12-27-2018 12:27 PM
My guess would be there's other devices interfering. Is this the best to measure these?
show interface wifi1 _counter
--------------------------------------------------------------------------------
215 recv eol interrupts
1397703 rx frames
15868 rx data frames
14596 rx unicast data frames
855 rx multicast data frames
417 rx broadcast data frames
1381003 rx management frames
832 rx control frames
1181 rx unknown frames
111 rx BAR (Block Ack Request) frames
1398884 11n rx frames
703 rx chip antenna errors
962 rx Retries
6% rx retry rate
685040 rx CRC errors
684600 rx CRC errors with no physical error
32% rx CRC rate
54040 tx data frames
9544 tx unicast data frames
32683 tx multicast data frames
11813 tx broadcast data frames
--------------------------------
52669 tx WMM best effort data frames
60 tx WMM background data frames
616 tx WMM video data frames
695 tx WMM voice data frames
1504421 tx management frames other than beacon
1462603 tx beacon frames
1 tx global timeout interrupts
1 tx carrier sense timeout interrupts
140 tx BAR (Block Ack Request) frames
742 tx aggregated completions
5936 tx retries
2444 tx RTS failures
3492 tx retries
2893 unicast data tx retries
2444 unicast data tx RTS failures
449 unicast tx retries
23% unicast data tx retry rate
1390 tx retries of sub frames
44513 tx frames with no ack marked
4556 tx frames with short preamble
781 tx frames with an alternate rate
304 tx frame errors
304 too many hw retries
511 tx available buffers
512 rx PHY error summary count
61905 tx frames succeed
rx phy err:
1 transmit override receive
511 OFDM restart
-128(invalid) tx rssi of last ack
13 rx rssi from histogram
21730 periodic calibrations
01-08-2019 01:17 PM
Hi Sam,
I checked the radio profiles first (it seemed the more safe approach in production).
In the 2.4Ghz everything from 1 to 11 Mbps was already put on N/A and in the 5ghz area everything from 6 to 9 Mbps.
I'll start playing around by lowering the power of the neighbours.
12-27-2018 02:39 PM
It will be an issue for neighboring devices regardless of channel because you still have to share the same air space. The auto channel settings will allow the APs to choose different channels based on channel cost, which takes interference in to account as well. So it's likely that neighboring APs did move to different channels to help with the issue, but it will still cause some problems for you.
12-27-2018 02:36 PM
Hi Sam, Great help. I appreciate it. I'll be mailing that VHM ID soon.
I just want to clarify: this specific access point is operating at channel 13 at this point. So is the -55 dBm still an issue as it happens on channel 1? I thought that was the point of auto adjusting channels?
Or am I misunderstanding this and mixing things up ?
12-27-2018 02:13 PM
Thanks for the neighbor output, I'm seeing three neighbors in the -50's, which is too loud so that would be causing some overlap and interference. Fortunately they are Aerohive devices, so we should be able to adjust their power down to fix this.
To do this we want to go to the Monitor Tab> Select the check box next to the AP> Modify> and lower the power on the drop down menu. If you have Maps uploaded to your HiveManager with the APs placed in their respective locations, and if you can provide your VHM ID, I can take a look and give you specific power recommendations. In general, I recommend lowering by 2dBm from wherever you are now per test, so as to avoid dramatic shifts of coverage. If you lower by 2dBm, then let the APs run for at least an hour (so we can be sure to get at least one background scan in) and then look at the neighbors list again to see if we still have any -50 values, that would be the best way to dial in on the correct power level.
Additionally, if you could disable the lower data rates on your SSIDs, this will require a stronger connection between the AP and the client device. This will mean that the client device needs to be closer to the AP before connecting. Not only will this reduce the chances of unintentional roaming, but it will also reduce the distance the packets need to travel, which will also reduce collision and interference on your network. To disable the lower data rates we need to go to Configuration> Open the Network Policy> Open the SSID> Expand radio and rates> In the 2.4 GHz 11/bg Rate Setting box, we want to turn 1Mbps-9Mbps to N/A. In the 5.0 GHz 11a Rate Setting box we want to turn 6Mbps and 9 Mbps to N/A.
Finally, we want to check the Radio Profile. The first thing we want to check is your Channel Width, which you can see by expanding Channel and Power. We want to see this at less than 80MHz channel width. This is because the large the channel width, the more overlap we have between channels, and if there is a problem on one channel (RF Interference for instance) it is now shared with all of the channels we are combining.
The second thing we want to check is that we have Short Guard Interval disabled, which we can see by expanding Radio Settings. This setting is designed to shorten the time we have to wait between when we send one packet and when we can send the next one. The problem with this is, if we have an environment with interference, we would just be sending more packets in to a crowded environment faster, increasing the damage we are causing. If this is enabled, please disable it.
I usually also recommend band steering on the 2.4GHz profile as 5GHz typically has less interference than 2.4GHz, but I can't say that for sure in this case and you've stated you see the CRC failures on both, so I think we should hold on on this change for now.
Once you are able, please push these changes out to your APs. A delta configuration push should be enough so we won't have to interrupt client traffic. To push a delta configuration out to your APs you will want to go to Monitor> Select the check box next to the AP> Update> Advanced> Upload and Activate Configuration> Make sure it says Delta at the top (if it doesn't, click on "Settings" in the right hand corner, select one of the options for Delta and click on the floppy disk icon in the top right to save these settings)> Upload.
After we've let these new settings run for about an hour, we will want to pull tech data. if you can send that tech data to me either on this post or directly at communityhelp@aerohive.com, I can take a look and let you know how the interference levels are doing.
Sorry for the wall of text, please let me know if I can clarify anything.