Gdk-CRITICAL **: gdk_cursor_new_for_display: assertion 'GDK_IS_DISPLAY (display)' failed; TypeError: constructor returned NULL

[Saneesh]

Hello,
I have successfully configured DIGITS in AWS with the supported system like Ubuntu 14.04, CUDA 7.0, GPU etc.

When I trying create the dataset through the browser, the following error is showing:

(parse_folder.py:13268): Gdk-CRITICAL **: gdk_cursor_new_for_display: assertion 'GDK_IS_DISPLAY (display)' failed
Default value for log_file "/home/ubuntu/digits/digits/digits.log" invalid:
You do not have write permissions
Traceback (most recent call last):
File "/home/ubuntu/digits/tools/parse_folder.py", line 21, in <module>
digits.config.load_config()
File "/home/ubuntu/digits/digits/config/load.py", line 147, in load_config
option.apply()
File "/home/ubuntu/digits/digits/config/caffe_option.py", line 232, in apply
import caffe
File "/home/ubuntu/caffe/python/caffe/__init__.py", line 1, in <module>
from .pycaffe import Net, SGDSolver
File "/home/ubuntu/caffe/python/caffe/pycaffe.py", line 14, in <module>
import caffe.io
File "/home/ubuntu/caffe/python/caffe/io.py", line 2, in <module>
import skimage.io
File "/usr/local/lib/python2.7/dist-packages/skimage/io/__init__.py", line 15, in <module>
reset_plugins()
File "/usr/local/lib/python2.7/dist-packages/skimage/io/manage_plugins.py", line 89, in reset_plugins
_load_preferred_plugins()
File "/usr/local/lib/python2.7/dist-packages/skimage/io/manage_plugins.py", line 69, in _load_preferred_plugins
_set_plugin(p_type, preferred_plugins['all'])
File "/usr/local/lib/python2.7/dist-packages/skimage/io/manage_plugins.py", line 81, in _set_plugin
use_plugin(plugin, kind=plugin_type)
File "/usr/local/lib/python2.7/dist-packages/skimage/io/manage_plugins.py", line 251, in use_plugin
_load(name)
File "/usr/local/lib/python2.7/dist-packages/skimage/io/manage_plugins.py", line 295, in _load
fromlist=[modname])
File "/usr/local/lib/python2.7/dist-packages/skimage/io/_plugins/matplotlib_plugin.py", line 4, in <module>
import matplotlib.pyplot as plt
File "/usr/local/lib/python2.7/dist-packages/matplotlib/pyplot.py", line 109, in <module>
_backend_mod, new_figure_manager, draw_if_interactive, _show = pylab_setup()
File "/usr/local/lib/python2.7/dist-packages/matplotlib/backends/__init__.py", line 32, in pylab_setup
globals(),locals(),[backend_name],0)
File "/usr/local/lib/python2.7/dist-packages/matplotlib/backends/backend_gtk3agg.py", line 11, in <module>
from . import backend_gtk3
File "/usr/local/lib/python2.7/dist-packages/matplotlib/backends/backend_gtk3.py", line 55, in <module>
cursors.MOVE          : Gdk.Cursor.new(Gdk.CursorType.FLEUR),
TypeError: constructor returned NULL

How can I fix this issue?

[lukeyeager]

@joyofdata addressed this in his blog post:
http://www.joyofdata.de/blog/gpu-powered-deeplearning-with-nvidia-digits/

I have personally never run into this issue on EC2.

[Saneesh]

Hello Luke,
I'm facing this issue when I create the dataset using DIGITS in the browser. So the solution he mentioned is when I use ssh command I have use like "ssh -Xi ..." but in my case I'm not accessing using ssh. I'm accessing through the browser to create the dataset.

[gheinrich]

Hi Saneesh, are you not using SSH to start the DIGITS server? You may want to have a look at @jashmenn's solution on BVLC/caffe#861 (comment). Can you try adding these two lines at the top of your /home/ubuntu/digits/tools/parse_folder.py:

import matplotlib
matplotlib.use('Agg')

[Saneesh]

Now I could login through ssh and my issue is solved!
These are the steps I have followed
username@local:# sudo -s
root@local:# ssh -X -v -i /home/username/Downloads/keypair.pem ubuntu@ec2-xx-xx-xx-xx.compute-x.amazonaws.com
ubuntu@amazon:# sudo -s
root@amazon:# nohup ./digits-devserver &

Now, access the site and create the dataset!

[gheinrich]

Thanks Saneesh, out of curiosity, would you mind trying this:

• log in with SSH without using the -X option
• add the two lines from Gdk-CRITICAL **: gdk_cursor_new_for_display: assertion 'GDK_IS_DISPLAY (display)' failed; TypeError: constructor returned NULL #187 (comment)
• create a new dataset

I am interested to know whether this works. Thanks!

[Saneesh]

Hello gheinrich,
I tried that first but it didn't worked for me, it showed same error.

[nszceta]

@gheinrich I did not have luck with modifying parse_folder.py. However, commit 28b7fee should fix the gtk3 backend issue completely.

[nszceta]

@Saneesh please try again with commit 28b7fee applied

[nszceta]

@lukeyeager I ran into this issue on a fresh Ubuntu 14.04 LTS HVM install on AWS EC2 g2.2xlarge instance

[choosehappy]

I also had an issue with this; to resolve i simply used a virtual framebuffer, and everything worked without a problem.

On ubuntu:
sudo apt-get install xvfb
xvfb-run ./digits-server

[christopher5106]

sudo pip uninstall matplotlib
sudo apt-get install python-matplotlib

works for me

[zhengzhang828]

christopher5106 method works! thank you!

[mikechen66]

Solution to 'GDK_IS_DISPLAY (display)' failed

It seems that the error is a minor error during the ssh connection from a computer to another computer. While I run the test application of class_activation_maps.py (keras examples), it shows the same error.

$ python3 class_activation_maps.py /home/nvidia/Downloads/galaxy.jpg
Unable to init server: Could not connect: Connection refused
Unable to init server: Could not connect: Connection refused

(image_ocr.py:1257): Gdk-CRITICAL **: 19:44:35.690: gdk_cursor_new_for_display: assertion 'GDK_IS_DISPLAY (display)' failed

Using TensorFlow backend.
.........

However, while I run the following both the commands and the application. It can successfully show the splendid galaxy makeup.

$ echo $DISPLAY
:0
$ python3 class_activation_maps.py /home/nvidia/Downloads/galaxy.jpg

Using TensorFlow backend.
2020-05-01 19:45:14.877288: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.0

Notes:

Another solution is to describe the above-metioned commands in bashrc.

Cheers,

Mike