I am trying to get every last drop out of our internet. Most towers have a
MikroTik Routing, Netonix doing POE and Switching, Fed by AF5x, and Rockets or
Prism for AP's. Sometimes seeing some slow downs I cant account for.
Wondering if MTU or Flow control needs adjusted.
The MTU settings are
unless you are passing VLAN's MPLS/VPLS or PPPoE over the AP's everything
else looks fine.
On Mon, Feb 19, 2018 at 10:38 AM, Steve Barnes wrote:
> I am trying to get every last drop out of our internet. Most towers have
> a MikroTik Routing, Netonix doing POE and Switching, Fed by AF5x, and
> R
Are you running air control or anything else that can aggregate the performance
of your site in some nice easy to read graphs?
From: ubnt_users-boun...@wispa.org [mailto:ubnt_users-boun...@wispa.org] On
Behalf Of Steve Barnes
Sent: Monday, February 19, 2018 7:38 AM
To: Ubiquiti Users Group
Subj
We have found, every time, without exception there are only a few causes:
* Not enough backhaul. :)
* Sector is heavily loaded.
* Customers with excessively weak signals pulling the sector down
* #1 INTERFERENCE. Gotta LOVE airmagic and airview on prism aps.
YMMV,
JP
On Mon, 19 Feb 2018 16:38:
Yeah everything is on AIRcontrol. Just can figure it out.
At peak times I can run TCP Btest from PC through CPE to Router on Local tower2
and get 45-50 Meg. (10Mhz channel)
Btest from Router Tower2 to Router tower1 and get 200Meg. (AF5x link 250M
capacity Only 50-100M used).
Tower 1 Connect
I'll plug that Preseem is working really well for us to help find issues
exactly like you're talking about. It is more than monitoring and equipment
stats like AirControl. It actually gives you visibility into the customer
experience. I'd recommend doing their free 30-day trial. I promise you'll
wa
I've got the same issue. I cant identify exactly when it started, just
crept up over time.
On Mon, Feb 19, 2018 at 11:38 AM, Steve Barnes wrote:
> I am trying to get every last drop out of our internet. Most towers have
> a MikroTik Routing, Netonix doing POE and Switching, Fed by AF5x, and
> R
My symptoms exactly Steve. I'm starting to wonder if it's a UBNT thing. My
Cambium sector does not show those issues although to be fair I do have a
few UBNT towers that run great. IDK!
On Mon, Feb 19, 2018 at 2:55 PM, Steve Barnes wrote:
> Yeah everything is on AIRcontrol. Just can figure it o
If it was MTU related it would likely not be a “slowness” issue but more of a
on-off-on-off experience for the end user. Sounds interference or congestion
related to me. Too many connections in connection tracking tables? Broadcast
traffic?
::://Shawn C. Peppers\\:::
> On Feb 19, 2018, at
Guys,
We have had a very similar problem recently with non Cloud Core mikrotik
routers. Nothing seemed wrong, but the router capacity to process packets was
maxed out even though the CPU showed only around 70-80% max. After replacing
the router with a CCR, everything was much better. I recommen
What Travis said about checking your routers is valid.
But to be honest, it's more than likely just customer usage keeps
increasing and you have a new bottleneck somewhere that didn't exist
before. What worked fine in October might be choked now after the holidays
and all the new devices, streamin
Travis, I appreciate the input and cant speak for Steve but I've been
running a CCR for years. Currently it's a CCR1036 being fed with a 2gig
fiber circuit with peak traffic of 700-800 megs. LOL, I'm starting to think
it's the CCR!
On Mon, Feb 19, 2018 at 5:35 PM, Travis Allen
wrote:
> Guys,
> W
Preseem has a 30 day free trail
It will help you find and fix the issues
And I bet you will want to keep it
We did
Mitch Koep
On 2/19/2018 4:28 PM, RickG wrote:
My symptoms exactly Steve. I'm starting to wonder if it's a UBNT
thing. My Cambium sector does not show those issues although to
I’d consider x86 over CCR if your at those numbers
Jon Langeler
Michwave Technologies, Inc.
> On Feb 19, 2018, at 6:05 PM, RickG wrote:
>
> Travis, I appreciate the input and cant speak for Steve but I've been running
> a CCR for years. Currently it's a CCR1036 being fed with a 2gig fiber ci
Our CCR1072 pulls 2g out of 10g and doesn't break a sweat.
YMMV,
JP
On Mon, 19 Feb 2018 19:03:26 -0500, Jon Langeler wrote
> [UTF-8?]Iâd consider x86 over CCR if your at those numbers
>
> Jon Langeler
> Michwave Technologies, Inc.
>
> On Feb 19, 2018, at 6:05 PM, RickG wrote:
>
>
>
> Travi
CCR is fine. One of ours pushes 2 gbps at night and averages 5% CPU
utilization so we have plenty of headroom.
On Feb 19, 2018 6:11 PM, "J Portman" wrote:
> Our CCR1072 pulls 2g out of 10g and doesn't break a sweat.
> YMMV,
> JP
> *On Mon, 19 Feb 2018 19:03:26 -0500, Jon Langeler wrote*
> > [UTF
Good to hear! It does also depend on # of firewall rules, queues, NAT, etc
Jon Langeler
Michwave Technologies, Inc.
> On Feb 19, 2018, at 7:13 PM, Darin Steffl wrote:
>
> CCR is fine. One of ours pushes 2 gbps at night and averages 5% CPU
> utilization so we have plenty of headroom.
>
>> O
Have you tried disabling Pause Frames on the Netonix switch? We've had an
issues with Airfibers and that feature on a few sites.
From: ubnt_users-boun...@wispa.org [mailto:ubnt_users-boun...@wispa.org] On
Behalf Of Steve Barnes
Sent: Monday, February 19, 2018 10:56 AM
To: Ubiquiti Users Group
S
Steve,
Unthrottled what does your download and upload speed look like?
> On Feb 19, 2018, at 21:01, Chris J. Ruschmann wrote:
>
> Have you tried disabling Pause Frames on the Netonix switch? We’ve had an
> issues with Airfibers and that feature on a few sites.
>
> From: ubnt_users-boun...@wis
Are you disabling pause everywhere? Or just Airfiber?
Jon Langeler
Michwave Technologies, Inc.
> On Feb 19, 2018, at 9:01 PM, Chris J. Ruschmann wrote:
>
> Have you tried disabling Pause Frames on the Netonix switch? We’ve had an
> issues with Airfibers and that feature on a few sites.
>
>
20 matches
Mail list logo
