Hello Jakob, Thanks for the bug report! I'm working on a new mobility for the scenario, so I'll include the changes you pointed out too.
> Also, as far as I can tell from google and bing maps, there are fewer lanes > within the roundabout and some of its approaching edges. This looks like an > importer bug that was fixed in 0.25.0. The network should probably be rebuilt. > Note, that I did not check other roundabouts but the problem likely applies > to more than one. For the record, LuST Traffic is a realistic but simplified version of Luxembourg, so the fact that the they don't match, it's quite common in many places. I't not only a matter of topology but of mobility too. Anyway, I'm working on it! All the Best! Lara --- Lara Codecà PhD Candidate --- Interdisciplinary Centre for Security, Reliability and Trust Office: Weicker Site, C026 > On 19 Jan 2016, at 09:30, Jakob Erdmann <[email protected]> wrote: > > Hello Matej (and Hello Lara), > thank you for reporting the problem again. I fixed another bug in [19749] > and the crash you were seeing should be gone now. > As to the teleports, this has its roots in the LuST scenario itself (without > re-routing it is simply more pronounced) so I'm including Lara in my response: > > Jamming and teleporting: > - the roundabout <roundabout nodes="-10122 -15272 -25332 -26768 -7046 -7348 > -8018 -8304" edges="-32770#0 -32770#1 -32770#2 -32770#3 -32770#4 -32770#5 > -32770#6 -32770#7"/> > has unsuitable connectivity. This is causing lots of deadlocks by vehicles > that need to change lanes in opposite directions. > I'm attaching a screenshot of what it should look like instead. Basically, > all lanes within the roundabout should allow to continue within the > roundabout. > > Also, as far as I can tell from google and bing maps, there are fewer lanes > within the roundabout and some of its approaching edges. This looks like an > importer bug that was fixed in 0.25.0. The network should probably be rebuilt. > Note, that I did not check other roundabouts but the problem likely applies > to more than one. > > Other things: > - the option --lanechange.allow-swap is deprecated (and known to be buggy). > In the scenario it is causing "emergency-stop" warnings > - the option --device.rerouting.pre-period at it's default value of 1 is > slowing the scenario down very much. (there are plenty of vehicles with > delayed departure). Either these vehicles should be discared > (--max-depart-delay) or the pre-period should be increased > > best regards to both of you, > Jakob > > > 2016-01-18 22:36 GMT+01:00 Matěj Kubička > <[email protected]>: > An update: SUMO does not die on segfault with rerouting disabled, but it does > not work properly. Vehicles start teleporting massively after 600 seconds > (time-to-teleport). Summary reads: > > Simulation ended at time: 100000.00vehicles TOT 241899 ACT 67017 > Reason: The final simulation step has been reached. > Performance: > Duration: 20983445ms > Real time factor: 3.33596 > UPS: 166300.532682 > Vehicles: > Inserted: 241899 (Loaded: 295979) > Running: 67017 > Waiting: 54080 > Teleports: 265576 (Collisions: 22, Jam: 146278, Yield: 88302, Wrong Lane: > 30974) > Emergency Stops: 7 > > Notice the count of teleports (265576), it was supposed to be about 12. > > ~Matej. > > > On 18.1.2016 16:16, Matěj Kubička wrote: > Hello, > the problem persists in 19716. It is much less likely to occur, but it > still happens. I run the same setup as last time and SUMO segfaults > after ~600 seconds. It also gives same nonsense warning(s) prior to the > segfault. Stdout: > > Simulation started with time: 30000.00 > Step #30497.00Warning: Vehicle 'h53498c1:1' performs emergency stop at > the end of lane '-31232_1' because there is no connection to the next > edge (decel=-22.99), time=30497.00. > Step #30587.00 (992ms ~= 1.01*RT, ~5135.08UPS, vehicles TOT 68989 ACT 50 > Segmentation fault (core dumped) > > ~Matej. > > On 12.1.2016 15:38, Matěj Kubička wrote: > Thank you for fixing it! I really appreciate it. > Matej. > > On 12.1.2016 15:35, Jakob Erdmann wrote: > Hello, > the problem is fixed in the development version 19716 > (http://sumo.dlr.de/wiki/Downloads#SUMO_-_Latest_Development_Version) > You cannot use old state files and must recreate them with this > version instead. > Thank you for reporting the problem. > > regards, > Jakob > > > 2016-01-08 18:20 GMT+01:00 Matěj Kubička > <[email protected] > <mailto:[email protected]>>: > > Maybe the problem is not in the state file - it runs fine from loaded > state if I omit the --begin parameter. Also, there seem to be > different > state file formats between 0.24 and 0.25 and the problem exists in > both > versions.. > > Thank you, > Matej. > > On 8.1.2016 13:36, Matěj Kubička wrote: > > Here goes the command that generated the sate file: > > > > sumo -c LuSTScenario.complete.mobility.sumo.cfg > --save-state.times 30000 > > -e 30010 > > > > Matej. > > > > > > On 8.1.2016 12:44, Jakob Erdmann wrote: > >> It looks as if the state file contains inconsistent data. > >> The fastest work-around would be to avoid saving/loading states > until > >> the problem is fixed. > >> If you can send me the command that was used to create the > state file > >> I can try to figure out what goes wrong. > >> regards, > >> Jakob > >> > >> 2016-01-07 23:22 GMT+01:00 Matěj Kubička > >> <[email protected] > <mailto:[email protected]> > >> <mailto:[email protected] > <mailto:[email protected]>>>: > >> > >> Update: the problem persists with 0.25.0. Also, if I omit > the --begin > >> parameter it won't die, but all the vehicles will teleport. M. > >> > >> On 7.1.2016 22:11, Matěj Kubička wrote: > >> > Hello, > >> > I have been trying to run sumo from saved state, but it > dies first > >> > simulationStep() after initialization. Its is killed on > SIGABRT > >> after > >> > OS detects double free() or memory corruption. There is also > >> plenty of > >> > nonsense warning statements prior to that so we probably > deal with a > >> > dangling pointer messing with the memory. > >> > > >> > I am using sumo 0.24, built about 2 months back using > sources from > >> > SourceForge. I know there is a new version, but I didn't > get to > >> update > >> > it yet. It was ran on Linux (3.13.0-74-generic kernel). > >> > > >> > The scenario and saved state is available for download from: > >> > https://idata.lss.supelec.fr/index.php/s/mvHTQxVZ9sdwwCw > >> > > >> > You can run it using "sumo -c > >> LuSTScenario.complete.mobility.sumo.cfg > >> > --load-state state_30000.00.sbx -b 30000". I've verified > that > >> this is > >> > with correct config file. > >> > > >> > For its output see enclosed stdout dump. I've ran valgrind's > >> memcheck > >> > on it so you can see where the problems originate from. > First > >> invalid > >> > read is in MSLane::succLinkSec, folowed by warnings of > type 'Vehicle > >> > performs emergency stop [...] because there is no > connection to the > >> > next edge'. > >> > > >> > Thank you, > >> > Matej. > >> > > >> > PS: Being a proper PhD student I am late to finish my > >> experiments. So, > >> > if there is a quick solution to this I would be happy to > hear about > >> > it.. :-) > >> > > >> > > >> > > >> > > >> > > >> > > ------------------------------------------------------------------------------ > >> > > >> > > >> > _______________________________________________ > >> > sumo-user mailing list > >> > [email protected] > <mailto:[email protected]> > >> <mailto:[email protected] > <mailto:[email protected]>> > >> >https://lists.sourceforge.net/lists/listinfo/sumo-user > >> > >> > > ------------------------------------------------------------------------------ > >> _______________________________________________ > >> sumo-user mailing list > >> [email protected] > <mailto:[email protected]> > >> <mailto:[email protected] > <mailto:[email protected]>> > >> https://lists.sourceforge.net/lists/listinfo/sumo-user > >> > >> > > > > ------------------------------------------------------------------------------ > > _______________________________________________ > > sumo-user mailing list > > [email protected] > <mailto:[email protected]> > > https://lists.sourceforge.net/lists/listinfo/sumo-user > > > > ------------------------------------------------------------------------------ > _______________________________________________ > sumo-user mailing list > [email protected] > <mailto:[email protected]> > https://lists.sourceforge.net/lists/listinfo/sumo-user > > > ------------------------------------------------------------------------------ > Site24x7 APM Insight: Get Deep Visibility into Application Performance > APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month > Monitor end-to-end web transactions and take corrective actions now > Troubleshoot faster and improve end-user experience. Signup Now! > http://pubads.g.doubleclick.net/gampad/clk?id=267308311&iu=/4140 > _______________________________________________ > sumo-user mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/sumo-user > > ------------------------------------------------------------------------------ > Site24x7 APM Insight: Get Deep Visibility into Application Performance > APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month > Monitor end-to-end web transactions and take corrective actions now > Troubleshoot faster and improve end-user experience. Signup Now! > http://pubads.g.doubleclick.net/gampad/clk?id=267308311&iu=/4140 > _______________________________________________ > sumo-user mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/sumo-user > > > <LuST_roundabout_fixed.png> ------------------------------------------------------------------------------ Site24x7 APM Insight: Get Deep Visibility into Application Performance APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month Monitor end-to-end web transactions and take corrective actions now Troubleshoot faster and improve end-user experience. Signup Now! http://pubads.g.doubleclick.net/gampad/clk?id=267308311&iu=/4140 _______________________________________________ sumo-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/sumo-user
