Re: [WIRELESS-LAN] Mixing ac AP types

[James Andrewartha]

On 05/02/15 22:50, Cosgrove, John wrote:
> I am designing a new wireless placement for a 5 story building and I
> have been considering mixing 802.11ac AP types.

Is the switching existing? If so, 802.3af vs at would also inform the
decision.

> Meaning.  Some Cisco 1702’s,2702’s and 3702’s.  Placement depending on
> estimated client densities.
>  
> Example.  Conference rooms may have 3702’s yet open areas with less
> people population may have a 1702 or 2702.

I'm not particularly familiar with the Cisco range, but
http://www.cisco.com/c/en/us/td/docs/wireless/technology/apdeploy/8-0/Cisco_Aironet_3700AP.html
(which covers [123][67]00[ei]) seems to have a pretty good feature
comparison.

Lee, note it says "A single GbE cable is fine for Wave-1. While it is
true 802.11ac (Wave-2) will exceed GbE speeds, there is no need or
requirement for cabling greater then GbE for 802.11ac Wave-1. Installers
wishing to future proof new installations should consider pulling CAT-6a
cables  and either another CAT6a or a CAT5e cable (this
allows you to fall back to 2 GbE ports) for some iterations of Wave-2
and/or support 10GbE should this emerge as the method. 10GbE has some
challenges such as PoE standardization. Again, for the foreseeable
future, a single GbE is all that is needed."

Plus Cisco appear to have announced multi-gigabit switches for Q2:
http://www.cisco.com/c/en/us/solutions/enterprise-networks/catalyst-multigigabit-switching/index.html

> I usually hear from people to “Keep it all consistent and the same” and
> I remember in the old days if you mixed “G” in with “b-only” ap’s often
> clients would grab the “G” and never let go no matter how bad the signal
> got.  I am thinking if I at least keep things in the same “family” of
> technology it should work out.

I don't think there would be a problem mixing AP types within the same
technology, from the client point of view it'd just be another roaming
decision. There is more information about how clients decide to roam
now, eg http://support.apple.com/en-au/HT203068 and it seems to be
mostly RSSI based. So if you account for the smaller coverage provided
by the lower-end models (per the diagrams in the above Cisco document)
then there shouldn't be any problems.

See also
http://community.arubanetworks.com/aruba/attachments/aruba/WLAN-Pro-Conf-EU-2014/1/1/WLANPro_EU_MobileDevices%20v1.0-airheads.pdf
for more info on what handover is like now (thanks to powersaving) and
how it should be in an ideal world with 802.11k.

> AP’s are a huge multiplier in a project cost and I was wondering if
> anyone else looked at approaching it this way.

I have a similar challenge in that I'm going to be upgrading some of our
n APs to ac this year. In terms of reducing cost, we have a fairly dense
deployment and so I'm still wishing for APs with a single 802.11ac 5GHz
radio since I turn off a fair few of my 2.4Ghz radios already, and my
client base is 80% 5GHz. Since we're a K-12 1:1 iPad school, I can at
least predict where the ac clients are going to be as we go through our
3 year refresh cycle.

-- 
James Andrewartha
Network & Projects Engineer
Christ Church Grammar School
Claremont, Western Australia
Ph. (08) 9442 1757
Mob. 0424 160 877

**
Participation and subscription information for this EDUCAUSE Constituent Group 
discussion list can be found at http://www.educause.edu/groups/.

Re: [WIRELESS-LAN] Wireless client test bed

[Jason Watts]

Fair points all and pretty much reinforces my gut feeling about it, but I value 
the perspectives of this list and wanted to be sure I wasn't overlooking 
something. There's certainly no substitute for reality.

-- 
Jason Watts
Pratt Institute, Academic Computing
Senior Network Administrator


> On Feb 5, 2015, at 7:52 PM, Jeffrey Sessler  wrote:
> 
> And even if you do replicate it 100%, there is more to a problem then just 
> the device.
>  
> For example, we ran into an issue with an iMac and wifi connectivity where we 
> couldn't reproduce it in our lab. By pure accident, we discovered that the 
> problem manifested itself only if the iMac was rotated so its rear was at a 
> 45 deg angle relative to the position of the AP's antenna's. Any other 
> orientation and it had no issues.
>  
> Jeff
> 
> >>> On Thursday, February 05, 2015 at 9:33 AM, in message 
> >>> <82a99f9da460534f98254bbc299e7612012d946...@ex10mbn01.ua-net.ua.edu>, 
> >>> "Watters, John"  wrote:
> Agreed. With 60,000+ devices on our campus there are hundreds of unique 
> combinations of hardware & software. It would be impossible to get even 75% 
> of the different setups in a test lab. We have neither the budget, the time, 
> the space, nor definitive info on what hardware/software combinations are 
> needed to even attempt a good test bed.
>  
>  
> -jcw  
> 
>   
> 
> John Watters   The University of Alabama
> Office of Information 
> Technology
> 205-348-3992
>  
> From: The EDUCAUSE Wireless Issues Constituent Group Listserv 
> [mailto:WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU] On Behalf Of Lee H Badman
> Sent: Thursday, February 05, 2015 11:27 AM
> To: WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU
> Subject: Re: [WIRELESS-LAN] Wireless client testbed
>  
> Stick with the actual problem machine. You'll never replicate all the 
> variables between hardware/software/user config.
>  
> -Lee Badman
> 
> 
> 
> On Feb 5, 2015, at 9:53 AM, Jason Watts  wrote:
> 
> I'm wondering what folks do to test/recreate issues that clients are having 
> when there are such a mix of OS's, patch levels, client chipsets, and drivers.
> Does anyone have a testbed system that works well for recreating a client 
> setup?
> 
> I'm imagining a system that may be configured with multiple boot disks 
> (various OSes) and an array of PCI and USB wireless adapters with different 
> chipsets.
> 
> 'Is anyone doing anything like this or do you simply work through all issues 
> with the client machine if it's accessible and time permitting?
> 
> --
> Jason Watts | Senior Network Administrator
> W1
> PRATT INSTITUTE
> Academic Computing
> 
> 
> ** Participation and subscription information for this EDUCAUSE 
> Constituent Group discussion list can be found at 
> http://www.educause.edu/groups/.
> ** Participation and subscription information for this EDUCAUSE 
> Constituent Group discussion list can be found at 
> http://www.educause.edu/groups/.
> ** Participation and subscription information for this EDUCAUSE 
> Constituent Group discussion list can be found at 
> http://www.educause.edu/groups/.

**
Participation and subscription information for this EDUCAUSE Constituent Group 
discussion list can be found at http://www.educause.edu/groups/.

Re: [WIRELESS-LAN] Wireless client test bed

[Jeffrey Sessler]

And even if you do replicate it 100%, there is more to a problem then just the 
device.

For example, we ran into an issue with an iMac and wifi connectivity where we 
couldn't reproduce it in our lab. By pure accident, we discovered that the 
problem manifested itself only if the iMac was rotated so its rear was at a 45 
deg angle relative to the position of the AP's antenna's. Any other orientation 
and it had no issues.

Jeff

>>> On Thursday, February 05, 2015 at 9:33 AM, in message 
>>> <82a99f9da460534f98254bbc299e7612012d946...@ex10mbn01.ua-net.ua.edu>, 
>>> "Watters, John"  wrote:

Agreed. With 60,000+ devices on our campus there are hundreds of unique 
combinations of hardware & software. It would be impossible to get even 75% of 
the different setups in a test lab. We have neither the budget, the time, the 
space, nor definitive info on what hardware/software combinations are needed to 
even attempt a good test bed.
 
 
-jcw
  

  
John Watters   The University of Alabama
Office of Information Technology
205-348-3992
 
From: The EDUCAUSE Wireless Issues Constituent Group Listserv 
[mailto:WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU] On Behalf Of Lee H Badman
Sent: Thursday, February 05, 2015 11:27 AM
To: WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU
Subject: Re: [WIRELESS-LAN] Wireless client testbed
 
Stick with the actual problem machine. You'll never replicate all the variables 
between hardware/software/user config.
 
-Lee Badman



On Feb 5, 2015, at 9:53 AM, Jason Watts  wrote:
I'm wondering what folks do to test/recreate issues that clients are having 
when there are such a mix of OS's, patch levels, client chipsets, and drivers.
Does anyone have a testbed system that works well for recreating a client setup?

I'm imagining a system that may be configured with multiple boot disks (various 
OSes) and an array of PCI and USB wireless adapters with different chipsets.

'Is anyone doing anything like this or do you simply work through all issues 
with the client machine if it's accessible and time permitting?
-- 
Jason Watts | Senior Network Administrator
W1
PRATT INSTITUTE
Academic Computing



** Participation and subscription information for this EDUCAUSE 
Constituent Group discussion list can be found at 
http://www.educause.edu/groups/. 
** Participation and subscription information for this EDUCAUSE 
Constituent Group discussion list can be found at 
http://www.educause.edu/groups/. 
** Participation and subscription information for this EDUCAUSE 
Constituent Group discussion list can be found at 
http://www.educause.edu/groups/. 

RE: [WIRELESS-LAN] Wireless client test bed

[Watters, John]

Agreed. With 60,000+ devices on our campus there are hundreds of unique 
combinations of hardware & software. It would be impossible to get even 75% of 
the different setups in a test lab. We have neither the budget, the time, the 
space, nor definitive info on what hardware/software combinations are needed to 
even attempt a good test bed.


-jcw
  [UA Logo]

John Watters   The University of Alabama
Office of Information Technology
205-348-3992

From: The EDUCAUSE Wireless Issues Constituent Group Listserv 
[mailto:WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU] On Behalf Of Lee H Badman
Sent: Thursday, February 05, 2015 11:27 AM
To: WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU
Subject: Re: [WIRELESS-LAN] Wireless client testbed

Stick with the actual problem machine. You'll never replicate all the variables 
between hardware/software/user config.

-Lee Badman


On Feb 5, 2015, at 9:53 AM, Jason Watts 
mailto:jwa...@pratt.edu>> wrote:
I'm wondering what folks do to test/recreate issues that clients are having 
when there are such a mix of OS's, patch levels, client chipsets, and drivers.
Does anyone have a testbed system that works well for recreating a client setup?

I'm imagining a system that may be configured with multiple boot disks (various 
OSes) and an array of PCI and USB wireless adapters with different chipsets.

'Is anyone doing anything like this or do you simply work through all issues 
with the client machine if it's accessible and time permitting?
--
Jason Watts | Senior Network Administrator
W1
PRATT INSTITUTE
Academic Computing


** Participation and subscription information for this EDUCAUSE 
Constituent Group discussion list can be found at 
http://www.educause.edu/groups/.
** Participation and subscription information for this EDUCAUSE 
Constituent Group discussion list can be found at 
http://www.educause.edu/groups/.

**
Participation and subscription information for this EDUCAUSE Constituent Group 
discussion list can be found at http://www.educause.edu/groups/.

Re: [WIRELESS-LAN] Wireless client testbed

[Lee H Badman]

Stick with the actual problem machine. You'll never replicate all the variables 
between hardware/software/user config.

-Lee Badman



On Feb 5, 2015, at 9:53 AM, Jason Watts 
mailto:jwa...@pratt.edu>> wrote:

I'm wondering what folks do to test/recreate issues that clients are having 
when there are such a mix of OS's, patch levels, client chipsets, and drivers.
Does anyone have a testbed system that works well for recreating a client setup?

I'm imagining a system that may be configured with multiple boot disks (various 
OSes) and an array of PCI and USB wireless adapters with different chipsets.

'Is anyone doing anything like this or do you simply work through all issues 
with the client machine if it's accessible and time permitting?

--
Jason Watts | Senior Network Administrator
W1
PRATT INSTITUTE
Academic Computing


** Participation and subscription information for this EDUCAUSE 
Constituent Group discussion list can be found at 
http://www.educause.edu/groups/.

**
Participation and subscription information for this EDUCAUSE Constituent Group 
discussion list can be found at http://www.educause.edu/groups/.

Wireless client testbed

[Jason Watts]

  
  
I'm wondering what folks do to test/recreate issues that clients are
having when there are such a mix of OS's, patch levels, client
chipsets, and drivers.
Does anyone have a testbed system that works well for recreating a
client setup?

I'm imagining a system that may be configured with multiple boot
disks (various OSes) and an array of PCI and USB wireless adapters
with different chipsets.

Is anyone doing anything like this or do you simply work through all
issues with the client machine if it's accessible and time
permitting?

-- 
  Jason Watts |
  Senior Network Administrator
  W1
  PRATT
  INSTITUTE
  Academic Computing
  
  

  

**
Participation and subscription information for this EDUCAUSE Constituent Group discussion list can be found at http://www.educause.edu/groups/.

Mixing ac AP types

[Cosgrove, John]

I am designing a new wireless placement for a 5 story building and I have been 
considering mixing 802.11ac AP types.

Meaning.  Some Cisco 1702's,2702's and 3702's.  Placement depending on 
estimated client densities.

Example.  Conference rooms may have 3702's yet open areas with less people 
population may have a 1702 or 2702.

I usually hear from people to "Keep it all consistent and the same" and I 
remember in the old days if you mixed "G" in with "b-only" ap's often clients 
would grab the "G" and never let go no matter how bad the signal got.  I am 
thinking if I at least keep things in the same "family" of technology it should 
work out.

AP's are a huge multiplier in a project cost and I was wondering if anyone else 
looked at approaching it this way.

Looking forward to your thoughts and suggestions.

Thank you.

John Cosgrove
Network Specialist
Penn State Hershey Medical Center
Penn State College of Medicine
Hershey, PA


**
Participation and subscription information for this EDUCAUSE Constituent Group 
discussion list can be found at http://www.educause.edu/groups/.

Re: [WIRELESS-LAN] Cisco WiSM-2 HA?

[Oliver Elliott]

Hi Danny

We have had some issues with HA mode in the past running 7.4.110. ping to
the gateway would routinely fail causing a failover, at one point both the
primary and secondary failed at once causing an outage. At this point i
disabled HA mode and waited for new software. When 7.6 came around I tried
to upgrade. I had all kinds of issues with 7.6 unrelated to HA mode and
didn't get to run it long enough to even try, having to revert to 7.4.121.
I have recently completed upgrading all of them to 8.0.110 and enabling HA
mode and this time things have worked out much better. We have since
noticed a failed gateway ping on one of our HA pairs, but the new 8.0 logic
means that does not cause a failover. We have some strange issues with the
6500 VSS at the moment so that could relate.

If the issues we've seen up to now have been caused by this VSS issue,
8.0.110 seems to have been good so far, still have an ongoing TAC case so
we'll see what happens.

Oli

On 4 February 2015 at 21:59, Danny Eaton  wrote:

> All,
>
>
>
> I’ve been working with TAC on issues with one of my two
> WiSM-2 HA clusters, and today they’ve finally asked me to break my HA, in
> order to test if the problem still happens – and/or replace the hardware
> itself.  Obviously, I can’t break the HA, because they were ordered as
> such, and the licensing won’t work if I do.
>
>
>
> So, I’m running 7.6.130.0 right now.  Is anyone else
> seeing any odd failovers on their WiSM-2’s in an HA cluster environment?
>
>
>
> Is anyone running the 8.0.110.0 code, and if so, have you
> had any negative experiences?
>
>
>
> Feel free to respond on, or off list.
>
>
>
>Respectfully,
>
>
>
>Danny Eaton
>
>
>
>Snr. Network Architect
>
>Networking, Telecommunications, & Operations
>
>Rice University, IT
>
>Mudd Bldg, RM #205
>
>Jones College Associate
>
>Office - 713-348-5233
>
>Cellular - 832-247-7496
>
>dannyea...@rice.edu
>
>
>
>Soli Deo Gloria
>
>Matt 18:4-6
>
>
>
> G.K. Chesterton, “Christianity has not been tried and found wanting.  It’s
> been found hard and left untried.”
>
>
>
>
>
>
>
>
> ** Participation and subscription information for this EDUCAUSE
> Constituent Group discussion list can be found at
> http://www.educause.edu/groups/.
>
>


-- 
Oliver Elliott
Senior Network Specialist
IT Services
University of Bristol
e: oliver.elli...@bristol.ac.uk
t: 0117 39 (41131)

**
Participation and subscription information for this EDUCAUSE Constituent Group 
discussion list can be found at http://www.educause.edu/groups/.