[VicenteIR]

Hi!

One of trains in my Timetable is stuck by Deadlock reason at station entrance. The F5 HUD screen shows that the train 266 stops at Entrance signal regards to active deadlock by train 156. Train 156 is at station stop with $wait=[train]/notstarted command for another train behind it. Each train has its own path at location but they both has a passing path at common station line in opposite directions. 156 is arriving to the station line. A station line of 266's path is empty, a "passing path line" is empty too.

https://i.ibb.co/G2yykc3/wait-issue.png
Both trains are AI trains in this Timetable Activity. When I was running the Timetable before with other train as the Player Train, it was not any issues at this location with those AI Trains - 266 was always entering to the location as defined in Timetable.
https://i.ibb.co/SvSFVQ1/elvas.png

I attaching two log files: one - restarting from save - because the situation occurs 4.5 hours after the Player service starts, and the second - a part from the start with the paths parsing.

Regards
Oleg

Attached File(s)

•  OpenRailsLog.txt (31.09K)
  Number of downloads: 214
•  OpenRailsLog1.txt (19K)
  Number of downloads: 189

[roeter]

Well, given the number of warnings in the log-file you have some cleaning-up to do, but I do not think any of those have any significance for this situation.
There's three things that puzzle me.
First is that there is a deadlock set, for allthough the two trains do have a (short) common section in opposite directions, this section is protected by signals but without any signals along the common section and this should prevent the deadlock from being activated, with the deadlock processing leaving the handling of the trains to the signals.
The second puzzle is that train 156 seems to have sprung a deadlock while in 'wait' state. That should not be possible.
The third puzzle is that you state that it did work until recently. There have been no significant changes to the timetable code for over a year now, so that can not now have caused this problem. The only thing I can think of is that in running another player train might have just changed the timing of some of the trains involved. Sometimes, just a few seconds difference in timing can cause problems, with the sequence of path allocation changing because of that small difference.

The only thing I can think of to avoid this deadlock is to set an explicit wait for train 156 at this location, to have it wait for 266.

Regards,
Rob Roeterdink

[Gro.Bi]

May it be that train #156 is a 'non-existent train' because it has no separate line in the F5-list?

@Rob: is it possibly an issue comparable to what is discussed at 'non-existent train in timetbale-mode'?

There was also a problem about a train that already has left the simulation, but still caused problems to actual trains.

Regards,
Gro.Bi

[roeter]

Gro.Bi, on 13 April 2021 - 03:03 AM, said:

May it be that train #156 is a 'non-existent train' because it has no separate line in the F5-list?

@Rob: is it possibly an issue comparable to what is discussed at 'non-existent train in timetbale-mode'?

There was also a problem about a train that already has left the simulation, but still caused problems to actual trains.

Regards,
Gro.Bi

Did you get an error report on train 156? If not, then it should still exist allthough it is strange that it's not shown in the F5 hud.
One thing which might help to give some clue as to what is going on is to run that '156' as player train, and check out what happens to it.
Other useful information could come from sitting around as an observer and check on what goes on with '156'. I know this will all cost time but at this stage I have no clue as to what happened here.

Regards,
Rob Roeterdink

[VicenteIR]

No, I didn't. The train is shown in the F5 Hud, but not in the same page with 266 =)

I unchecked a "Forced Red at station stops" - was checked by mistake... and tested the situation again with the same train as a player train. Unexpected, 266 does get clear state and does arriving (get a little delay and a previous train - 264 - moved first). Deadlock in this test "moved" to train 271 behind 266 and this time 271 was stuck.
This test shown in the first video. Please, take a look, behind 156 train 158 stops at Entrance signal and doesn't using a Passing Path station line.
https://www.youtube....h?v=uJ9rCOIS4PY


The second video shows a test with an "observe" train as a Player Train. That train is far from the station (about 80 km) and have not common track sections with any path in Timetable. Actually, this is what I expecting from the action in this station at that time. All AI traffic runs without any problem.
https://www.youtube....h?v=gkMdWYtijGM


Attached log file is for "observe train" as a player train Activity.

Rob, I can understand that the Program can not accept my method to have a "spare" station line with opposite passing paths. The question is why when a player doesn't passing the station, everything working properly, otherwise I have active Deadlock "over" signal at STOP state.

Regards
Oleg

Attached File(s)

•  OpenRailsLog1.txt (74.07K)
  Number of downloads: 283

[VicenteIR]

Now I running the Activity with 156 as the player train. Waiting for 6116 at location.
https://i.ibb.co/CzJLmXK/Run-Activity-2021-04-15-02-34-58-37.png
271 and 158 are already here =)

Regards
Oleg

[VicenteIR]

It seems that I am in a big troubles!
When I observing the Timetable, everything is ok. When I running a train in Timetable, for this train, it looks perfect. But what happens behind it - it is really mess.
At starting location of train 156 there are 3 trains stuck in similar situation, one of them doesn't stop, "slips" through a signal, passed it at danger and was removed from Activity for SPAD and out of control reasons.

Quote

Warning: Train 2032:monday (160) passing signal 1172 at 346.2 at danger at 0.18

Information: Train 2032:monday (160) is removed for out of control, reason : SPAD

Upd: I'll remove a passing paths from 156's path and will try again.

Gro.Bi, thank you for your answer and for the link. It really looks like similar problem and as you can see, finally, I get an "non-existant" train too =)

roeter, on 05 January 2020 - 11:36 AM, said:

Problem found - well, two problems, actually, and in this case the first problem triggered the second.
Will work on a solution, but meantime don't use passing paths through stations, things could go wrong.

Regards,
Rob Roeterdink

Rob, is this still relevant? Am I need to remove all of the passing paths from the trains paths in my Timetable?

Regards
Oleg

[Gro.Bi]

I found numerous lines in the attached log-file 'Information: Invalid passing path defined for train XXX at section XXX: overlaps with other passing path'. I have this already found in own timetables and guess that then a passing path will never be used by the named train. The section XXX can be found in the trackviewer by 'File/Search by Index/Search track node'.
@Rob: please revise my statement if it should be incorrect.

Regards,
Gro.Bi

[VicenteIR]

Oh... Thanks! It was helpful. I founded the points. And even was check with one of spoken trains- 271 (has a warning at point 564 - exactly in that locatiom). You are absolutely right. The default path track of AI 271 was occupied by another AI and the train doesn't enter to empty alternative path. I didn't pay my attention on it before because trains often even doesn't stops in those situations in my Timetable. Also in this case 271 doesn't stop at the Red signal. 266 moving outside and 271 slips and gets Clear state without stop.
And what can you say about a game speed increasing? I don't know why, but has any idea that things could go wrong with it. Yesterday night I came back and ran 156 without using game speed increasing and this:

VicenteIR, on 14 April 2021 - 04:07 PM, said:

At starting location of train 156 there are 3 trains stuck in similar situation, one of them doesn't stop, "slips" through a signal, passed it at danger and was removed from Activity for SPAD and out of control reasons.

was not happened!
My next step - to come back to the train from the starting post, run my Timetable Activity with it as a player train at normal game speed. It will take a while (days maybe). Will report.

Regards
Oleg

[VicenteIR]

VicenteIR, on 15 April 2021 - 10:55 PM, said:

My next step - to come back to the train from the starting post, run my Timetable Activity with it as a player train at normal game speed. It will take a while (days maybe). Will report.

Doesn't help! Also removing the passing paths from 266's and 271's paths was leaving a Deadlock been active from AI 156. (Running an Activity with 156 as a player train doesn't detect any problems - any of both: 266 and 271 getting active Deadlock there).
What does help - I was modified a path of 156 itself by removing a passing path at location and returned to 266 and 271 those original paths with a passing paths - no Deadlock was activated.
Also a warnings for those trains at section 564 doesn't appear anymore.

It would be very helpful if in a warning message rather "overlaps with other passing path " has a specific train pointed to.

Regards
Oleg