Program not running?

vining

Got call from a customer a few hours ago and he reported the R4's weren't working but he said the TPs were. Not sure about that but I VPN'd into the house and tried to log into the Zigbee gateway but couldn't so I rebooted it and then I could log in no problem and it showed in the online tree, cool, I thought, all fixed. Remotes still didn't work or show online on the remote itself which works through code so I VNC'd into a TP and that was offline too. VNC'd into another TP, offline too and went to the set up pages which showed online but still my online indicater that runs through the program didn't, Hmmm. Check the online tree in NS and everything still shows up but still nothing works. Rebooted the main master still nothing, checked some constantly changing time variables in debug but they don't register or work. So now I'm thinking my program isn't running so I resend the program, still nothing, resend firmware and nothing. So I telnet in check the dipswitches and 1-5 off 6-8 on, normal. Show program info and nothing, no programs. Now this program has been working fine for years and it just stopped running the program for no apparent reason. All the communications work, button pushes, everything seems fine and if I didn't know better I say a gremlin flipped the prd switch but the PRD dispswitch shows to be off in telnet. Any hidden soft prd switch that I can try in telnet. No storm or lightning and nothing out of th ordanary happened according to this customer and he would tell me if he did something stupid so its just one of those fluke things, I guess. Is there like a 3 year diode tied to the PRD switch that could fail and my reboot today for a smaller problem created a bigger problem.


I guess I get to go there tomorrow now and swap this NI-3100 out but if anyone thinks of anything worth trying I'm all ears.

DHawthorne

I'm pretty sure I've seen something like this related to a device buss problem. The device might show online, but is not processing commands, eventually taking down the message queue. Try rebooting it and Telnetting in as soon as possible to see what kind of messages pass before it stops running.

The other thing is memory. Got comfortable margins in all your memory types? Running out of non-volatile or disk memory can do this too. When I was developing my logging program, I ran into this a few times when the logs got too big. I have a parameter that checks for free space and stops the logging then it gets too low, but I have discovered that that number varies with the system. Bigger and busier programs need more disk space free, it seems. If you have something writing to disk, it might be eating up needed free space.

vining

I was leaning towards a com bus porblem and when I swapped out the NI3100 with a back up NI4000 pre-loaded with the same program it acted the same way. No com ports were talking but when in the shop the tx lights were flashing (ports not connected to anything). With the back up Master still installed I pulled all the com port connections thinking that soemthing was bringing down the bus but still nothing. So master/dev firmware updates, clean disc and alot of cursing and then the program reload and she started to talk (tx only since everything was still disco'd). Connected the com ports back up and most things worked but after more pulling of hairs I decided my Master & Dev firmware were out of wack and I tired of trying to figure out which dev firmware went with what master firmware so re-installed the original NI3100 master less com ports, ran clean disc, reloaded the program and tx lights started flashing. Connected the com ports and everything seemed happy. So maybe it was memory and the clean disc did the trick but it was very bizarre. This was the main master of a 3 master system and I wasn't running your devtermlogger or storing much more than a handful of files whose sizes change only slightly. I think my various memory types are good but that's one of many areas I still don't understand all that well but systems been reliable for years and apparently just had a itch (glitch) that needed to be scratched.

The system does check in every day but maybe I need to update it with your logger so it can send me more than just a daily hello. I believe I started using your logger mod after this job.

DHawthorne

Come to think of it, since the last update, I have had a few isolated incidents when the master firmware and device firmware (for the controller chassis) weren't in sync. I have also had a few incidents where loading and reloading did nothing at all until I did a "clean disk" then loaded again. I really hate it when stuff fails and you don't know why, but in a way it's even worse when it starts working again and you don't know why (I have a few of those out in the wild too, and they make me very nervous).