RSSI and Jitter are ok, but ping latency are HORRIBLE!

Products PMP
1

Hello, I’ve a weir problem and any help will be appreciated.

I’ve one 5200AP (w/o reflector) and two 5200SM (with reflector) associated and registrated to the AP. The SM are located at one distance of 5Km .

Please look at the registration levels and details in the AP:


LUID: 002 : MAC: 0a-00-3e-00-4f-d9 State: IN SESSION (Encrypt Disabled)
      Site Name : FOL
      Software Version : CANOPY 7.0.7 Mar 02 2005 15:06:58
      Software Boot Version : CANOPYBOOT 3.0
      FPGA Version : 051104 (DES)
      Session Timeout: 26, AirDelay 185
      Session Count: 1, Reg Count 1, Re-Reg Count 0
      RSSI (Avg/Last): 1591/1598   Jitter (Avg/Last): 2/2   Power Level (Avg/Last): -63/-62
      DnRate: 3500 DnLimit: 10000 UpRate: 3500 UpLimit: 10000 (kbit)

LUID: 003 : MAC: 0a-00-3e-00-e4-1b State: IN SESSION (Encrypt Disabled)
      Site Name : aeropuerto
      Software Version : CANOPY4.2.1 Apr 16 2004 15:23:05
      Software Boot Version : CANOPYBOOT 2.7
      FPGA Version : 062403 (DES)
      Session Timeout: 25, AirDelay 196
      Session Count: 3, Reg Count 3, Re-Reg Count 0
      RSSI (Avg/Last): 1229/1224   Jitter (Avg/Last): 4/4   Power Level (Avg/Last): -70/-70
      DnRate: 10000 DnLimit: 10000 UpRate: 10000 UpLimit: 10000 (kbit)


The link test for the LUID02 :


Stats for LUID: 2 Test Duration: 8
Downlink RATE: 2488896 bps
Uplink RATE: 659008 bps
Downlink Efficiency: 63 Percent
Max Downlink Index: 96
Actual Downlink Index: 60
Expected Frag Count: 38889
Actual Frag Count: 64754
Uplink Efficiency: 93 Percent
Expected Frag Count: 10297
Actual Frag Count: 11028


The link test for the LUID03 :


Stats for LUID: 3 Test Duration: 8
Downlink RATE: 1667520 bps
Uplink RATE: 615808 bps
Downlink Efficiency: 58 Percent
Max Downlink Index: 96
Actual Downlink Index: 56
Expected Frag Count: 26055
Actual Frag Count: 45989
Uplink Efficiency: 92 Percent
Expected Frag Count: 9622
Actual Frag Count: 10370


The ping to the LUID02 and LUID03 :


64 bytes from 201.93.223.169: icmp_seq=88 ttl=63 time=233 ms
64 bytes from 201.93.223.169: icmp_seq=89 ttl=63 time=658 ms
64 bytes from 201.93.223.169: icmp_seq=90 ttl=63 time=53.7 ms
64 bytes from 201.93.223.169: icmp_seq=91 ttl=63 time=93.3 ms
64 bytes from 201.93.223.169: icmp_seq=92 ttl=63 time=25.7 ms
64 bytes from 201.93.223.169: icmp_seq=93 ttl=63 time=151 ms
64 bytes from 201.93.223.169: icmp_seq=94 ttl=63 time=23.6 ms
64 bytes from 201.93.223.169: icmp_seq=95 ttl=63 time=115 ms
64 bytes from 201.93.223.169: icmp_seq=96 ttl=63 time=55.5 ms
64 bytes from 201.93.223.169: icmp_seq=97 ttl=63 time=125 ms
64 bytes from 201.93.223.169: icmp_seq=98 ttl=63 time=461 ms
64 bytes from 201.93.223.169: icmp_seq=99 ttl=63 time=483 ms
64 bytes from 201.93.223.169: icmp_seq=100 ttl=63 time=120 ms

— 201.93.223.169 ping statistics —
100 packets transmitted, 99 received, 1% packet loss, time 99154ms
rtt min/avg/max/mdev = 23.634/267.457/1174.612/251.094 ms, pipe 2


As you can see, the latency is very terrible !! and the data transfer from adn to remote are very slow.

What can I do to fix this problem ?

2

Hi,

What software version on your AP?
according to your explanation, your SM’s have different version of software.
try to upgrade all your stuff to 7.2.9 n enable hardware scheduling (if you use Advantage AP)

thanks

regards,
mrdlnf

3

I agree, upgrade to 7.2.9 and re-test. One other thing to check is the alignment as the downlink on both units looks poor. Have you run the spectrum analyzer at both ends? Possibly there is local interference which is affecting the receive power.

4

The most common cause of this I find is interference try chaging freqs if at all possible

5

I’ll agree, could be interference. However, in my experience you’d usually see re-registrations and a higher jitter when that happens. Your link test shows the radios are re-transmitting frames, but its mostly the downlink that is affected.

Check the status at the SM side. Its possible the SM are receiving interference and the AP is not. You’ll notice what’s shown at the AP’s session list is different from what the SM says on its status page. Its just a difference in perspective - they are values calculated on the received signal.

You can check for a clear channel by turning off your AP, and run the spectrum analyser from an SM. You need your own AP turned off when you check it, so you can tell if there’s any other signals on the same frequency you are currently trying to use.

Also, you really need everything on the same firmware version. Get the CNUT tool and update the AP, then both your SM to 7.2.9 using the autoupdate feature in CNUT. After that, you may think about HW scheduling if you have a newer advantage AP. Use the CNUT tool to enable HW scheduling, as your older SMs will need a specific firmare for it.

I have an AP with about 152 SMs on it, some have reflectors at the distance you’re trying to achieve. We did nothing special other than put the reflector on the SM. Our pings run about 8 to 15ms with HW scheduler (was 25 - 50 before HW scheduling).

6

you also have to graph the used bandwidth (and errors) with mrtg/rrd/cacti whatever you like, to see, if there is no other unneeded traffic in the cell what can make you poor performance

7

looks like interference to me

8

yep looks like interfearence. the sm’s are talking to the ap’s fine this is why you see good levels and uplink effic above 90%. The problem is the sm’a are not recieving the ap well. this is why the downlink effic is low. What are the levels on the sm’s status page? put the sm’s in spectrum analyzer mode and see. You also could have a bad transmitter in the ap if the signal is low.