5.8 & 2.4; Coexistence Problems?

[magister]

Greetings;

I am administering a 2.4 ghz network that worked fine for months, but suddenly, I have started having problems with my long-distance length which is a point-to-point, mac-filtered, unicast and encrypted run of about 16 miles.

I am slowly working through all of the possibilities, but since everything works in a lab setting, everything worked for months and when redeployed after testing, everything works for a few days; Plus, because I've switched-out the antenna cable and reversed the two bridges (Cisco BR350) and due to some occasional "funny" log-entries; I've started reaching for every straw.

Sometime around the time I started having problems keeping a connection another agency installed a 5.8 ghz Trangolink system that is within two miles of me at one end and half a mile at the other. So, basically we are transversing the same terrain which is composed of little more than the desert floor and at a significantly lower elevation.

Is anyone familiar with the Trangolink system and/or does anyone have any thoughts as to whether, a 5.8 system blasting at me might cause my 2.4 to have problems?

Thanks in Advance;
R

[Mr Lodin]

Unless there is something wrong with their 5.8 GHz equipment it won't (by far) reach even noise floor levels in the 2.4 GHz region.
Do your antennas work properly? A faulty pig-tail could reduce your much needed power dramatically. Do you have microwave line-of-sight at all times?

[magister]

Thanks for the quick response and I appreciate your input and any input that might you may offer, or any suggestions that may be forthcoming from others.

I've brought the remote Yagi in twice and because I'm limited to power sources, I've only been able to test shorter distances; But, it reaches and performs on my tests. Plus, when I redeploy after the tests, the antenna and network works about half the time. Nothing fails back at the lab and I've left it operational for days.

The other party has also installed a pair of AP's running 2.4 through directional antennas but not pointed directly at me, pretty much directly in my line of sight, except they are at ground level and I am going from mountain to mountain, so I should be a couple of hundred feet over their heads. (Though, their 5.8 is essentially at the same elevation as myself at both ends)

I inherited this project because no one else wanted to do it. Basically, the client had purchased a lot of the equipment including the antennas using somebody else's list and then, he showed the closet full of junk to people and after a year of searching, I agreed to take the challenge.

According to Cisco, the antennas they had purchased coupled with the cables would get 17 miles @ 1mps in a perfect world and we are doing 15.82. The system did work without fail for months and though I can't pinpoint the beginning of the problems to the precise date the other network was installed, because their's went-up piecemeal and they have had some downtime due to hardware issues; I'm trying to weigh all possibilities.

Our signal strength hovered around 5% before the other network went-up and since the problems started, we can't seem to get over 1% when it is remotely installed. So, I'm wondering if they may have a hand in our reduction and when it goes down, 1% is not giving us much margin of error.

Of course, there could be some other reason for the problems; Right now, I'm studying a couple of possibilities; But, because I am totally unfamiliar with 5.8 and have never had anyone install two access points, one of which is on the same channel as myself, directly in my LOS. I'm grasping at straws. Otherwise, I start-out on a ridge that's about 300' above the ground and shoot over a desert that gets to more than a 1000' below me, before it rises back-up on the other side and though, the wired-side is on a hill and not a ridge, within half a mile the world drops-down about 500'.

I only learned of the other system exactly 10 days ago and finally, they have agreed to communicate about their setup. You mention that "unless something is wrong with their 5.8"; Right this minute, they are communicating; Could you suggest something, I might have them check?

Any ideas are more than welcome;
I thought I was finished with this project;

Thanks;
R

[oshea85]

5% signal strength? That sounds like your problem right there; I'm surprised the linked ever worked acceptably. Was it always this low, or is this something new? Maybe you've got a bad radio...if one's lost Tx power, it would still work in the lab. But, I'd have to say, for almost 16 miles, I would've definitely gone with a 2' or 3' parabolic antenna on both ends. Overbuild, overbuild, overbuild...

First thing I would do is get the time synced on those bridges; set up an ntp server, and point both bridges at it. RTFM on how to do this. You want them synced to the same source; do both sites have independent access to the internet hopefully? If so, you could pick a public NTP server. If not, since you're link is shaky, sync is going to be sporadic, but might be good enough.

Then, clear the logs and interface counters.

Run a link test between the bridges, observe results. Try running a carrier-busy test on the root bridge, you're looking for somebody on your channel that you may not be aware of. You will want to run that test a bunch of times to get a decent picture of what's going on. to be thorough, I like to swap the root and non-root bridges (through configuration) and run the carrier-busy test from the other side.

Scrutiny of the logs, now that the time is synced up, may reveal something useful.

If you're desperate, and can provide remote access to the bridges, I'd be willing to take a look.

PS - have you tried changing channels? The carrier test would give you a clue if this is needed.

[magister]

Unfortunately, we are starting to stray from the subject-heading, but because I'm getting to my wits-end and it looks like my initial question may have been answered; I'm game as long as ya'll are and I appreciate any help, anyone can offer. As I mentioned in my first posts, I'm in the middle of the desert and I'm not dealing with an everyday setup; So, I'm finding myself just talking to folks in the clients office, sometimes getting a brainstorm or ruling something out, but most of the time just watching their eyes glaze over because unpacking a Dell is pretty much their computer knowledge.

With that said...

5% may be a slight exageration; It may have been closer to 15; It's been a couple of months since the thing worked, reliably and due to access limitations; I am only granted access to the remote relay, six times a month on pre-determined, evenly-spaced days and I can only go out with an escort. And, I was also barred from the area for the entire month of July. Nonetheless, the signal-strength was never high and 5% is in my memory, plus I know that it's now stuck at 1% when the network is up, because I've been watching it like a hawk and keeping much better records. (After it was working and the client was happy, I pretty much paid no attention and it just chugged-along. Though, whenever I was at the remote location for some other reason, I would cycle the power on everything and other than that, I don't think I even did many warm reboots after everyone was pleased)

And, I agree that it would be better to have a parabolic at each end, but the client had an associate, contact a friend who made-up a list that works in an urban setting. I was the third tech on the job (Though, the only one who actually did anything productive and who actually made it work). But, when I was brought onboard, a 13dBi Yagi and a 21dBi Dish was what I was given for the long-distance link. I've considered making them buy a parabolic, but they're two years into the project (I've been involved for a year) and tens of thousands of dollars have fallen into this black hole. Personally, I'd rather be positive that an antenna upgrade would do the trick and try to get it back where it was, before I try to force another expenditure down their throat. (After all, they just renewed my contract and hired me for another job within the past month)

Anyway, back to your questions;

There is nothing at the remote end except for the networking equipment and a couple of cameras. It is a closed area and the network exists, solely for the cameras. There's no remote internet access outside of this system and everything is solar-powered. But, I agree with the concept and since this time, I tried to pretty-much sync the times for just this situation, if I can get-in from behind the remote bridge and if the thing isn't back-up by that time, I will try to go through the remote log in the the field when I'm allowed back-out, next Wednesday.

I've switched-out the antenna cable at the remote end, tried an entirely new wire, field-tested an offbrand Yagi that I have for another project and only recently, reversed the bridges. And apparently, because the thing has "halfway" lost contact since my initial post, I have another symptom that I am currently researching to see if it'll shed any light and hopefully, give me an answer.

Since reversing the bridges and also redefining the radio port as primary, which assigned it a different mac-addy and I gave the radio and ethernet port separate IP's; On the last trip-out, you could not log into the remote bridge from the mainland at first, though it did show-up as an authenticated/associated on the root bridge. I did not press the young lady who was testing across the network, primarily because I did not know the question; But as part of my recent routine, I reversed the antenna cables between the two devices in that box and when I rebooted, she had direct contact. Though, I believe the contact was most likely, the result of the hard reboot and not the act of changing cables; But, that is what I did.

The thing has since, halfway lost contact (and in case it may have any bearing, per usual, it lost contact during a rainstorm; Other rainstorms have occurred since my "inexplicable" problem arose and it has always lost the connection when it's buried in a cloud, but this time it only halfway came back). Anyway, the "mainland" bridge still sees the remote unit, calls it authenticated/associated, but it truncates the bridge's name and says it's IP is "unassigned".

I've tried several, both cold and warm reboots on the root bridge and I sat here and deauthenticated the remote bridge, repeatedly for probably an hour. The root always "sees" the remote after a reboot. After a deauthentication, it doesn't always immediately reauthenticate; Though, the root still "sees" it and after a couple more tries, it does reauthenticate. (Sometimes, during this test, it would immediately reauthenticate and sometimes it would take up to five deauthentications for it to happen)

This causes me to think away from a hardware problem, because the root "sees" the remote and packets are being exchanged with what appears to be no more errors than has been the norm, lately. It's just that the IP is being read as "unassigned", the name is truncated by a single character and I can't get past the remote bridge to the devices on the other side.

Any ideas would be appreciated and I will post another message with more details on the competing system(s), shortly; But, as someone who has been using forums since Matt invented the things; I realize that too many topics in a single post can get confusing.

Thanks Again;
Random Thoughts are Certainly Welcome;
R

[oshea85]

you should not be playing with the primary port settings, as strange things sometimes happen...the fact that it goes out in a rainstorm tells me its at least partly a power issue.

Put the bridges back to default settings, then reconfigure them. All you should be configuring is IP address on both sides, SSID, root/non-root, WEP if you're using it.

15% RSSI is still very low. Is the signal strength percieved as the same by both bridges?

The authentication could be crapping out because the quality of teh link is so poor...sometimes it works, other times is doesn't. If you look at the radio errors, are you seeing lots of retries, CRC errors, etc? These would be due poor link quality...

[countywifi]

But, because I am totally unfamiliar with 5.8 and have never had anyone install two access points, one of which is on the same channel as myself, directly in my LOS.

Have them turn off there APs then check to see if better. You can find antennas very cheap out there

OSHEA85 is right on the money

[magister]

Thanks countrywifi; Based on oshea's apparent agreement that I'm probably looking at power/signal strength problem, coupled with the fact that I learned today, their AP-construct wasn't where I thought it was and instead, both of their antennas are oriented, pretty much towards my root from a whole other location. (I thought they formed a T to my signal, directly in my path, but instead they are forming a V from much further back and theoretically, both of their AP's are broadcasting towards my root, especially considering the width of the signal and when you plot something real on a topo).

According to my stats from the rootside, I am getting quite a high number of retries; But, only a handful of maxes. The other party did volunteer to change their channel and I have inquired as to whether they would be willing to do it, tomorrow; So that, I may be able to regain full contact. But, the holiday weekend may cause them to delay.

Originally, I was doing my long-distance on channel 1; I inquired, once I learned of their existence about their channel and when I had not gotten a response in 10 days, I switched to channel 11 and I have since learned that they are using, both.

I don't know if I can get them to shut the whole thing down without being a tremendous pain; But, they did express a willingness to change one of their channels and now that I've "halfway" lost contact and got the coordinates of their AP-setup; I emailed a request, late this afternoon and we'll see what they say about tomorrow.

I'll report back here, as soon as I know;

Otherwise, though I was the only thing for miles in any direction except a Vortac, when I started; There is now, the Trango pointed directly towards me, a 2.4 V-shaped cloud in my path and there's a fairly-new cluster of experimental "pods" out the sidedoor from the root that are screaming at each other in the 900mHz range. All of which, were suspiciously installed around the same time as I started having problems. Plus, I originally mounted my parabolic on a new tower that now boasts antennas for Motorola two-ways; The Motorolas have pretty much been there from the start, the antennas moved to the new tower, not too long after it was installed. And, because I mentioned it, I pretty much shoot right over the Vortac which is about seven or eight miles out; It broadcasts at a low frequency (116.9 or 122.1, I don't recall offhand) and it has always been there, though of course, I guess DHS could've theoretically installed some kind of jamming device.

And, don't even get me started telling you about the night, something that resembled that thing they keep under the desert on SG-1, ambled over my head...

Any or All Thoughts are Welcome;
Thanks;
R

PS) And, I also agree with oshea that one shouldn't blindly mess with the Primary Port setting; It does cause extreme funniness and thankfully, I did it in the lab and only after the problem started. There isn't much documentation available on the function, but unless I'm misreading something, it did enable the beacon to stay alive so that I might play with it, without waiting and driving-out the hour or so to my relay. But, it really isn't for the faint-of-heart, 99.999% of people should not mess with it and then, only if absolutely necessary. Also, it should only be done in a lab setting and only when there's no remote access, other than the radio.

[countywifi]

Hey I made a little mistake and thought the BR 350 had + 24dbm,
only has + 20dbm. So the headroom I said you were close to is even lower. If the 13 dbi antenna is the local one then swap it for a 22dbi unit. It will definately improve S/N. Put your equip on channel 6. Find out if they are vertical or horizontal and if you can just change your polarity opposite their equip. You must get more signal........................