We have been “bombarded” with radar detects on multiple locations the last few days. We haven’t changed the configs or put up new equipment in theese locations, it just started… Trying to change frequencies didn’t help at all. After a while, we downgraded from version 22.1 to 22.0.2 on all the locations, and there haven’t been a single radar-detect since…
The locations this started happening, has been running several years without radar-detections (of course some random ones, but not for a longer period of time like now), so it seems to us that this has something to do with the 22.1 release. We are using GPS sync (of-course), and the timing has been calculated using the co-location-tool - so our whole network should be in sync (as mentioned - it has been running fine for years).
We see radar-detects both on ap’s as well as sm’s…
We were seeing this as well, and thought it might be the firmware, but went back through the logs and noticed it was happening before we upgraded. Our solution was to position the radio on a channel outside of the FCC TDWR range, now we are no longer getting DFS hits. I think this range is more sensitive, so if there is any wifi in the area that is in the same range, it just continuously gets a hit.
Thanks for your reply, but we are not using frequencies near the TDWR band, and this was not an issue before upgrade. After we downgraded all of the ap’s, the fals dfs hits were gone… Seems like the issue only happens in areas where we have several towers and ap’s - “isolated” towers do not seem to have the same problems. Another thing we are seeing on fw 22.1, is sm’s (especially the one I have at my home) having periodicly very poor snr and modulation - I’ve tried both 450b-HG as well as 450i integrated, and even moved the antenna several places on my house, but nothing helped until I downgraded it to 22.0.2. Since then (probably about 2 months ago - I haven’t had any issues what so ever - modulation downstream stable at 7x mimo-b, and uplink 8x mimo-b… I am positive that there are some issues with the 22.1 fw, but it might be due to some configuration issues in combination with the fw - I do’nt know really…?
I can’t speak to DFS channels as we haven’t pushed out to those yet, though I can speak to your reply about SNR and modulation with what we’re seeing below. We have a small test segment on 22.1.1 running before we push it out to the rest of our network next week. About 130 SM’s spread between 13 450i AP’s, No M’s yet because I ran out of time before our maintenance/changes freeze for the holiday week.
So far on the test batch. I have seen improved LQI (at least via graphing in Meastro) on about 10-15 SM’s, and no changes to SNR. Modulation wise, I have not seen any degradation, rather I have seen some improvement and stabilization on a few that have good signal, but previously had unstable modulations rate before the upgrade.
What kind of distances and signal levels are you looking at? Also how does the spectrum analysis look?
Where the false radar issue was happening, our distances are between 1-5km to the sm’s, and the signal levels mostly between -55 to -65 downstream, and -54 to -70 upstream. It has been running for 7 years whitout having constant radar hits like we experienced with the 22.1 fw. Performance-wise, the 22.1 seemed to give use quite a big gain in aggregated throughput (on the 450m’s - almost 30%) - but we cannot run this version when there are constant false radar…
Regarding the intermittent drops in modulation and snr I was experiencing, the distance is 4km with downstream signal of -66 and upstream signal of -61. As mentioned, the issues with bad snr and modulation, only occurs on fw version 22.1 on the sm (the ap is still on 22.1 on this 450m, because we don’t see the false radar on “isolated” towers.
Unfortunatly I do not have recordings of spectrum analysis, and really don’t think that it will give us any answers anyways - as the problems only occurs on fw 22.1…
BTW - all of the other SM’s on the same ap as I am connected, are still on fw 22.1 ass well as the ap, but at my house, I had the same issues on both the 450b-hg, 450b-mid AND 450i sm. Only when I downgraded the fw the issue was gone - even mounting the antennas at different locations on my house didn’t help. Maybe it has something to do with the angle of which I am connected to the ap, as there are not any other sm’s in the same bearing from the ap as I am (I am well within the 90 degree sector, and we have other customers further to the “edge” of the sector working fine).
I get the issue is only present on 22.1 so no need for the passive aggressive dot dot dot’s lol
The main reason for the Spectrum analysis question was to see if there was a change in that between firmware. I haven’t seen it in a while, but I do recall running some old Ubiquiti gear and seeing false DFS hits due to some odd sensitivity after a FW upgrade. It presented in the spectrum analysis, and appeared to have been detecting the incoming RF from the SM’s and misinterpreting what it was, causing a DFS hit.
Weird one for sure, I’d suggest opening a ticket if no one from Cambium responds here
Thanks once again for your reply! Sorry about the dot dot dot’s - wasn’t meant to be offending or attacking in any way, it’s my way of writing… (there it was - it’s meant more as pointing out something I find peculiar.)
The false dfs hits I suspect might have something to do with the sm’s as you describe with the ubiquiti, but that is just a gutfeeling. The other hypothesis is that some configuration gets automatically changed when installing new fw, and then the gps-sync of the towers isn’t aligning correctly anymore - but I couldn’t find anything obvious when checking.
I’ll try upgrading to either 22.1.1 or the next release at a later point to see if we experience the same issues in the area, and if the problem persists, I’ll open a ticket with cambium I think.
I hate having to upgrade such a big area without knowing if it’ll “break” or not, as this is an area with holiday-cabins, where there is almost no traffic during the week, and loads in the weekend - when it’s much more difficult to troubleshoot because of us techs having the weekends off. Such issues also takes quite a bit of effort to troubleshoot, as it is a good idea to gather as much information as possible when the issues is “showing itself” - just downgrading back to older fw is the fastest way to resolve it of course, but without digging a bit, and downgrading immediatly, we will probably never find the cause of them.
No worries on the …, I do the same a lot especially to convey that I’m frustrated.
Your hypothesis sounds reasonable. There were a few bugs noted in regards to max distance changing, so its plausible that other config changes could happen.
For the updates, I am currently rolling our active gear over to 22.1.1 due to a bug in 22.0.2. Its a rather annoying one that pops up after a device has been up for 50 days or more. The device show offline in maestro and remote management becomes more difficult. I’d recommend jumping into 22.1.1 if you ant to upgrade, but I will let you know If I see any bugs. I have had it running on a small test bed of about 120 radios and haven’t had any issues yet.
So far so good. About 1300 devices swapped over to 22.1.1, with no big issues. Just a few took longer than expected to reconnect (sometimes 30 minutes, sometimes its on the next day). They just say “Waiting for AP session establishment”. It’s a little odd because they were already connected to Maestro prior to the upgrade.
Aside from that oddity, all seems to be well on 22.1.1 so far.
Well - we upgraded most of our network to 22.1.1 (skipping 22.1) the 4th of december, and haven’t noticed any issues som far - so hopefully the bugs we saw in 22.1 is fixed. We are keeping our fingers crossed!
