Am I losing my F'n mind?! NX-2200 processors acting up.
feddx
Posts: 183
I have 2 different AMX systems. Both are running on NX-2200 processors on the latest FW and running NetLinx code I made from scratch. And BOTH of them will run for a few days, and when I come back, nothing is working until I reset at least the processor.
Both systems are for MDUs. Mainly display controls, tuners, and the audio routing therein.
In one system I am using a large number of networked DX-TX and RX 4K60s (connected 1 TX directly to 1 RX for video signal) and a couple of EXB-IR4s. I have 10 areas, each with a TP (MD/MT-702) and a display and a tuner. Some of the systems have extra inputs and a few other bells and whistles. The only serial control I'm doing on that site is for a DSP, a video switch, and a surround sound processor. I am running IP control to the LEA amps, and every tuner and every display is controlled via IR off of the controller and the EXBs (for the CATV tuners) and the DX-RX4K60s for the displays. I have been to this site many times for failures, and I always need to reboot the processor to connect with it and get control of it, and sometimes I need to reboot the DX-RX boxes as they are no longer passing video (I even have them rebooting every night at midnight to ensure they stay "fresh"). I thought maybe it was an errant processor, so I set up a USB drive on the back of the NX-2200 to collect logging data and when I go back (next week) I will pull the logs to see what is happening.
I was certain this was a one off, until I arrived at the other site today.
The second site is again an MDU, NX-2200, and several DX-TX and RX boxes, though this time they are just dumb boxes passing video. There are 12 displays all being controlled IP, 10 tuners being controlled IR, a DSP (IP control) 2 TPs, 5 KPs, and that's it. When I got the the site today after being away for just a few days, I found that things weren't working and I could not get access to the processor's web interface (to check to see if some displays were registered) without a reboot.
Is this something people are seeing commonly in AMX NX-2200 masters? (Master FW is 1.8.183, Device FW is 1.50) In any AMX master? Or am I just the victim of a messed up joke?
Thanks in advance.
"If there is no pain like a broken heart, why is it thinking of my divorce makes me so happy?"
Comments
you may (re)activate the SD card Audit logging, to see longer term what's happening on the systems.
The audit logging to the SD card is deactivated since fw 1.6.205 / 1.8.121, to prevent tearing out the sd card.
You could try resetting the NX from one site to factory default including default firmware (holding down the ID button for longer than 20 seconds while powering on). I'm assuming here that the factory default firmware will be v1.6.175, which has been really stable for me on dozens of sites. Or if that isn't the factory default, downgrade to it? Not sure if the device firmware will go back to v1.1.48 or if that is necessary.
The v1.6 firmware has been around a lot longer and is likely better field tested. Of course check first if there is something you really need from the new firmware, but I'm not sure for an NX that will be much .
If you have access to the NX console (telnet or USB serial device), you can even downgrade the NX to factory firmware by "manage firmware" console command ... no need of that "holding that ID button" thing
Old dogs, new tricks sort of thing
If its running the latest 1.8 firmware I've had similar on NX1200/2200
Basically it just stops working. Its pingable, you can't telnet into it, Netlinx studio can't connect but goes through the 15 second countdown then does nothing. Only thing that works is a reboot - whihc has to be a power cycle.
I've kind of tied it down to a memory leak. I was writing a module for Autonomics MMS and it worked perfectly on 1.6.205 and overnight left running was effectively locked out in the morning.
When I watch button presses in diagnostic, you can see it eating memory. I've reported it
Just go back to 1.6.205 if you have the 1.8 firmware - it just works. I only upgraded out of curiostity.
Hope that helps
I've also had nothing but problems with 1.8 and seen the same things where it will be responsive after reboot and 24 hours later its unresponsive. Timelines won't always start, duet modules don't always start. 1.8.183 and 1.8.192. looks like we're stuck with 1.6.205 since its the last 1.6 firmware. If you do roll back, make sure you factory default, then load factory firmware then upgrade to 1.6.205. otherwise due to the security changes you might end up locked out.
Are your systems M2M connected? I got reports with similar symptoms from a setup with 3 controllers in M2M, tested v1.8.183 and v1.8.192. Sometimes over night, but also sometimes just after 10 minutes the masters stop working, and they all (!) need a reboot to get to work again. Downgraded to 1.6.193 everything works stable again.
I first thought it may be a M2M issue, and so it would be interesting if your setups are also M2M, or standalone.
All ours are standalone, interestingly enough, I've got a 2265 running 1.8.183 and had little to no issues, which came from the factory with 1.8.74 firmware. Maybe it has something to do with the conversion from 1.6 to 1.8.
good point.. maybe when upgrading from 1.6 to 1.8, a factory reset is required, similar when downgrading from 1.8 to 1.6.
The only difference between NX and DVX4K is afaik that the NX firmware increases Duet mem from 64 to 128M when detecting it is running in a DVX4K...
Not sure if this helps but older modules were encrypted. Any master running fw 1.6.x is able to decrypt them while fw 1.8.x is not.
All of the modules on developer.amx.com have been decrypted now so they all work with 1.8.x FW but if you are using older modules that were downloaded before the change then your modules will not load. You would need to download a new copy of that module and load it to the controller
https://help.harmanpro.com/magic-value-error
Sorry for not getting back to you guys earlier.
THANK YOU!!
I think this is my issue. I reverted the processor to 1.6.179 (its factory FW) and seemingly no issues. It's odd in the MANY calls I've had with AMX, this hasn't been mentioned.