Out of range bug still present in 1.0.8

[reinzor]

Version: 1.0.8

terminate called after throwing an instance of 'std::out_of_range'
  what():  vector::_M_range_check: __n (which is 18446744073709551615) >= this->size() (which is 0)

Follow up of https://github.com/SICKAG/sick_safetyscanners/issues/58

Bug still seems to be present in the latest version

I am also seeing these warning messages coming by:

[ WARN][/front_scan][1662478255.752680645]: Skipping data, sizes do not match, actual size is smaller then expected size! If this occurs please report with a stacktrace if the driver crashes at some other place. 
[ WARN][/front_scan][1662478255.754574851]: Expected minimum size: 5464
[ WARN][/front_scan][1662478255.755154463]: Actual size: 5060
[ WARN][/front_scan][1662478255.755987363]: Skipping all data for this message.

[ WARN][/front_scan][1662492345.148584654]: Skipping data, sizes do not match, actual size is smaller then expected size! If this occurs please report with a stacktrace if the driver crashes at some other place. 
[ WARN][/front_scan][1662492345.148653641]: Expected minimum size: 5108
[ WARN][/front_scan][1662492345.148678426]: Actual size: 5060
[ WARN][/front_scan][1662492345.148694592]: Skipping all data for this message.

[ WARN][/front_scan][1662520262.934361006]: Field Number Beams has a value larger then the expected Number of Beams for the laserscanners. Skipping this measurement.
[ WARN][/front_scan][1662520262.934413448]: Max expected beams: 2751
[ WARN][/front_scan][1662520262.934442805]: Number beams according to the datafield: -1056956303

[ WARN][/front_scan][1662674352.304261438]: Field Number Beams has a value larger then the expected Number of Beams for the laserscanners. Skipping this measurement.
[ WARN][/front_scan][1662674352.304358731]: Max expected beams: 2751
[ WARN][/front_scan][1662674352.304468138]: Number beams according to the datafield: -1899954145

[ WARN][/front_scan][1662765657.677277037]: Field Number Beams has a value larger then the expected Number of Beams for the laserscanners. Skipping this measurement.
[ WARN][/front_scan][1662765657.677343470]: Max expected beams: 2751
[ WARN][/front_scan][1662765657.677374916]: Number beams according to the datafield: 2031291393

[ WARN][/front_scan][1662796320.543376042]: Skipping data, sizes do not match, actual size is smaller then expected size! If this occurs please report with a stacktrace if the driver crashes at some other place. 
[ WARN][/front_scan][1662796320.543490062]: Expected minimum size: 5067
[ WARN][/front_scan][1662796320.543518815]: Actual size: 5060
[ WARN][/front_scan][1662796320.543539003]: Skipping all data for this message.

[lenpuc]

Hi, I am still not sure where they originate from. Could you provide me with the stack trace of where the program is crashing. Otherwise I will not be able to implement a check to prevent the driver from crashing.

[reinzor]

Thanks for your quick reply. This happened on a release build so I do not have more information at the moment. When this happens again and I can obtain a stack trace, I will post it here.

[mechatheo]

I can confirm, we also keep seeing this error.

we found out that we can trigger this error quite reliably if we call the status_overview service in parallel. Increasing the service call frequency increased the crash likelihood for us.

Not quite sure though whether an error must be present. If i recall correctly we see it most often when the laser reports a contamination warning

[FabianSchurig]

Still an issue. 🚀

@puck-fzi are you able to reproduce it?

[lenpuc]

Hi together, sadly I am still unable to reproduce the error. I would need a stack trace to see where it happens.

If that is not possible, if you could give me a detailed description how you can reproduce the error that might be helpful as well. On my setup I can let the system run without crashes.

[lenpuc]

Hopefully, this is resolved through the PR https://github.com/SICKAG/sick_safetyscanners/pull/139

Please reopen if not.