RT ocasional error

[TerraNova]

Again & again & again !

This funny behavior isn't really funny :)!

Ocasion happened again...

Take a look at this logs!

[eek]

Hello,

I've looked at your error file and it looks as though you're experiencing a DWarn error which means that something unexpected but not fatal has occured on your system.

If this is actually hindering you from running your RT system, you should probably call or email NI for direct support. We can then directly troubleshoot this tricky behavior and look to find you a resolution. You can find out how contact NI here.

We look forward to hearing from you soon.

Cheers,

Emilie K | National Instruments

[TerraNova]

Thanks, Emilie.

And the question is WHAT could be a reason for such unexpected RT system event.

The wander is in the fact that in previous versions (<8.2.1) it was absolutely fine with THE SAME LV code :).

Anyway, thanks a lot, Em.

[Sima]

Sharonoff, can you describe what your code does? Maybe post it? Like I said earlier, when we upgraded to 8.2.1 our same code seemed to be tripping on the timed loops, and when we replaced those with regular while loops, the problem has not reoccured since.

Do you see any other errors besides the the exception code? Are you watching the VGA output of the RT system when it crashes? Do you also get any kernel errors at all?

[TerraNova]

Thank you, Sima!

No more on VGA - just those things I've written in the very first post.

As far as I can see, EVERY activity on RT stops after.
RT doesn't reboot, doesn't answers on PING. But!

After manually turning off and turning on it boots up RT along with the sentence on VGA: "Reboot due to system Error". Then it works fine until next crash (Enjoy;)!!!).

... And, about Timed Loops ...
I really and actively use them!
Problem is that 0,5 ms RT-determinism is needed, that's why I'm afraid I won't be able to replace them with regular ones.

Thanx for your feedback!!!

P.S.: 8.5 just arrived, but I have serious doubts about wether it'll get me rid of this issue (see second post) :(.

[TerraNova]

And, by the way, Sima! Would you please be so kind to send me those logs from your RT you mentioned some post earlier! Thanks.

[Sima]

Sharonoff, my error logs pointed specifically to timed loop threads, which directed our debugging trail that way. I can't say that you and I experienced the same problem, since I was told that the exception error 0xC0000005 is a generic error in an RT crash, but you could try to replace your timed loops with while loops to see if that at least eliminates the crash. The workaround we got from NI was to use while loops and the RT wait timers (not the regular ones). According to them, this should provide pretty good timing without the overhead of the timed loops, unless you need to use any of the extra features that timed loops present.

Another thing, what's your CPU & memory usage on the RT? Are they high? Stable?

[Sima]

By the way, are those the complete error logs? They look to be truncated. Towards the end of the log there is usually some memory info.

[TerraNova]

Hi!

Thats those truncated logs!

...And a string from it...
Thread name: LV TL[50100]

Is this about Timed Loop?

Thank you in adv.

Message Edited by Sharonoff on 11-15-2007 10:29 AM

[TerraNova]

Hope, this log woud be a use for Emilie K either
🙂

Thanks!