SM's Disconnecting

Products ePMP
21
@ninedd wrote:
@dkeltgen wrote:

What's your logging level on the AP and SMs?  The default setting will not log the disconnect reason.

Hmmm...  Where do I configure that?

Under Configuration->System->Syslog. Just check all the boxes

syslog.png

22

AH - thank you.  I had been over the GUI with a fine-toothed comb... but not fine enough!  :)

23

OK, now that I have logging of all things checked on, this is what I see in the SM's when a bunch of them disassociate at the same time.

Nov 24 08:28:55 SM NAME kernel: [676116.380000] SM disassociated from AP[00:04:56:ca:0e:0c] F=2457 11nght10. Reason: 48 (COMMUNICATION LOST)
Nov 24 08:28:59 SM NAME kernel: [676120.400000] TPC set initial Tx-Power to 30dbm
Nov 24 08:29:15 SM NAME kernel: [676136.920000] SM associated with AP[00:04:56:ca:0e:0c]
Nov 24 08:29:15 SM NAME kernel: [676136.940000] connection is UP

So, ''Communication Lost'' on each of the SM's doesn't tell me much. :)  Maybe 'Reason 48' means something to you?

On the AP, a tthe same time - this is what I see

Nov 23 18:48:49 AP NAME kernel: [626441.810000] SM[00:04:56:c9:66:25] aid=4 disassociated. Reason: COMMUNICATION LOST
Nov 23 18:49:06 AP NAME kernel: [626458.570000] SM[00:04:56:f9:58:03] aid=1 disassociated. Reason: COMMUNICATION LOST
Nov 24 06:46:39 AP NAME kernel: [669511.850000] SM[00:04:56:f9:5b:03] aid=2 disassociated. Reason: COMMUNICATION LOST
Nov 24 08:28:55 AP NAME kernel: [675647.050000] SM[00:04:56:f9:58:03] aid=1 disassociated. Reason: COMMUNICATION LOST
Nov 24 08:29:08 AP NAME kernel: [675660.590000] SM[00:04:56:c9:66:25] aid=4 disassociated. Reason: COMMUNICATION LOST
24

Hi,

Please find belov link to KB wjith description of Registration State Codes:

http://community.cambiumnetworks.com/t5/ePMP-Troubleshooting/Network-Entry-State-Definitions-and-Courses-of-Action/m-p/40712#U40712

Communication Lost means the SM cannot be reached by AP polling, resulting in SM disassociation.

Please check RSSI and SNR levels.

Also please check with Spectrum Analyzer tool whether there is any bursty interference on the channel which may cause Radio Link quality degradation.

Thank you.

25

OK, thank you.

The RSSI's are between -56 and -73, and the SNR's are all about 30-40 range - so that should be pretty decent.

However, those two that are close to -70 are Force200's so I've been planning on changing those out to Connectorized radios and 22 dBi Grid Antennas, that should put them at -65 or so, and maybe that'd make a difference.  However, there are connections on here which are -64/64 signal and -38/38 SNR and they both unlinked 5:25 ago

So, maybe it is just occasionaly LOUD noise, but that'll be tricky to catch on the Spectrum - because I'd have to basically watch it live for 6 hours at a time.  Also, it seems strange because the logs show these clients seem to reassociate within seconds and performance it good for them right away it seems.

However - the simplest thing for me to try on this end is to repalce the Force200 with ePMP1000-22dB Antennas and see if an extra 6 dBi has a good effect on them.  It does seem like the client with the best signal (-56) does have the least disconnection problems.

RDH-N_RSSI_SNR_NOV24_16.jpg
RDH-N_RSSI_SNR_NOV24_16.jpg906×266 74.7 KB