开云体育

I have beenhaving a problem with JMRI locking up during operating sessions.? A dispatcher panel using Panel Pro (version 4.10+R419243ewith version 1.8.0 for Java) is used to throw mainline switches and to keeptrack of train locations using manually activated sensors.? The track diagram is on an external monitorfor a laptop (Windows 10) that communicates at 9600 baud with an NCE DCC systemhaving the latest EPROM.? A typical scenariohas the OS starting at 1:15 pm, and about 45 min later the screen locks up andJMRI does not respond.? Usually, thereare 6 to 8 trains running.? I restart thelaptop and then JMRI to get back running and things work OK until the end ofthe OS at 5:00 pm.? I have copied below someof the lines from the file for the NCE Command Monitor.? These lines are for the times right beforeand after the lockup.? The last line is when I restarted JMRI.? Can you tell mehow to prevent JMRI from locking up??Thanks for your help.

Ted Smith

?

2018-03-07 13:56:57,708 ncemon.NceMonBinary? ?ERROR- Unhandled reply type code: 2 [AWT-EventQueue-0]

2018-03-07 13:56:58,132 ncemon.NceMonBinary? ERROR - Unhandled command code: 143[AWT-EventQueue-0]

2018-03-07 13:57:08,140 jmrix.AbstractMRTrafficController???? WARN?- Timeout on reply to message: 8F EC 10 consecutive timeouts = 0 innce.NceTrafficController [nce.NceTrafficController Transmit thread]

2018-03-07 14:02:17,430 util.Log4JUtil??????????????????????? INFO? - * JMRI log ** [main]



[Non-text portions of this message have been removed]

See below:

On 9 Mar 2018, at 1:12 PM, Ted Smith tfsmith365@... [jmriusers] <jmriusers@...> wrote:

2018-03-07 13:56:57,708 ncemon.NceMonBinary ERROR- Unhandled reply type code: 2 [AWT-EventQueue-0]

2018-03-07 13:56:58,132 ncemon.NceMonBinary ERROR - Unhandled command code: 143[AWT-EventQueue-0]

These two error messages are spurious and occur if you have the NCE Command Monitor open. This has been fixed in the V4.11.x series.

2018-03-07 13:57:08,140 jmrix.AbstractMRTrafficController WARN - Timeout on reply to message: 8F EC 10 consecutive timeouts = 0 innce.NceTrafficController [nce.NceTrafficController Transmit thread]

This message tells me the command station didn't reply to a memory read request (an 8F command), usually because the command station interface has locked up (which you already knew).

A very difficult problem to track.
--
Dave in Australia

The New England Convention 2018

Dave,

?

Thanks for your response.

?

To clarify, is thetimeout due to the NCE command station or to JMRI?

?

It is interestingto me that JMRI ran well for the 3 hours to the end of the OS after I restartedthe laptop and JMRI, even though the number of trains and activity level waspretty much the same as before the lockup.?I could not restart only JMRI because the com port was open and I do notknow how to release it other than restarting the laptop.

?

Do you have anythoughts on what I can do to keep from having the lockup problem?

?

Ted in Iowa City



[Non-text portions of this message have been removed]

The command station locks up, not JMRI.

We have a similar situation sometimes with a decoder "Read All Sheets" and it can be unlocked by power-cycling the NCE command station WITHOUT touching or quitting JMRI.

Try that when you get a lockup to see if it frees things up.

For a long time we believed the "Read All Sheets" lockup was an NCE command station issue, but we then discovered an as-yet-unresolved circumstance where JMRI can send overlapping CV read commands.

So at this stage we could not make a call on your issue. Please try the command station power cycle, report back and then please raise an issue at <> so we have some background information and can get back to you for testing if needed.
--
Dave in Australia

The New England Convention 2018