Client Association Successful / Failed

[td_td]

How will the SCG / ZD determine a failed associations and what is the main reason of the failed association?

Will it related to frequency jam? or Will it related to the maxium number of client the AP supported?

[td_td]

That area have 6 APs will all of them nearly 200 client during that time.

AP not install close to metal

[michael_brado]

200 clients all the time  -  There is the Key I think.

Under your WLAN configuration, Advanced Options is a Max Clients setting, with default value of 100.
You might try increase to value 120?

Otherwise, I analyzed your AP support info (that got stuck in SPAM filter) above, and had this interpretation.
( I do NOT see high PHY errors )

TD,

 

    You can sort ZD
debugs by client MAC address with Client Association debug

component selected, or sort SCG logs for your client MAC
messages to see the

association/disassociation activity.  It’s not easy to say a “main reason” without

information.  If you
have 100+ clients on your AP radios, you may reach user limits

if set on the WLAN(s), or maximum encrypted clients (don’t
use WEP/TKIP) for

your AP radios. 

     Your issues do
not appear to be interference related for the particular AP that

you share this support info from.  The 2.4G Radio0 is typically the most
affected

by EMF and noise, but your AP says:

 

Histogram of PHY errors per second (pcttime in each range)
   0  1-500 ..1K  ..2K  ..5K ..10K ..20K ..50K
.100K  more
   2.......86.....8.......2......1.......0........0.......0.........0..........0

 

Or 86% of packets seen over the last 2 minute sample period,
contained less

that 1000 PHY errors per second.  Only 2% of packets had over 2K PHY errors,

and only 1% exceeded 5000 PHY errors per second.

 

Bad RF/interference near the AP would be represented with
higher than 5K

PHY errors per second (bad). 
This is very clean for 2.4G.

 

Athstats from Radio1 (5G) are typically cleaner than the
2.4G band, and your

AP reports:

 

Histogram of PHY errors per second (pcttime in each range)
   0  1-500 ..1K  ..2K  ..5K ..10K ..20K ..50K
.100K  more
  14........1.....0.....0......0........0........0.......0.........0.........0

 

With 0% of packets over 1000 PHY errors per second, (very
clean).

 

What kind of “failed association” problems are you having,
with what type of

client?  Do other type
of clients use the same WLAN and do they have the same

problems, or only one type of client?  Does it happen for them all the time, or

only sometimes, in certain areas?

 

These are the type of questions Tech Support would ask you.

[seanmuir]

When is said I see high PHY errors I was not relating to the last 2mins as the AP might not even be busy when you took that snapshot - was it?

The value I am looking at is the total PHY which has 10 figures on the 2.4 as opposed to 8 on the 5GHz - if the AP's are rebooting like you say, and the uptime is 1 day or 1 hour then you have an issue!

I would cross reference the uptime of the AP before ruling out PHY errors.

Apologies for not being clearer.

Also if this is a HD deployment there are some manual tweaks like rpm,max-aid, rts/cts and proxy arp which might be of benefit to you once you have this issue solved.

[td_td]

The AP support log is get after 1 hour of the congested period since we are not able to get the real time log of the AP at this moment.

Understand that 100 user per ssid per radio will be the bottleneck on this case. Will there any alarm or log that I can get it on the SCG to know I already reach the limit?

For those client even they can connect to the SSID the performance is reported not able to open the web site. Will there any log that can be check about this too?

Since I check on the Event / Alarm on the SCG by sorting some report fail client mac address I can only see normal activity like client jointed / client disconnect / client roam ......

I checked after client report fail and on those AP that reach limit on 5G will got a lot of failed association and on the testing environment if it reach the limit we can find there will have failed association too. So just want to know will there any general idea that will cause failed association on the AP report.

Even the AP show there still have room for 2.4G but our client with only 2.4G end device is also not able to connect to the SSID. 

The above incident happened only once and I guess it should be the number of client higher than expected.

Also understand that I can increase the value to a higher figure but even the current setting if the user is able to join the SSID they also report can't access internet so the customer end user experience is my most concern of it. 

AlphaDog, Sean and Michael thanks for let me know more about it.

[seanmuir]

What is the AP up time? it would be handy to understand if PHY errors are an issue!

Also its not just the wifi that can produce issues like slow web page resolve i.e. what's your backhaul capacity and what router are you using.

You also need to understand your session concurrency values, as loading up APs is not a great idea if all of those clients want traffic and by the sounds of it more are trying to connect, so maybe you dont have enought AP's.

I would usually do the following at the pre-deployment stage:

1. Determine the area capacity
2. What is my target demographic percentage of the area capacity.
3. What is the application demand within that area
4. Choose AP based on above values
5. Use AP choice to determine AP count.

I would also look at implementing a syslog server to dump the AP syslog too as that would be useful, unless you have a Flexmaster or SCI.