As of approx. 4p on the 18th of May the 7 day retransmission data in cnMaestro shows a sudden and continuous increase on the uplink side on most (not all) of our AP’s, even on towers over 60 kilometers apart. I am not convinced the data is real and wondering if it might be a cnMaestro glitch? The increase only shows on cnMaestro with the 7 day graphs…switching to the 24 hour graph the uplink does not show at all. Downlink is fine. We have 4 towers approx. 20 to 25 kilometers apart. This week we changed a primary link from epmp 1000 to epmp 3000L. Very possibly related to that but I have checked parameters on all APs, including subscriber MTR levels, up/ down rations, GPS, etc. and all as recommended I believe. Attached is an example, but a similar but lesser jump happened on the AP’s of a tower 45 kilometers away. CPU usage has remained low or even gone down.
I am adding screen shots of 14 more AP’s throughout our network because I see now that the CPU data shows perhaps more clearly some glitch in the data…just cannot be right. Looks like a data gathering issue. Did something change at that time?
Thank you for this report! Do you see the same numbers if you open GUI of some AP?
after reading this i checked my aps … same thing ALL my Aps remote locations … has same readings in cnmaestro !!
Here is two examples
Something appears to have happened on 5/19 around 3am. I think this is when cnMaestro cloud was updated to 3.1.1-r16. We’re basically seeing something like what looks to be the CPU stats swapped for the Retransmission rate on at least ePMP radios. Here’s an instance where you can see something drastically changes with the reporting here:
That’s what it felt like, a switching of data in a couple of fields of cnMaestro, makes sense. Since we were changed equipment on towers that same day I was concerned, but the network appears to be fine. I would appreciate being advised when this gets resolved. Thank you!
We are looking into the issue and will revert with the updates shortly.
The UL Re-transmission graph issue in AP was identified and will be fixed and updated in the cloud servers ASAP. The CPU % calculation was rectified in 3.1.1 after it was found it was not plotted properly earlier insync with the device values.