
> To: 'WISPA General List'
> Subject: Re: [WISPA] Bullet5M disassembly?
> 
> Really simple.  Right at the N connector is a threaded ring.  You'll 
> see the grooves on the top side of it.  Take a pair of channel locks 
> and screw off the ring.  That's what holds the entire thing in.  At 
> the RJ45 connector, you can push in and at the same time pull the N 
> connector out of the housing.
> 
> Bob-
> 
> 
> 
> -Original Message-
> From: wireless-boun...@wispa.org [mailto:wireless-boun...@wispa.org]
> On Behalf Of Greg Ihnen
> Sent: Saturday, September 04, 2010 2:56 PM
> To: WISPA General List
> Subject: [WISPA] Bullet5M disassembly?
> 
> Can anyone tell me how to disassemble the bullet5M? (yes, I googled, 
> searched the UBNT forum, WISPA archive, etc and didn't find anything) 
> I have one that little black ants got into. It's been acting funny and 
> locking up and needing to be reset. I just noticed the ants today and 
> gassed them. The bullet is still running but I think I'd better open 
> it up and clean out any dead ants and any debris they might have left
behind.
> 
> Greg
> 
> 
> --
> --
> 
> WISPA Wants You! Join today!
> http://signup.wispa.org/
> --
> --
> 
> 
> WISPA Wireless List: wireless@wispa.org
> 
> Subscribe/Unsubscribe:
> http://lists.wispa.org/mailman/listinfo/wireless
> 
> Archives: http://lists.wispa.org/pipermail/wireless/
> 
> 
> 
> --
> --
> 
> WISPA Wants You! Join today!
> http://signup.wispa.org/
> --
> --
> 
> 
> WISPA Wireless List: wireless@wispa.org
> 
> Subscribe/Unsubscribe:
> http://lists.wispa.org/mailman/listinfo/wireless
> 
> Archives: http://lists.wispa.org/pipermail/wireless/
> 
> 
> 
> --
> --
> WISPA Wants You! Join today!
> http://signup.wispa.org/
> --
> --
> 
> WISPA Wireless List: wireless@wispa.org
> 
> Subscribe/Unsubscribe:
> http://lists.wispa.org/mailman/listinfo/wireless
> 
> Archives: http://lists.wispa.org/pipermail/wireless/





WISPA Wants You! Join today!
http://signup.wispa.org/


 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/





WISPA Wants You! Join today!
http://signup.wispa.org/


 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/




WISPA Wants You! Join today!
http://signup.wispa.org/

 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/

Re: [WISPA] mikrotik vs ubiquiti

[Robert West]

$300 bucks per day seems damn reasonable considering what it is.

I'd be down with that.

Bob-



-Original Message-
From: wireless-boun...@wispa.org [mailto:wireless-boun...@wispa.org] On
Behalf Of Butch Evans
Sent: Sunday, September 05, 2010 1:40 PM
To: WISPA General List
Subject: Re: [WISPA] mikrotik vs ubiquiti

On Fri, 2010-09-03 at 14:15 -0700, Forbes Mercy wrote: 
> I keep adding filters as traffic presents itself but help and training 
> is very expensive and extraordinarily technical

While I would disagree that training is "very expensive", I would have to
agree that it is very technical in nature.  My training sessions are
normally under $300/day for students (not counting hotels/flights/etc.). 

> On my backhauls
> when one Mikrotik goes down its not unusual for the foul traffic to 
> permeate throughout (yes I'm bridged) the network and take down other 
> Mikrotik's and often requires a drive to reboot then they work fine 
> again, irritating, yes but still great equipment.

Training would be especially good if you could learn something that would
keep you from having to roll a truck even once every 2 weeks.  It wouldn't
take long to pay for that.

> Ubiquiti is a monster for power and throughput, it's menus are basic 
> but filters entry options are slim and limited to IP rather than by 
> protocol so some things sneak through that wouldn't with Mikrotik.

This, unfortunately, is one "cost" of less expensive gear.  FWIW, you have
most of the same functionality available in both platforms, but it's just
not in the GUI for UBNT.

> I promised an analogy so here goes, I feel from experience that 
> Mikrotik is the Linux of equipment, you better know what you're doing 
> when you buy it.

UBNT is linux, too.  :-)

> Ubiquiti is like Windows, pretty GUI driven, and simplified at a 
> reasonable cost.

You have access to iptables and more in the ssh/telnet interface with
Ubiquiti.

--

* Butch Evans   * Professional Network Consultation*
* http://www.butchevans.com/* Network Engineering  *
* http://store.wispgear.net/* Wired or Wireless Networks   *
* http://blog.butchevans.com/   * ImageStream, Mikrotik and MORE!  *






WISPA Wants You! Join today!
http://signup.wispa.org/


 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/




WISPA Wants You! Join today!
http://signup.wispa.org/

 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/

Re: [WISPA] My fun (kidding) weekend on call

[Robert West]

My weekend has been profitable from local hotels not believing me that Time
Warner is having an issue and they all insisting that I make a service call
to tell them just that but in person.  Total take from a walk in, test and
walk out.  Nearly $700 bucks.  Weekend rate, $95 bucks per hour.

 

Thanks time Warner.  You suck.

 

Yes you also rock.  Maybe we could coordinate our efforts sometime.

 

At least it pays the rent.

 

Bob-

 

 

 

 

 

From: wireless-boun...@wispa.org [mailto:wireless-boun...@wispa.org] On
Behalf Of Forbes Mercy
Sent: Sunday, September 05, 2010 2:52 PM
To: WISPA General List; Butch Evans
Subject: [WISPA] My fun (kidding) weekend on call

 

Thanks for the comments Butch, this has been a busy on-call weekend with the
bandwidth manager dropping twice (It actually shut the power off on the
server) and several Mikrotik towers refusing to come back up until rebooted.
To try to battle this I went to the WIKI for Mikrotik and entered this
string:

/ ip firewall filter
add chain=input connection-state=established comment="Accept established
connections"
add chain=input connection-state=related comment="Accept related
connections"
add chain=input connection-state=invalid action=drop comment="Drop invalid
connections" 
add chain=input protocol=udp action=accept comment="UDP" disabled=no 
add chain=input protocol=icmp limit=50/5s,2 comment="Allow limited pings" 
add chain=input protocol=icmp action=drop comment="Drop excess pings" 
add chain=input protocol=tcp dst-port=22 comment="SSH for secure shell"
add chain=input protocol=tcp dst-port=8291 comment="winbox" 
# End of Edit #
add chain=input action=log log-prefix="DROP INPUT" comment="Log everything
else"
add chain=input action=drop comment="Drop everything else"

Immediately after implementing this WhatsUp indicated:

 
 bandwdith manager
  HTTP(Down at least 5 min)
 

So how can I protect my bandwidth manager and still monitor it at the same
time?  I guess I could disable HTTP monitor and do pings on the monitor
software. 

Three more quick questions:
1) I didn't put in these lines because I wasn't sure what IP's to use, same
problem when I installed PRTG I'm not sure what IP's I need to monitor
within the system to watch:



# Edit these rules to reflect your actual IP addresses! # 
add chain=input src-address=159.148.172.192/28 comment="From Mikrotikls
network" 
add chain=input src-address=10.0.0.0/8 comment="From our private LAN"

2)  The "add chain=input action=log log-prefix="DROP INPUT" comment="Log
everything else"" command indicates there is a log that I can watch foul
traffic, where do I find that log?

3) Is there more sets of examples for firewalls for my Mikrotik routers
somewhere, I'm searching WIKI's right now.

Thanks for the help, hope your weekend is going well, I already have logged
100 miles just chasing outages down in the last 12 hours.

Forbes

On 9/5/2010 10:39 AM, Butch Evans wrote: 

On Fri, 2010-09-03 at 14:15 -0700, Forbes Mercy wrote: 
  

I keep adding filters as traffic presents itself but help and 
training is very expensive and extraordinarily technical 


 
While I would disagree that training is "very expensive", I would have
to agree that it is very technical in nature.  My training sessions are
normally under $300/day for students (not counting
hotels/flights/etc.). 
 
  

On my backhauls 
when one Mikrotik goes down its not unusual for the foul traffic to 
permeate throughout (yes I'm bridged) the network and take down other 
Mikrotik's and often requires a drive to reboot then they work fine 
again, irritating, yes but still great equipment.


 
Training would be especially good if you could learn something that
would keep you from having to roll a truck even once every 2 weeks.  It
wouldn't take long to pay for that.
 
  

Ubiquiti is a monster for power and throughput, it's menus are basic 
but filters entry options are slim and limited to IP rather than by 
protocol so some things sneak through that wouldn't with Mikrotik.


 
This, unfortunately, is one "cost" of less expensive gear.  FWIW, you
have most of the same functionality available in both platforms, but
it's just not in the GUI for UBNT.
 
  

I promised an analogy so here goes, I feel from experience that Mikrotik 
is the Linux of equipment, you better know what you're doing when you 
buy it.  


 
UBNT is linux, too.  :-)
 
  

Ubiquiti is like Windows, pretty GUI driven, and simplified at a 
reasonable cost.


 
You have access to iptables and more in the ssh/telnet interface with
Ubiquiti.
 
  

 



-

Re: [WISPA] mikrotik vs ubiquiti

[Robert West]

I use No Ack only on links longer than 5 miles.  I had problems with No Ack
at shorter distances.  But I put AirMax on it all, not just because it's
TDMA but because it will screw with those trying to hack into it.

 

Bob-

 

 

 

From: wireless-boun...@wispa.org [mailto:wireless-boun...@wispa.org] On
Behalf Of Scott Carullo
Sent: Sunday, September 05, 2010 6:28 PM
To: WISPA General List
Subject: Re: [WISPA] mikrotik vs ubiquiti

 

I'm not sure your assessment of UBNT not recommending to use airmax on PTP
as a general statement hold true.  It is unlikely that they would have built
in a specific PTP noack mode into airmax configuration if it was their
suggestion not to use it.  I use it on lots of links and it works very well.

You are incorrect in saying airmax and nstream cannot increase throughput in
comparison with 802.11n.  In the real world in the wild a lot of times it is
only airmax or nstream that will even let a link perform reliably,
regardless of what the textbook says. Not to mention it allows our
throughput to increase in contrast to your statement.

Scott Carullo
Technical Operations
877-804-3001 x102

   

 

  _  

From: "Vyacheslav Vasilyev" 
Sent: Sunday, September 05, 2010 3:50 PM
To: "WISPA General List" 
Subject: Re: [WISPA] mikrotik vs ubiquiti




2010/9/5 Jeromie Reeves 

On Sun, Sep 5, 2010 at 2:10 AM, Vyacheslav Vasilyev 
wrote:
>

What is a Rockets PPS with airmax on?

Ubnt does not recomend to use Airmax On in ptp due to  lower performance. We
did not test  it. 

 Ubnt ,  MT  and any other atheros 802.11n based products  have  aprox equal
max throughput in  standard 802.11n mode.

But when airmax and nstreme are ON  they have  different performance. 

Tecnically Airmax is  polling ( round robin algorithm ) like nstreme or
turbocell/ outdoor router Proxim .   Ubnt polling uses latest Atheros
chipset clock timing  ( ubnt  calls it "tdma" ) , that may be usefull only
in ptmp .

 Nstreme 2 is  Nstreme 1, that also  uses clock of atheros chipset

So both  airmax and Nstreme 1,2 can not increase max throughput in ptp in
comparison with standard 802.11n (hardware atheros aggregation On, 2 chains)
in ptp in ideal conditions( no interference).

Nstreme 1,2  is able to improve  802.11n link in comparison with standard
802.11n  ( Nstreme Off)  mode in  presence of interference  or/and multipath
fading  due to it's feature of link parameter adaptation  according packets
losses rate.

I  do not  know is  Airmax support link adaptation ( modulation ) or not . I
suppose -not yet.

 

  > We tested TDMA freebsd Sam Lefler MAC 802.11a implementation .at the
simular
> platform ( Alix, CM9) . It  also has poor  throughput at small packet size
> ( but much better then standard 802,11a)  and it is may be improved  by
> using more powerfull h/w.

Ive read up on Sams work and have been very impressed. I looked at the BSD
TDMA
driver back in late 08 or early 09 (been a while), when I did it was
missing glue and needed
a bit of polish. It looked like adding GPS sync to it would have only
been a matter of getting
2 or more AP's to hold then start on the same signal and keep the
timing windows synced.

 

 Software TDMA  Linux/freebsd  implementation  based on 802.11 chipset
hardware is separate issue . I  think it may be useful  in ptp and our test
showed promising results. With regards to ptmp  IMHO it is not viable. There
is standard fixed TDMA BWA  techhology called fixed wimax 802.16-2004/2009.
There is  802.16-2004 miniPCI cards - ASIC hardware TDMA implementation .
There is TDMA 802.16-2004  BS/CPE  Linux based software. For what  a lot of
people want full software TDMA implemenation?

 Vyacheslav Vasilyev
  UNIDATA
  Fixed  BWA solution




WISPA Wants You! Join today!
http://signup.wispa.org/



WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/

 




WISPA Wants You! Join today!
http://signup.wispa.org/

 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/

Re: [WISPA] Trango, UBNT & Alvarion equipment for sale

[Jim Wilson]

Cameron Kilton  midcoast.com> writes:

> 
> I have:
> 
> (9) Trango Fox 5580 Units for sale $60 Each or $480 if you buy all of them.
> 
> (2) Trango 5800 units $75 each
> 

> 
> Contact me and I'll see what I have for a real good price.
> 

Do you still have any 5580 or other trango gear?
Thanks
Jim
5092941174





WISPA Wants You! Join today!
http://signup.wispa.org/

 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/

Re: [WISPA] mikrotik vs ubiquiti

[Greg Ihnen]

Today I wondered the same thing about the noack.

My experience running a very short distance (1000 yards) PtP using two NS5M's 
with the TX turned almost all the way down, in the jungle where there is no 
interference what so ever is that there's no speed gain in using Airmax. But 
soon the PtP will become a PtMP and I'm expecting the TDMA of Airmax to 
outperform 802.11n.

Greg

On Sep 5, 2010, at 5:57 PM, Scott Carullo wrote:

> I'm not sure your assessment of UBNT not recommending to use airmax on PTP as 
> a general statement hold true.  It is unlikely that they would have built in 
> a specific PTP noack mode into airmax configuration if it was their 
> suggestion not to use it.  I use it on lots of links and it works very well.
> 
> You are incorrect in saying airmax and nstream cannot increase throughput in 
> comparison with 802.11n.  In the real world in the wild a lot of times it is 
> only airmax or nstream that will even let a link perform reliably, regardless 
> of what the textbook says. Not to mention it allows our throughput to 
> increase in contrast to your statement.
> 
> Scott Carullo
> Technical Operations
> 877-804-3001 x102
> 
> 
> 
> 
> From: "Vyacheslav Vasilyev" 
> Sent: Sunday, September 05, 2010 3:50 PM
> To: "WISPA General List" 
> Subject: Re: [WISPA] mikrotik vs ubiquiti
> 
> 
> 
> 2010/9/5 Jeromie Reeves 
> On Sun, Sep 5, 2010 at 2:10 AM, Vyacheslav Vasilyev  
> wrote:
> >
> What is a Rockets PPS with airmax on?
> Ubnt does not recomend to use Airmax On in ptp due to  lower performance. We 
> did not test  it.
>  Ubnt ,  MT  and any other atheros 802.11n based products  have  aprox equal 
> max throughput in  standard 802.11n mode.
> But when airmax and nstreme are ON  they have  different performance.
> Tecnically Airmax is  polling ( round robin algorithm ) like nstreme or 
> turbocell/ outdoor router Proxim .   Ubnt polling uses latest Atheros chipset 
> clock timing  ( ubnt  calls it "tdma" ) , that may be usefull only  in ptmp .
>  Nstreme 2 is  Nstreme 1, that also  uses clock of atheros chipset
> So both  airmax and Nstreme 1,2 can not increase max throughput in ptp in 
> comparison with standard 802.11n (hardware atheros aggregation On, 2 chains) 
> in ptp in ideal conditions( no interference).
> Nstreme 1,2  is able to improve  802.11n link in comparison with standard 
> 802.11n  ( Nstreme Off)  mode in  presence of interference  or/and multipath 
> fading  due to it's feature of link parameter adaptation  according packets 
> losses rate.
> I  do not  know is  Airmax support link adaptation ( modulation ) or not . I 
> suppose -not yet.
>  
>   > We tested TDMA freebsd Sam Lefler MAC 802.11a implementation .at the 
> simular
> > platform ( Alix, CM9) . It  also has poor  throughput at small packet size
> > ( but much better then standard 802,11a)  and it is may be improved  by
> > using more powerfull h/w.
> Ive read up on Sams work and have been very impressed. I looked at the BSD 
> TDMA
> driver back in late 08 or early 09 (been a while), when I did it was
> missing glue and needed
> a bit of polish. It looked like adding GPS sync to it would have only
> been a matter of getting
> 2 or more AP's to hold then start on the same signal and keep the
> timing windows synced.
>  
>  Software TDMA  Linux/freebsd  implementation  based on 802.11 chipset 
> hardware is separate issue . I  think it may be useful  in ptp and our test 
> showed promising results. With regards to ptmp  IMHO it is not viable. There 
> is standard fixed TDMA BWA  techhology called fixed wimax 802.16-2004/2009. 
> There is  802.16-2004 miniPCI cards - ASIC hardware TDMA implementation . 
> There is TDMA 802.16-2004  BS/CPE  Linux based software. For what  a lot of 
> people want full software TDMA implemenation?
>  Vyacheslav Vasilyev
>   UNIDATA
>   Fixed  BWA solution
> 
> 
> 
> WISPA Wants You! Join today!
> http://signup.wispa.org/
> 
> 
> WISPA Wireless List: wireless@wispa.org
> 
> Subscribe/Unsubscribe:
> http://lists.wispa.org/mailman/listinfo/wireless
> 
> Archives: http://lists.wispa.org/pipermail/wireless/
> 
> 
> 
> 
> WISPA Wants You! Join today!
> http://signup.wispa.org/
> 
> 
> WISPA Wireless List: wireless@wispa.org
> 
> Subscribe/Unsubscribe:
> http://lists.wispa.org/mailman/listinfo/wireless
> 
> Archives: http://lists.wispa.org/pipermail/wireless/




WISPA Wants You! Join today!
http://signup.wispa.org/

 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/li

Re: [WISPA] mikrotik vs ubiquiti

[Scott Carullo]

I'm not sure your assessment of UBNT not recommending to use airmax on PTP 
as a general statement hold true.  It is unlikely that they would have 
built in a specific PTP noack mode into airmax configuration if it was 
their suggestion not to use it.  I use it on lots of links and it works 
very well.

You are incorrect in saying airmax and nstream cannot increase throughput 
in comparison with 802.11n.  In the real world in the wild a lot of times 
it is only airmax or nstream that will even let a link perform reliably, 
regardless of what the textbook says. Not to mention it allows our 
throughput to increase in contrast to your statement.

Scott Carullo
Technical Operations
877-804-3001 x102



From: "Vyacheslav Vasilyev" 
Sent: Sunday, September 05, 2010 3:50 PM
To: "WISPA General List" 
Subject: Re: [WISPA] mikrotik vs ubiquiti

2010/9/5 Jeromie Reeves 

On Sun, Sep 5, 2010 at 2:10 AM, Vyacheslav Vasilyev  
wrote:
>

What is a Rockets PPS with airmax on?

Ubnt does not recomend to use Airmax On in ptp due to  lower performance. 
We did not test  it. 
 Ubnt ,  MT  and any other atheros 802.11n based products  have  aprox 
equal max throughput in  standard 802.11n mode.
But when airmax and nstreme are ON  they have  different performance. 
Tecnically Airmax is  polling ( round robin algorithm ) like nstreme or 
turbocell/ outdoor router Proxim .   Ubnt polling uses latest Atheros 
chipset clock timing  ( ubnt  calls it "tdma" ) , that may be usefull only  
in ptmp .
 Nstreme 2 is  Nstreme 1, that also  uses clock of atheros chipset
So both  airmax and Nstreme 1,2 can not increase max throughput in ptp in 
comparison with standard 802.11n (hardware atheros aggregation On, 2 
chains) in ptp in ideal conditions( no interference).
Nstreme 1,2  is able to improve  802.11n link in comparison with standard 
802.11n  ( Nstreme Off)  mode in  presence of interference  or/and 
multipath fading  due to it's feature of link parameter adaptation  
according packets losses rate.
I  do not  know is  Airmax support link adaptation ( modulation ) or not . 
I suppose -not yet.
 
  > We tested TDMA freebsd Sam Lefler MAC 802.11a implementation .at the 
simular
> platform ( Alix, CM9) . It  also has poor  throughput at small packet 
size
> ( but much better then standard 802,11a)  and it is may be improved  by
> using more powerfull h/w.

Ive read up on Sams work and have been very impressed. I looked at the BSD 
TDMA
driver back in late 08 or early 09 (been a while), when I did it was
missing glue and needed
a bit of polish. It looked like adding GPS sync to it would have only
been a matter of getting
2 or more AP's to hold then start on the same signal and keep the
timing windows synced.

 
 Software TDMA  Linux/freebsd  implementation  based on 802.11 chipset 
hardware is separate issue . I  think it may be useful  in ptp and our test 
showed promising results. With regards to ptmp  IMHO it is not viable. 
There is standard fixed TDMA BWA  techhology called fixed wimax 
802.16-2004/2009. There is  802.16-2004 miniPCI cards - ASIC hardware TDMA 
implementation . There is TDMA 802.16-2004  BS/CPE  Linux based software. 
For what  a lot of people want full software TDMA implemenation?
 Vyacheslav Vasilyev
  UNIDATA
  Fixed  BWA solution



WISPA Wants You! Join today!
http://signup.wispa.org/



WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/





WISPA Wants You! Join today!
http://signup.wispa.org/

 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/

Re: [WISPA] mikrotik vs ubiquiti

[Vyacheslav Vasilyev]

2010/9/5 Jeromie Reeves 

> On Sun, Sep 5, 2010 at 2:10 AM, Vyacheslav Vasilyev 
> wrote:
> >
> What is a Rockets PPS with airmax on?
>
Ubnt does not recomend to use Airmax On in ptp due to  lower performance. We
did not test  it.
 Ubnt ,  MT  and any other atheros 802.11n based products  have  aprox equal
max throughput in  standard 802.11n mode.
But when airmax and nstreme are ON  they have  different performance.
Tecnically Airmax is  polling ( round robin algorithm ) like nstreme or
turbocell/ outdoor router Proxim .   Ubnt polling uses latest Atheros
chipset clock timing  ( ubnt  calls it "tdma" ) , that may be usefull only
 in ptmp .
 Nstreme 2 is  Nstreme 1, that also  uses clock of atheros chipset
So both  airmax and Nstreme 1,2 can not increase max throughput in ptp in
comparison with standard 802.11n (hardware atheros aggregation On, 2 chains)
in ptp in ideal conditions( no interference).
Nstreme 1,2  is able to improve  802.11n link in comparison with standard
802.11n  ( Nstreme Off)  mode in  presence of interference  or/and multipath
fading  due to it's feature of link parameter adaptation  according packets
losses rate.
I  do not  know is  Airmax support link adaptation ( modulation ) or not . I
suppose -not yet.

  > We tested TDMA freebsd Sam Lefler MAC 802.11a implementation .at the
simular
> platform ( Alix, CM9) . It  also has poor  throughput at small packet size
> ( but much better then standard 802,11a)  and it is may be improved  by
> using more powerfull h/w.

> Ive read up on Sams work and have been very impressed. I looked at the BSD
> TDMA
> driver back in late 08 or early 09 (been a while), when I did it was
> missing glue and needed
> a bit of polish. It looked like adding GPS sync to it would have only
> been a matter of getting
> 2 or more AP's to hold then start on the same signal and keep the
> timing windows synced.
>

 Software TDMA  Linux/freebsd  implementation  based on 802.11 chipset
hardware is separate issue . I  think it may be useful  in ptp and our test
showed promising results. With regards to ptmp  IMHO it is not viable. There
is standard fixed TDMA BWA  techhology called fixed wimax 802.16-2004/2009.
There is  802.16-2004 miniPCI cards - ASIC hardware TDMA implementation .
There is TDMA 802.16-2004  BS/CPE  Linux based software. For what  a lot of
people want full software TDMA implemenation?
 Vyacheslav Vasilyev
  UNIDATA
  Fixed  BWA solution



WISPA Wants You! Join today!
http://signup.wispa.org/


WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/



WISPA Wants You! Join today!
http://signup.wispa.org/

 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/

[WISPA] My fun (kidding) weekend on call

[Forbes Mercy]

Thanks for the comments Butch, this has been a busy on-call weekend
with the bandwidth manager dropping twice (It actually shut the power
off on the server) and several Mikrotik towers refusing to come back up
until rebooted.  To try to battle this I went to the WIKI for Mikrotik
and entered this string:

/ ip firewall filter
add chain=input connection-state=established comment="Accept
established connections"
add chain=input connection-state=related comment="Accept related
connections"
add chain=input connection-state=invalid action="" comment="Drop
invalid connections" 
add chain=input protocol=udp action="" comment="UDP" disabled=no 
add chain=input protocol=icmp limit=50/5s,2 comment="Allow limited
pings" 
add chain=input protocol=icmp action="" comment="Drop excess pings" 
add chain=input protocol=tcp dst-port=22 comment="SSH for secure shell"
add chain=input protocol=tcp dst-port=8291 comment="winbox" 
# End of Edit #
add chain=input action="" log-prefix="DROP INPUT" comment="Log
everything else"
add chain=input action="" comment="Drop everything else"

Immediately after implementing this WhatsUp indicated:

bandwdith manager 

HTTP(Down at least 5 min) 

So how can I protect my bandwidth manager and still monitor it at the
same time?  I guess I could disable HTTP monitor and do pings on the
monitor software. 

Three more quick questions:
1) I didn't put in these lines because I wasn't sure what IP's to use,
same problem when I installed PRTG I'm not sure what IP's I need to
monitor within the system to watch:

# Edit these rules to reflect your actual IP addresses! # 
add chain=input src-address=159.148.172.192/28 comment="From Mikrotikls network" 
add chain=input src-address=10.0.0.0/8 comment="From our private LAN"
2)  The "add chain=input action="" log-prefix="DROP
INPUT" comment="Log everything else"" command indicates there is a log
that I can watch foul traffic, where do I find that log?

3) Is there more sets of examples for firewalls for my Mikrotik routers
somewhere, I'm searching WIKI's right now.

Thanks for the help, hope your weekend is going well, I already have
logged 100 miles just chasing outages down in the last 12 hours.

Forbes

On 9/5/2010 10:39 AM, Butch Evans wrote:

  On Fri, 2010-09-03 at 14:15 -0700, Forbes Mercy wrote: 
  
  
I keep adding filters as traffic presents itself but help and 
training is very expensive and extraordinarily technical 

  
  
While I would disagree that training is "very expensive", I would have
to agree that it is very technical in nature.  My training sessions are
normally under $300/day for students (not counting
hotels/flights/etc.). 

  
  
On my backhauls 
when one Mikrotik goes down its not unusual for the foul traffic to 
permeate throughout (yes I'm bridged) the network and take down other 
Mikrotik's and often requires a drive to reboot then they work fine 
again, irritating, yes but still great equipment.

  
  
Training would be especially good if you could learn something that
would keep you from having to roll a truck even once every 2 weeks.  It
wouldn't take long to pay for that.

  
  
Ubiquiti is a monster for power and throughput, it's menus are basic 
but filters entry options are slim and limited to IP rather than by 
protocol so some things sneak through that wouldn't with Mikrotik.

  
  
This, unfortunately, is one "cost" of less expensive gear.  FWIW, you
have most of the same functionality available in both platforms, but
it's just not in the GUI for UBNT.

  
  
I promised an analogy so here goes, I feel from experience that Mikrotik 
is the Linux of equipment, you better know what you're doing when you 
buy it.  

  
  
UBNT is linux, too.  :-)

  
  
Ubiquiti is like Windows, pretty GUI driven, and simplified at a 
reasonable cost.

  
  
You have access to iptables and more in the ssh/telnet interface with
Ubiquiti.

  







WISPA Wants You! Join today!
http://signup.wispa.org/

 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/

Re: [WISPA] mikrotik vs ubiquiti

[Butch Evans]

On Fri, 2010-09-03 at 14:15 -0700, Forbes Mercy wrote: 
> I keep adding filters as traffic presents itself but help and 
> training is very expensive and extraordinarily technical 

While I would disagree that training is "very expensive", I would have
to agree that it is very technical in nature.  My training sessions are
normally under $300/day for students (not counting
hotels/flights/etc.). 

> On my backhauls 
> when one Mikrotik goes down its not unusual for the foul traffic to 
> permeate throughout (yes I'm bridged) the network and take down other 
> Mikrotik's and often requires a drive to reboot then they work fine 
> again, irritating, yes but still great equipment.

Training would be especially good if you could learn something that
would keep you from having to roll a truck even once every 2 weeks.  It
wouldn't take long to pay for that.

> Ubiquiti is a monster for power and throughput, it's menus are basic 
> but filters entry options are slim and limited to IP rather than by 
> protocol so some things sneak through that wouldn't with Mikrotik.

This, unfortunately, is one "cost" of less expensive gear.  FWIW, you
have most of the same functionality available in both platforms, but
it's just not in the GUI for UBNT.

> I promised an analogy so here goes, I feel from experience that Mikrotik 
> is the Linux of equipment, you better know what you're doing when you 
> buy it.  

UBNT is linux, too.  :-)

> Ubiquiti is like Windows, pretty GUI driven, and simplified at a 
> reasonable cost.

You have access to iptables and more in the ssh/telnet interface with
Ubiquiti.

-- 

* Butch Evans   * Professional Network Consultation*
* http://www.butchevans.com/* Network Engineering  *
* http://store.wispgear.net/* Wired or Wireless Networks   *
* http://blog.butchevans.com/   * ImageStream, Mikrotik and MORE!  *





WISPA Wants You! Join today!
http://signup.wispa.org/

 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/

Re: [WISPA] mikrotik vs ubiquiti

[Jeromie Reeves]

On Sun, Sep 5, 2010 at 2:10 AM, Vyacheslav Vasilyev  wrote:
>
> 2010/9/5 Jeromie Reeves 
>>
>> On Sat, Sep 4, 2010 at 2:39 PM, Vyacheslav Vasilyev 
>> wrote:
>> Airmax was on or off? What were the single direction speeds for each?
>
> In all   Rocket ptp link tests  Airmax   is   off.
> Rocket
"In comparison we  tested in lab  the same board RB411AH with CM9
802.11a card  ( Nstreme  On, packet aggregation frame policy  is
3200). In this mode unit has 39K pps. In Nstreme off mode this unit
has only 4K pps. "

What is a Rockets PPS with airmax on?

> -max simplex ( one direction) throughput BW  20 Mhz
> -( udp,1470), UL/DL= 75.3/78.4 mbps
> -(udp,64 bytes) UL/DL= 11.4/12.3 mbps
> Max simplex throughput (udp, 1470) BW 40 MHz
> ( udp,1470), UL/DL= 95.5/95.5 mbps - limited by 100 BaseT
> -(udp,64 bytes) UL/DL= 7.6./8.6 mbps
> RB411, MPLS;  5GHz-only-N; Nstreme OFF Polling disable,CSMA enable
> showed aprox  the same results.
> Nstreme ON in 802.11n mode gives worse results at small packet traffic.
>>
>> > We noticed  that internal bandwidth test ( between wireless interfaces )
>> > at
>> > small packets shows much higher throughput than via Ethernet+wireless.
>> MT, Ubnt, or both? With or with out NStream/NStream2/Airmax? Running
>> as a bridge, a router, WDS?
>
> All MIPS platform (ubnt, mt ) and also we tested Alix (x86) have  problem of
> poor performance at small packet traffic between ethernet-wireless,
> wireless-wireless in any mode ( bridge, routing..Nstremeairmax, ) that
> cases  low pps and problems with passing multiservice trafffc( data, voip,
> p2p-utorrent etc).
> Problem is connected with MAC802.11a/b/g and also 8022.11n  implemetation.
> Using more powerfull CPU does not resolve it.
> We tested TDMA freebsd Sam Lefler MAC 802.11a implementation .at the simular
> platform ( Alix, CM9) . It  also has poor  throughput at small packet size
> ( but much better then standard 802,11a)  and it is may be improved  by
> using more powerfull h/w.

Ive read up on Sams work and have been very impressed. I looked at the BSD TDMA
driver back in late 08 or early 09 (been a while), when I did it was
missing glue and needed
a bit of polish. It looked like adding GPS sync to it would have only
been a matter of getting
2 or more AP's to hold then start on the same signal and keep the
timing windows synced.


>  Vyacheslav Vasilyev
>  UNIDATA
>  Fixed  BWA solution



WISPA Wants You! Join today!
http://signup.wispa.org/

 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/

Re: [WISPA] mikrotik vs ubiquiti

[Vyacheslav Vasilyev]

>
>  We tested TDMA freebsd Sam Lefler MAC 802.11a implementation .at the
> simular platform ( Alix, CM9) . It  also has poor  throughput at small
> packet size  ( but much better then standard 802,11a)  and it is may be
> improved  by using more powerfull h/w.
>
>
Sorry freebsd8  tdma Sam 802.11a implementation was tested with
Wiston DCMA-82 (CM12 ) .


>   Vyacheslav Vasilyev
>  UNIDATA
>  Fixed  BWA solution
>
>>  >
>> 
>> > WISPA Wants You! Join today!
>> > http://signup.wispa.org/
>> >
>> 
>> >
>> > WISPA Wireless List: wireless@wispa.org
>> >
>> > Subscribe/Unsubscribe:
>> > http://lists.wispa.org/mailman/listinfo/wireless
>> >
>> > Archives: http://lists.wispa.org/pipermail/wireless/
>> >
>>
>>
>>
>> 
>> WISPA Wants You! Join today!
>> http://signup.wispa.org/
>>
>> 
>>
>> WISPA Wireless List: wireless@wispa.org
>>
>> Subscribe/Unsubscribe:
>> http://lists.wispa.org/mailman/listinfo/wireless
>>
>> Archives: http://lists.wispa.org/pipermail/wireless/
>>
>
>



WISPA Wants You! Join today!
http://signup.wispa.org/

 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/

Re: [WISPA] mikrotik vs ubiquiti

[Vyacheslav Vasilyev]

2010/9/5 Jeromie Reeves 

> On Sat, Sep 4, 2010 at 2:39 PM, Vyacheslav Vasilyev 
> wrote:
> Airmax was on or off? What were the single direction speeds for each?
>
In all   Rocket ptp link tests  Airmax   is   off.
Rocket
-max simplex ( one direction) throughput BW  20 Mhz
-( udp,1470), UL/DL= 75.3/78.4 mbps
-(udp,64 bytes) UL/DL= 11.4/12.3 mbps
Max simplex throughput (udp, 1470) BW 40 MHz
( udp,1470), UL/DL= 95.5/95.5 mbps - limited by 100 BaseT
-(udp,64 bytes) UL/DL= 7.6./8.6 mbps
RB411, MPLS;  5GHz-only-N; Nstreme OFF Polling disable,CSMA enable
showed aprox  the same results.
Nstreme ON in 802.11n mode gives worse results at small packet traffic.

> > We noticed  that internal bandwidth test ( between wireless interfaces )
> at
> > small packets shows much higher throughput than via Ethernet+wireless.
> MT, Ubnt, or both? With or with out NStream/NStream2/Airmax? Running
> as a bridge, a router, WDS?
>
All MIPS platform (ubnt, mt ) and also we tested Alix (x86) have  problem of
poor performance at small packet traffic between ethernet-wireless,
wireless-wireless in any mode ( bridge, routing..Nstremeairmax, ) that
cases  low pps and problems with passing multiservice trafffc( data, voip,
p2p-utorrent etc).
Problem is connected with MAC802.11a/b/g and also 8022.11n  implemetation.
Using more powerfull CPU does not resolve it.
We tested TDMA freebsd Sam Lefler MAC 802.11a implementation .at the simular
platform ( Alix, CM9) . It  also has poor  throughput at small packet size
( but much better then standard 802,11a)  and it is may be improved  by
using more powerfull h/w.
 Vyacheslav Vasilyev
 UNIDATA
 Fixed  BWA solution

>  >
> 
> > WISPA Wants You! Join today!
> > http://signup.wispa.org/
> >
> 
> >
> > WISPA Wireless List: wireless@wispa.org
> >
> > Subscribe/Unsubscribe:
> > http://lists.wispa.org/mailman/listinfo/wireless
> >
> > Archives: http://lists.wispa.org/pipermail/wireless/
> >
>
>
>
> 
> WISPA Wants You! Join today!
> http://signup.wispa.org/
>
> 
>
> WISPA Wireless List: wireless@wispa.org
>
> Subscribe/Unsubscribe:
> http://lists.wispa.org/mailman/listinfo/wireless
>
> Archives: http://lists.wispa.org/pipermail/wireless/
>



WISPA Wants You! Join today!
http://signup.wispa.org/

 
WISPA Wireless List: wireless@wispa.org

Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless

Archives: http://lists.wispa.org/pipermail/wireless/