We have lust purchased a bunch of OmniTIK radios to test out and there seems to be a few issues.

1. I can not get any of the units to fire up the signal strength LED's
2. One of the units has a screw lodged in behind the signal strength LED and the window. this is a manufacturing quality control issue.
3. when set up and being tested the packet loss seems to be pretty high. This is with the unit under test in the workshop in what would have to be an optimum configuration.

We have updated the firmware to the latest version so as to be sure that this is not the issue.

1. I can not get any of the units to fire up the signal strength LED's

"signal strength LED's", are you sure? there are no such LEDs on the OmniTik. this is an AP device, there is no specific signal to show (if you have 40 clients, each has different signal), the LEDs on the back are ethernet LEDs

2. One of the units has a screw lodged in behind the signal strength LED and the window. this is a manufacturing quality control issue.

please send a photo to support, the problem is not clear

3. when set up and being tested the packet loss seems to be pretty high. This is with the unit under test in the workshop in what would have to be an optimum configuration.

how did you test this? can you post more specific information about this problem?

Im using these as AP's and Client devices. there is what appears to be the option to set the LED's as a signal meter in the LED settings but this just seems to turn them on.

The packet loss was what was given in the details under the bandwidth test option. the units were being tested in the workshop and transferring UDP as trying the same with TCP causes the cpu to run at 100%

As for the screw... here you go.

you can configure each led to do something in the led menu. see all the options. by default they show ethernet activity.

about the screw, please send the serial number to support and we will check. what's interesting, that omnitik doesn't even use screws inside.

SN: 2BBF01108D8B/121

And I did config the signal level LED's but they don't seem to work or seem to ignore the signal levels. I will test it a little more in the field and its not overly critical but would be nice to have this feature.

Other than that they are a nice little unit, clean and simple. Looking forward to seeing a 2.4Ghz version and possibly one with external antenna connectors. It would be nice to be able to daisy chain the PoE to another device.

Regards
Andrew

It would be nice to be able to daisy chain the PoE to another device.

It would be VERY nice.

you having ping packet loss when running on the UDP test, that's normal. UDP test doesn't mean the actual throughput it can handle.

then more better way is you use the TCP test with tcp count set to 20, then u can see the actual throughput it can carry when it's in live network.

you having ping packet loss when running on the UDP test, that's normal. UDP test doesn't mean the actual throughput it can handle.

then more better way is you use the TCP test with tcp count set to 20, then u can see the actual throughput it can carry when it's in live network.

you mean to say that there is no UDP traffic in actual networks ??

I was only stating what was reported by the test utilities provided in the device and not by an external device. UDP is critical to the base network's functionality and to be honest we should not be seeing such losses in an optimal test environment.

maybe I am spoiled with having used high grade DMR's but firstly I would not expect to find a screw lodged in the LED window of a commercial device. and secondly if i purchase a link that is supposed to provide a speed of X then i should be getting X and not X with high packet loss albeit UDP or TCP.

Regards
Andrew

I was only stating what was reported by the test utilities provided in the device and not by an external device. UDP is critical to the base network's functionality and to be honest we should not be seeing such losses in an optimal test environment.

I replied to comment above mine

secondly if i purchase a link that is supposed to provide a speed of X then i should be getting X and not X with high packet loss albeit UDP or TCP.

what speed is it supposed to provide? which speed are you getting (you didn't say)? you can't expect maximum speed if your client device is far away, or has a low gain antenna. you need to know some wireless basics, wireless is a very sensitive environment, the speed depends on too many factors. nobody guarantees speed in a wireless environment.

packet loss is not an indication of speed. ping is a low priority type of packet, if there is traffic running, it will have higher priority than ping.

The test environment spans some 5 or so meters and there was negligible interference in the area. the Omnitik has an integrated antenna so there should be no issues there with cable loss or other problems.

signal levels ranges between -37 to -43 after turning the power down a little.

When the link associated we it reported it was connected at either 6 or 9mbps however in the tests is showed a connection speed of 300/300 from memory.

I say from memory as these units have since been deployed on site as we had a pretty tight time frame to work to and something is better than nothing in this situation.

As for experience and understanding or wireless, I have Trango certification for both the DMR product and their earlier product, I have Bridgewaves Certification and over the last 7 years have deployed a wireless network that spans most of the greater Auckland area as well as doing consultancy to other vendors both in NZ and off shore.

When in get the chance I will unpack the other units I have here and config them the same then conduct the same speed tests but will this time do some screen captures for you.

Regards
Andrew

-37 to -43

there is your problem. signal should be -60 not more. you have much too much signal.

When the link associated we it reported it was connected at either 6 or 9mbps however in the tests is showed a connection speed of 300/300 from memory.

yes, when there is no traffic, it stays on low rates, when the traffic increases, it goes to higher rates

Thanks. I had a feeling that it may have been something like that. its not easy finding a place that you can set a test up in and you can really only turn the power down so much.
Normally I could shove a set of attenuation in line but its a bit hard to do with fixed antennas.

That gives me an idea Attenuation Condoms...

Radios should be all live in the next day or so so should be able to see what they do in the real world.
the installer had no issues putting them up so far.

Andrew

what is the distance, and what devices are the clients? let us know the results.

This one is not over a huge distance. just connected to one of the radios and getting about -71 and from memory the distance is about 1 to 1.2Km.

So far I have not done any tweaking, they have just been installed and configured to connect to each other.

We are using the omnitik as both client and AP in bridge mode due to the layout and may eventually set up them up to handle MESH.

status: running
duration: 41s
tx-current: 27.7Mbps
tx-10-second-average: 27.6Mbps
tx-total-average: 23.6Mbps
rx-current: 8.1Mbps
rx-10-second-average: 8.2Mbps
rx-total-average: 8.1Mbps
lost-packets: 238
random-data: no
direction: both
tx-size: 1500
rx-size: 1500

About the LEDs, see http://forum.mikrotik.com/viewtopic.php?f=3&t=55912

Hi Andrew,

When all your omnitiks are in production please post your performance results,i will be changing m new towers over to Omnitiks in the next month from Ubiquiti Airmax products.From what i read the Omnitik's have the edge over Airmax.