ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
2020-01-06 11:28:00 -0500 | received badge | ● Famous Question (source) |
2019-05-20 02:27:11 -0500 | marked best answer | Is there a package for topic diagnostics/message rate tracking and warning? I am interested in having something like a package with general nodes that subscribe to topics and publish warnings when messages aren't being sent or are coming in at the wrong rates and that publishes warnings when this happens. I have a multi-robot system that suffers from a wide variety of often easily-mitigated errors related to things like hardware issues, wireless issues, etc. The biggest issue is generally detecting these errors. All of the typical failures involve streaming data such as heartbeats, odometry, and camera data. An easy way to track these streams in a separate node would be incredibly valuable. If someone has not already written a tool for this, I will likely write one myself. |
2017-11-03 09:16:30 -0500 | received badge | ● Notable Question (source) |
2017-08-21 01:40:02 -0500 | commented answer | Is there a package for topic diagnostics/message rate tracking and warning? Thanks. This looks a little more complex than what I was hoping for, but it looks like it can be configured to meet my n |
2017-08-21 01:37:55 -0500 | received badge | ● Popular Question (source) |
2017-08-20 23:56:04 -0500 | received badge | ● Organizer (source) |
2017-08-20 23:55:13 -0500 | asked a question | Is there a package for topic diagnostics/message rate tracking and warning? Is there a package for topic diagnostics/message rate tracking and warning? I am interested in having something like a p |
2016-08-17 10:02:29 -0500 | received badge | ● Famous Question (source) |
2016-06-28 22:15:59 -0500 | received badge | ● Nice Answer (source) |
2015-10-27 15:20:16 -0500 | received badge | ● Famous Question (source) |
2015-10-27 15:20:16 -0500 | received badge | ● Notable Question (source) |
2015-06-22 01:31:58 -0500 | received badge | ● Popular Question (source) |
2015-06-09 03:48:13 -0500 | received badge | ● Self-Learner (source) |
2015-06-09 03:48:13 -0500 | received badge | ● Teacher (source) |
2015-06-03 13:01:27 -0500 | answered a question | rtabmap stereo with original (mono) Bumblebee The issue was that the left and right cameras were switched. Remapping each side in camera1394stereo fixed the problem. Rtab-map in Indigo warns that the baseline is negative which tipped me off. |
2015-06-03 12:58:56 -0500 | commented answer | rtabmap stereo with original (mono) Bumblebee I didn't get to updating earlier. Turns out that the left and right of the camera were switched. rtab-map in hydro does not give meaningful output for this. On indigo it warns that the baseline is negative. |
2015-06-03 00:15:43 -0500 | received badge | ● Student (source) |
2015-06-02 11:41:05 -0500 | asked a question | rtabmap stereo with original (mono) Bumblebee I am trying to set up an original Bumblebee stereo camera with rtabmap. Currently, the issue that I am running into is that stereo_odometry is crashing without giving a discernible error message. Relevant parts of my launch file look like: Camera: Rectification: Transform: Odometry: Any ideas what my issue could be? |
2015-02-20 13:22:13 -0500 | received badge | ● Taxonomist |
2014-09-30 12:09:06 -0500 | received badge | ● Popular Question (source) |
2014-09-30 12:09:06 -0500 | received badge | ● Notable Question (source) |
2014-09-24 12:13:22 -0500 | received badge | ● Enthusiast |
2014-07-09 11:31:13 -0500 | received badge | ● Famous Question (source) |
2014-07-09 11:28:02 -0500 | answered a question | Cleanly exit from python publisher ensuring all messages sent Answering my own question: I found my answer in the rospy source. There is a comment that explicitly states that sending of all messages before shutdown is not guaranteed. Instead, there is a check for whether the process is still running before writing, and if not it exits immediately. Here is an excerpt from |
2014-07-09 11:01:28 -0500 | asked a question | Does ROS-Matlab support parallel execution? I have been testing out the ROS-Matlab package and have had some issues with subscribers, which run in the background, being starved. Matlab's "Parallel Computing Toolbox" which I do not have supports various methods for implementing parallelism such as parallel execution of "for" loops via "parfor." Unfortunately, the actual implementation of the listener is in an obfuscated ".p" file. I have found no documentation related to parallelism in ROS-Matlab. Does anyone know anything related to parallelism in ROS-Matlab? |
2014-07-03 16:41:43 -0500 | received badge | ● Notable Question (source) |
2014-07-03 14:30:01 -0500 | received badge | ● Editor (source) |
2014-07-03 14:27:34 -0500 | received badge | ● Popular Question (source) |
2014-07-03 14:27:00 -0500 | received badge | ● Supporter (source) |
2014-07-03 14:12:14 -0500 | commented answer | Cleanly exit from python publisher ensuring all messages sent What I have been seeing is that some subscribers will get a message but not others so self-subscribing provides no guarantee (i.e. "rostopic echo" may get messages when another node doesn't) Good point about services, I was planning the switch. A delay may fix that problem, but it would be a hack. |
2014-07-03 12:48:32 -0500 | asked a question | Cleanly exit from python publisher ensuring all messages sent I have been working on a Python node that essentially publishes one-off messages and then exits. Due, I believe, to the threading in rospy, messages are not always sent if the script exits soon after publishing a message. I have also found no ways of truly guaranteeing that messages that sent messages are received. Here is some example code based roughly on the nodes demonstrated in the ROS tutorials: Publisher Subscriber: There are a number of modifications I have found that will get the messages to send:
There are also some exceptions where the message may otherwise be expected to be received:
On top of all this, I have found that I would like to know if there is any (hopefully documented) way that I can exit a python script with the guarantee that every subscriber to a topic that node publishes will have received every message that has been published. So far I have not been able to find anything to this effect. |