(messenger-libusb.cpp:42) control_transfer returned error, index: 768, error: No data available, number: 61 IntelRealSense #2028
Comments
|
Hi @SarenWraith Could you post the launch log from the terminal that is generated by the roslaunch instruction in a comment below, please? Thanks! |
|
I hope I am doing this right, I ran the command line
... logging to /home/cthulhu/.ros/log/4b79829e-f8f2-11eb-ac36-c006c3430634/roslaunch-jetson-8974.log started roslaunch server http://172.17.0.1:44321/ SUMMARYPARAMETERS
NODES auto-starting new master setting /run_id to 4b79829e-f8f2-11eb-ac36-c006c3430634 [ INFO] [1628500514.905290205]: Device with physical ID 2-1.1-3 was found. [ INFO] [1628500521.115738835]: Device with physical ID 2-1.1-3 was found. |
|
Yes, that is the correct log. Thanks very much! Your librealsense SDK and RealSense ROS wrapper versions are correctly matched, and depth and color streams are being published. Are the control_transfer returned messages only appearing a small number of times during launch and then stopping, or at a rate of once or twice per minute after launch? If so then they can be safely ignored. However, if they are being generated continuously without stopping then this can indicate a serious communication problem with the device. |
|
I am afraid they are being generated continuously without stopping. This is all being done on a jetson nano, which has a limited computing power. Is it possible the problem is due to the computing power? Or is there something more sinister going on? Also the command |
|
Jetson boards are very capable, so I would not strongly suspect processing limitations to be the problem. You can though check CPU usage percentage in the Ubuntu OS whilst an application is running by using a system monitoring application such as htop though. https://vitux.com/how-to-use-htop-to-monitor-system-processes-in-ubuntu-20-04/ However, a power supply related issue could potentially cause problems if there is power instability at the USB port. Do you have your camera directly connected to the Nano board (so that the camera's power needs are being supplied by the Nano) or are you using a USB hub, please? And if you are using a hub, is it a 'powered hub' that is powered by a mains electricity socket? |
|
I have the camera connected directly to the Nano, but I have been powering the Jetson Nano from the usb port rather than the barrel jack, I will attempt a run with a barrel jack connector and update you. |
|
The issue was not resolved but the error log looks a little different now: ... logging to /home/cthulhu/.ros/log/c29f2cfc-f8f7-11eb-933e-c006c3430634/roslaunch-jetson-7953.log started roslaunch server http://172.17.0.1:41137/ SUMMARYPARAMETERS
NODES auto-starting new master setting /run_id to c29f2cfc-f8f7-11eb-933e-c006c3430634 [ INFO] [1628502862.235032717]: Device with physical ID 2-1.1-3 was found. [ INFO] [1628502868.489244329]: Device with physical ID 2-1.1-4 was found. |
|
Is there any improvement if you set RViz to use compressed data instead of raw data, like in the link below? |
|
I just tried your suggested solution, I am afraid that the image still sends a No image received error. However in my case the fixed frame (under the global status tab) shows the error Fixed Frame [map] does not exist. The error summary remains the same, alas. |
|
Do you get the same warning messages if you leave initial_reset:=true out of the roslaunch instruction? Do you also still get the Fixed Frame error if you set Fixed Frame to camera_link instead of map?
|
|
Sorry for the late reply, The good news is that the global status is showing a green check mark as well as the fixed frame for the system. There is no image showing, however, and the error log seems much the same roslaunch realsense2_camera rs_camera.launch filters:=pointcloud depth_fps:=6 color_fps:=6 started roslaunch server http://172.17.0.1:42487/ SUMMARYPARAMETERS
NODES auto-starting new master setting /run_id to 0d789d26-f9a2-11eb-a46d-c006c3430634 [ INFO] [1628576000.825680893]: Device with physical ID 2-1.1-3 was found.
|
|
Interestingly enough, while the realsense viewer does work, it also outputs a similar error log 10/08 10:22:33,891 WARNING [548195520960] (messenger-libusb.cpp:42) control_transfer returned error, index: 300, error: No data available, number: 3d |
|
The ROS wrapper is overriding your custom roslaunch configuration of 6 FPS and using 30 FPS for depth and color. This is likely because since ROS wrapper version 2.2.22, the wrapper requires a custom stream definition to have three factors specified for each stream type used: width, height and FPS. Otherwise, the configuration is deemed invalid and the default configuration of that particular camera model is applied instead. Please try the roslaunch instruction below: roslaunch realsense2_camera rs_camera.launch filters:=pointcloud depth_width:=848 depth_height:=480 depth_fps:=6 color_width:=1280 color_height:=720 color_fps:=6 |
|
Good news is I am finally seeing an image on RVIZ, but the error log eventually returns to the 768 index after a while.
started roslaunch server http://172.17.0.1:42105/ SUMMARYPARAMETERS
NODES auto-starting new master setting /run_id to e9a57a58-f9a7-11eb-9f65-c006c3430634 [ INFO] [1628578519.483094149]: Device with physical ID 2-1.1-3 was found. |
|
Ok so I decided to try [ WARN] [1628578987.926859165]: Failed to meet update rate! Took 0.012039967000000000377 |
|
Possibly 6 FPS is not delivering frames quickly enough for RTABMAP. Do you see improvement if you change 6 to 15 in the roslaunch instruction, please? |
|
Ok so good news, the camera is showing a feed at a decent frame rate and the IMU is being processed as I can see it changing.
The map is not forming however like in the Here is the Error log: 10/08 11:14:31,428 WARNING [547667042752] (messenger-libusb.cpp:42) control_transfer returned error, index: 300, error: Resource temporarily unavailable, number: b started roslaunch server http://172.17.0.1:38497/ SUMMARYCLEAR PARAMETERS
PARAMETERS
NODES auto-starting new master setting /run_id to b226122e-f9aa-11eb-b5ea-c006c3430634 [ INFO] [1628579716.640465688]: Device with physical ID 2-1.1-3 was found. |
|
I note that in your launch logs, the gyro and accel topics are disabled. They are disabled by default from ROS wrapper 2.2.15 onwards. To enable them so that they are published, add the instructions below to your roslaunch instruction: enable_gyro:=true enable_accel:=true For example: roslaunch realsense2_camera opensource_tracking.launch filters:=pointcloud depth_width:=848 depth_height:=480 depth_fps:=15 color_width:=1280 color_height:=720 color_fps:=15 enable_gyro:=true enable_accel:=true |
|
I tried your command and when I did it the camera no longer functioned. I am nit sure why this happened. Hoever, the camera did track itself rotating
|
|
If you wish to use the opensource_tracking launch file to take advantage of the IMU of the D435i then Intel's D435i SLAM tutorial may be a good guide to use if you have not done so already (remembering to enable gyro and accel in the launch instruction, as the guide does not include that information) https://github.com/IntelRealSense/realsense-ros/wiki/SLAM-with-D435i |
|
I actually started with this guide before all of this, I followed the steps to install the imu-madgwick, the robot localization, and the rtab-map. There are improvements now that I have received your help, but the picture shown in the guide is still far off from what I currently have. |
|
If your setup is based on the SLAM D435i guide then may I confirm the following points with you:
|
|
Sorry for the late reply. Alright so I am receiving an image on the image pad on the RVIZ, obviously the IMU is no longer functional as it was checked off. My question is how to obtain the map shown intel SLAM guide?
000011981 https://github.com/IntelRealSense/realsense-ros/wiki/SLAM-with-D435i |
|
There is little documentation available about the Failed to meet update rate! error. When ROS does generate this error though, it sometimes provides this advice: "Try decreasing the rate, limiting sensor output frequency, or limiting the number of sensors". I believe that 'sensor output frequency' is referring to the IMU frequency. Like with setting an FPS for the image streams, RealSense ROS supports setting a 'frequency' speed for the gyro and accel streams in the roslaunch instruction. The minimum supported frequencies are 200 for gyro and 63 for accel. gyro_fps:=200 accel_fps:=63
Advice about the Not enough inliers warning is provided by Doronhi the RealSense ROS wrapper developer in the link below. Doronhi advises adjusting the position of the camera to see whether the inlier warning goes away. In regard to Out of frame resources! and the Exception was thrown during user processing callback warnings, Doronhi provided the advice below in #1161 "The "Out of frame resources!" message and the "Exception...processing callback!" could be related to lack of system resources - the app can't keep up with the incoming flow of images. For that you can reduce frame rate and resolution. Building with CUDA support should really help in that case on NVidia machines". |
|
Apologies but I won’t be able to respond as my office closed for the time being. I’ll respond again in two days |
|
No problem at all, @SarenWraith - thanks for the update! |
|
It worked!! For the record what I did was change the depth cloud and the color image under it to image_raw.
Alright but is there a way for the camera to retain that the wall it mapped is at this postion, because as I move it it is constantly forgetting the area it mapped. |
|
The link below contains a guide based around Intel's D435i SLAM tutorial for ROS that was created by a RealSense ROS user and expands upon the original guide. https://shinkansan.github.io/2019-UGRP-DPoom/SLAM In the section of the guide headed Localization with rtabmap, it suggests using RTABMAP to study a recorded pointcloud map file and obtain the relative position from that map. |
|
Hi, I also have the same issue: "WARNING [139828389717760] (messenger-libusb.cpp:42) control_transfer returned error, index: 768, error: No data available, number: 61", tried to lower the color_fps and to set initial_reset to true but I still get the message (even without using rviz). Did you find any solution to this? |
|
Hi @Giggus Do you require further assistance with the above problem, please? Thanks! |
|
Hi @MartyG-RealSense, yes I do. |
|
Hi @Giggus What computer / computing device are you using with the camera? Could you also post a roslaunch log in a comment please. Thanks! |
|
I run the L515 on a MSI-GF75-Thin-10UEK (gaming laptop)
Checking log directory for disk usage. This may take a while. started roslaunch server http://gc-GF75-Thin-10UEK:38937/ SUMMARYPARAMETERS
NODES auto-starting new master setting /run_id to b43d3096-0572-11ec-8be9-4dd7bb3c6bd4 [ INFO] [1629875075.855134182]: Device with physical ID 2-4-3 was found. |
|
@Giggus Is this a different MSI gaming PC from the one that you achieved sufficient performance on by reducing FPS to 6 in #1903 (comment) |
|
No, always the same |
|
@Giggus I see that you have already tried reversing the orientation of your L515's USB cable connector at the camera end in #1903 (comment) but it was not a suitable solution because you could only stream low resolution in the reverse orientation. Your ROS wrapper version 2.3.1 is correctly matched with librealsense SDK version 2.48.0, so that SDK and wrapper version pairing looks fine. Can you try the roslaunch instruction below please to launch with only 640x480 at 30 FPS depth enabled in order to see whether the default configuration's inclusion of 1280x720 color is causing any problems. roslaunch realsense2_camera rs_camera.launch depth_width:=640 depth_height:=480 depth_fps:=30 |
|
Same problem
... logging to /root/.ros/log/c50efb0a-0650-11ec-87c1-2cf05db64dd6/roslaunch-GF75-Thin-10UEK-17514.log started roslaunch server http://GF75-Thin-10UEK:40909/ SUMMARYPARAMETERS
NODES auto-starting new master setting /run_id to c50efb0a-0650-11ec-87c1-2cf05db64dd6 [ INFO] [1629970454.441019352]: Device with physical ID 2-4-5 was found. |
|
@Giggus Are you using the official 1 meter USB cable supplied with the camera or a USB cable of your own choice? |
|
@MartyG-RealSense The official one |
|
@Giggus Can the L515 stream depth and RGB normally in the RealSense Viewer program on your computer? |
|
@Giggus It sounds as though the L515 camera hardware is okay, since it is working normally in the RealSense Viewer, and that it is a software problem related to your RealSense ROS setup. Are you using any form of ROS networking (connecting another computer remotely to the gaming laptop that your L515 is attached to) Also, your roslaunch log has the message frame's time domain is HARDWARE_CLOCK. Timestamps may reset periodically. I have seen this message occurring in very few cases, and it suggests to me that Global Time may be disabled in your launch. Do you therefore see a positive difference if you include the instruction global_time_enabled:=true in your roslaunch instruction to enable global time? |
|
@MartyG-RealSense I am connecting to a remote robot but this error happens when I just launch the camera node on itself. I tried to disable " Global Time" but in any launch file the "global_time_enabled:=true" parameter actually exists... |
|
If a launch file does not contain global_time_enabled then it can be added as a rosparam in the launch file for use with L515 using details provided by Doronhi the RealSense ROS wrapper developer in #1607 (comment) |
|
Hi @Giggus Do you require further assistance with this case, please? Thanks! |
|
Hi @MartyG-RealSense, even after adding I still get the same warning |
|
Is the launch still exiting or can you now view the data even though the control_transfer returned error is still occurring? |
|
the launch never exits (unless I stop it myself) but I always have this warning message |
|
If the data is published correctly in RViz and does not seem to be affected by the warning messages then it is likely safe to ignore them. |
|
@MartyG-RealSense ok thanks. I thought it was related with the low performance I have |
|
Hi @Giggus Do you require further assistance with this case, please? Thanks! |
|
No thanks, I'll keep the warning message |
|
Thanks very much @Giggus for the update. As you do not require further assistance and @SarenWraith the creator of the case had their problem resolved earlier in #2028 (comment) I will close the case. Thanks again! |
I got to the same point referred to in: #1510
except that the image is not being displayed in Rviz even when using the add tab
I am doing this on the jetson nano but have not attempted repeating it on a real pc yet. Could that resolve my problem?
The text was updated successfully, but these errors were encountered: