PMP 450m Sessions Dropping

I've got 3 PMP 450m APs. The first two were running 15.0.1, and they seem quite stable. The third was installed with 15.0.2, and is now running 15.0.2.1 - this one has been randomly dropping sessions.

One of the recurring thing I'm seeing in the SM Session Logs of the  SM which do drop a lot are things like this:

02/08/2017 : 00:02:12 CST : Event: SMSESFAIL, MsgType: REGREQTOMAXERR, NewState: OOSERVICE, Flag 0
02/08/2017 : 00:02:34 CST : Event: SMSESACTIVATE, NewState: REGISTERING, Flag 0
02/08/2017 : 00:02:37 CST : Event: SMSESMSG, MsgType: REGREQTO, NewState: REGISTERING, Flag 0
repeated 2 times
02/08/2017 : 00:02:43 CST : Event: SMSESFAIL, MsgType: REGREQTOMAXERR, NewState: OOSERVICE, Flag 0
02/08/2017 : 00:03:01 CST : Event: SMSESACTIVATE, NewState: REGISTERING, Flag 0
02/08/2017 : 00:03:04 CST : Event: SMSESMSG, MsgType: REGREQTO, NewState: REGISTERING, Flag 0

This particular SM reports its Air Delay as "38050 ns, approximately 3.541 miles (18701 feet)" (AP Range is set to 15 miles)

I notice under Statistics-> Scheduler that the "Scheduler Error" field increments when these issues happen. Signal quality is not great, but reasonable, and Spectrum Analysis didn't show any noise. Any ideas? 

Tools → Link Status 5.4/5.7GHz MIMO OFDM - Subscriber Module
0a-00-3e-bx-xx-xx

Downlink Status
Receive Power :
-73.0 dBm ( -76.0 dBm V / -76.0 dBm H )
Signal Strength Ratio :
0.0dB V - H
Signal to Noise Ratio :
22 V / 21 H dB
Beacons :
100 %
Receive Fragments Modulation :
Path V:QPSK:76% 16-QAM:24%
Path H:QPSK:57% 16-QAM:43%
Latest Remote Link Test Efficiency Percentage :
NA %
BER Total Avg Results :
5.944864e-05
Beacons Received Last 15 minutes :
100/100/100% (min/avg/max)
Uplink Status
Transmit Power :
22 dBm (target power [41 dBm] exceeded maximum)
Max Transmit Power :
22 dBm
Power Level :
-72.2 (-77.0 V / -74.0 H) dBm
Signal Strength Ratio :
-3.0dB V - H
Signal to Noise Ratio :
31 dB V / 33 dB H
Latest Remote Link Test Efficiency Percentage :
NA %
Local Status
Session Status :
REGISTERED VC 34 Rate 8X/6X MIMO-B
Spatial Frequency :
1010
Latest Local Link Test Results
Stats for LUID: 18   Test Duration: 2   Pkt Length: 1714   Test Direction Bi-Directional

Link Test without Bridging
VC Downlink Uplink Aggregate Packet Transmit Packet Receive Actual Actual
34 41.11 Mbps 18.16 Mbps 59.28 Mbps,  4259 pps 2625 (1312 pps) 5894(2947 pps)


Efficiency
Downlink Uplink Efficiency Fragments
count Efficiency Fragments
count Actual Expected Actual Expected
100% 160623 160623 100% 70974 70974

Link Test ran on 08:46:16 02/08/2017 CST

Currently transmitting at:
VC 34 Rate 8X/6X MIMO-B

Hi,

Scheduler errors and re-registrations aren't my particular area of expertise but I know a couple of engineers here that are looking at that sort of thing, there are improvements for re-registrations coming in the 15.0.3 release (although I'm not sure you are describing exactly the same fault). If you could collect the engineering.cgi and send that on I could add it to the information they are looking through.

my email is

andrew.rimmerATcambiumnetworks.com

(replace the AT with @, obviously)

regards,

Andrew.

1 Like

I emailed you the engineering.cgi capture from the AP

Got it. Thank you.

I updated the firmware on this AP to 15.0.3 early Thursday morning, and I would have to say, at this time, this firmware loadout does not appear to result in any improvement on this issue.

Thanks for the feedback. I've passed this on to the engineers currently investigating this.

Was there a resolution to this issue. We are experancing the same issues on PMP 450 3.65

There might be totally different causes between this symptom occurring on a 450m vs. a 450 (or 450i).

We have addressed this issue further on R15.1, which will be released today (hopefully).

Regarding it happening on the 450 as you mention SBC... I think we need more information to try to help.  Maybe better to start a new thread (as this one pertains, and is likely specific to, the 450m).

Matt

We've been having this issue for over a year on the PMP450 3.65. Nothing has ever been resolved. Good luck and if you find something, please let me know.

1 Like

As mentioned to SBC above, we've resolved this issue which was specific to 450m.  Dropping sessions can be caused by numerous things, and typically the culprit is interference.

I would suggest starting a separate thread, and we can dig into it there.

I was the original poster, and I haven't seen that problem recur since I went to 15.1 firmware. I would regard the issue in my post as having been solved.