Metric mapping attempts to insert sweeps into mongodb

[heuristicus]

After completing a sweep action (do_sweep), there is an attempt to add the cloud into mongo, resulting in a couple of error messages:

bettyr_teardown.launch:

[ INFO] [1479114795.523293524]: Created semantic map summary xml.
[ INFO] [1479114795.524244875]: Summary XML created. Looking for metaroom for room with id: WayPoint19                                                                                                       
[ INFO] [1479114795.524285712]: Loaded 5 metarooms.
[ INFO] [1479114795.524309526]: Matching metaroom found. XML file: /home/strands/.semanticMap/metarooms/metaroom_0/metaroom.xml                                                                              
Loading complete cloud file name /home/strands/.semanticMap/metarooms/metaroom_0/complete_cloud.pcd
[ INFO] [1479114795.724186012]: Room "/home/strands/.semanticMap/20161114/patrol_run_41/room_2/room.xml" inserted with id 5829802bb31be85ea0b7a3d5
Constructed surfel map for /home/strands/.semanticMap/20161114/patrol_run_41/room_2/room.xml
Loading interior cloud file name /home/strands/.semanticMap/metarooms/metaroom_0/interior_cloud.pcd                                                                                                          
Loading denoised cloud file name
/home/strands/.semanticMap/metarooms/metaroom_0/denoised_cloud.pcd
[ INFO] [1479114795.982732257]: Reading observations from /home/strands/.semanticMap
[ INFO] [1479114796.082882090]: Done looking for observations. Found 183 observations.
[ ERROR] [1479114796.713614256]: Service call failed: service [/message_store/insert] responded with an error: error processing request: BSON document too large (143260710 bytes) - the connected server supports BSON document sizes up to 16777216 bytes.
[ INFO] [1479114796.730759968]: Complete cloud "20161114_patrol_run_41___2___complete_cloud" inserted with id

The error inserting is repeated in the UI launcher as well.

The most likely culprit looks like sweep_save.py

[heuristicus]

@jayyoung will check tomorrow morning to see if that is the correct place to look.

[nilsbore]

This is a known error coming from MongoDB's 16MB limit. I'm not sure if Rares has any intention of fixing this but it doesn't effect robot behaviour.

[RaresAmbrus]

The script tries to log everything from the sweep, including the complete cloud which is above the file size limit. Just commenting out the line where the complete cloud is sent to MongoDB should remove the error, but as Nils pointed out, it doesn't affect anything else, it's just Mongo complaining that logging a file that is too large was attempted.