Cannot view tiles in remote Jupyter environment

[MATRIX4284]

Raster layer map not displayed for TileClient with local files.

The raster layer comng from local file is not being displayed on map

[giswqs]

You need to provide information about OS and env. Providing the sample dataset would be helpful.

import localtileserver
print(localtileserver.Report())

[banesullivan]

Screenshots and code would also be helpful

[MATRIX4284]

import localtileserver
print(localtileserver.Report())

--------------------------------------------------------------------------------
  Date: Fri Feb 04 03:44:31 2022 IST

                     OS : Linux
                 CPU(s) : 32
                Machine : x86_64
           Architecture : 64bit
                    RAM : 125.9 GiB
            Environment : Jupyter
            File system : ext4

  Python 3.7.11 (default, Jul 27 2021, 14:32:16)  [GCC 7.5.0]

        localtileserver : 0.4.0
                  flask : 2.0.2
          flask_caching : 1.10.1
            flask_restx : 0.5.1
               requests : 2.27.1
               werkzeug : 2.0.2
                  click : 7.1.2
                 scooby : 0.5.11
            large_image : 1.10.0
large_image_source_gdal : 1.10.0
             cachetools : 4.2.4
                    PIL : 9.0.0
                 psutil : 5.8.0
                  numpy : 1.21.2
             palettable : 3.3.0
                 pyproj : 3.1.0
             osgeo.gdal : 3.3.1
             ipyleaflet : 0.14.0
              traitlets : 5.1.1
                shapely : 1.7.1
                 folium : 0.12.1.post1
             matplotlib : 3.5.0
--------------------------------------------------------------------------------

[MATRIX4284]

from localtileserver import get_leaflet_tile_layer, TileClient
from ipyleaflet import Map

# First, create a tile server from local raster file
tile_client = TileClient('/DISK003/SPATIAL_ANALYTICS/landuse.tif')

# Create ipyleaflet tile layer from that server
t = get_leaflet_tile_layer(tile_client)

# Create ipyleaflet map, add tile layer, and display
m = Map(center=tile_client.center())
m.add_layer(t)
m

[MATRIX4284]

the tile layer id not getting displayed in map

[banesullivan]

Would it be possible to share the landuse.tif file?

the tile layer id not getting displayed in map

I do not understand what you mean... could you include screenshots?

[MATRIX4284]

landuse.tiff.zip
it is clearly getting plotted in QGIS

[MATRIX4284]

attached the zip file of the tif as it was not allowing to attach tiff directly

[MATRIX4284]

you can try to visualize the tif in QGIS as a reference

[banesullivan]

I am having no issues visualizing this raster with localtileserver. From what I can tell, there isn't anything wrong with the file and localtileserver handles it as well as any of the examples.

Here is the code I used:

from localtileserver import get_leaflet_tile_layer, TileClient
from ipyleaflet import Map

# First, create a tile server from local raster file
tile_client = TileClient('landuse.tiff')

# Create ipyleaflet tile layer from that server
t = get_leaflet_tile_layer(tile_client)

# Create ipyleaflet map, add tile layer, and display
m = Map(center=tile_client.center())
m.add_layer(t)
m

and a screenshot of Jupyter:

further, the command-line utility has no issue showing this in CesiumJS:

localtileserver landuse.tiff

---

@MATRIX4284, in order to help you further, I need you to specify exactly what is going wrong:

the tile layer id not getting displayed in map

I'm not sure what you mean by this.

Would you please included screenshots of what is happening and describe what you expect to happen?

[MATRIX4284]

how come it is possible i am ruuning the same set of commands and i can see only the basemap but not any other layer

[banesullivan]

@MATRIX4284, I'm not sure. Are you seeing any errors? In the JavaScript console of your browser are you seeing any errors? in the network tab of the dev tools in your browser, are you seeing any 404s?

Is it possible you are running on a remote system where your web browser might not have access to the port on which localtileserver is serving the imagery?

[MATRIX4284]

As you can see i can see basemap layer but not the layer in showing the land usage.I am not abe to comprehend the reason why it is not showing up in my machine despite similar code snippets used.

[MATRIX4284]

is there any dependency on chrome version ??

[MATRIX4284]

[MATRIX4284]

os:ubunyu 16.04
python:

@MATRIX4284, I'm not sure. Are you seeing any errors? In the JavaScript console of your browser are you seeing any errors? in the network tab of the dev tools in your browser, are you seeing any 404s?

Is it possible you are running on a remote system where your web browser might not have access to the port on which localtileserver is serving the imagery?

yes i am using remote system where my jupyter is running at port 15000 .I am connecting to jupyer by ip:port

[MATRIX4284]

i just tested the docker with the volume the output is showing with
docker run -p 15003:8000 -v /DISK003/SPATIAL_ANALYTICS/:/data/ ghcr.io/banesullivan/localtileserver:latest

[MATRIX4284]

but python module when installed on os is not showing up the stuff neither the tile layer is showing up in jupyter

[banesullivan]

Aha. This is a duplicate of #29 which isn't fully solved. banesullivan/localtileserver-demo#1 also tracks this

Because you are running in a remote environment, you must set up your server to expose the port the on which localtileserver is serving the image tiles. localtileserver randomly chooses a port by default, so you will need to allocate a port and pass it to localtileserver. For example, use 8000:

tile_client = TileClient('landuse.tiff', port=8000)

I don't have a good solution to this use case (yet).

[banesullivan]

Release 0.4.1.dev0 provides an interface that allows for a solution to this and #29 and banesullivan/localtileserver-demo#1 using jupyter-server-proxy

This can get messy/complicated and I haven't documented (at all), but essentially localtileserver.TileClient now allows you to map the server's host and port to what your browser sees as the host (and port). To use this, specify port, client_port client_host, and client_prefix on init of TileClient. Or better yet, set the following environment variables where you are running Jupyter:

• LOCALTILESERVER_PORT=5555
• LOCALTILESERVER_CLIENT_PORT=8888
• LOCALTILESERVER_CLIENT_HOST=127.0.0.0
• LOCALTILESERVER_CLIENT_PREFIX=proxy/$LOCALTILESERVER_PORT

This effectively remaps the tile requests coming from ipyleaflet from localhost:5555 to 127.0.0.0:8888/proxy/5555 allowing your browser access

You can also use/see jupyter.Dockerfile:

• https://github.com/banesullivan/localtileserver/blob/main/jupyter.Dockerfile
• https://github.com/banesullivan/localtileserver/pkgs/container/localtileserver-jupyter

cc @giswqs

I will follow up in banesullivan/localtileserver-demo#1 with a demo and soon release 0.4.1

[giswqs]

@banesullivan You rock! This is exciting! Can't wait to try it out.

[giswqs]

It seems the binder system username is jovyan. Can you set those env variables automatically when the system username is jovyan?

[banesullivan]

Yep! You should be able to set env variable no matter the user

[banesullivan]

I just pushed 0.4.1 which removed LOCALTILESERVER_PORT and now requires you to set the following to work with jupyter-server-proxy:

• LOCALTILESERVER_CLIENT_HOST=127.0.0.1 - The host on which you launch Jupyter
• LOCALTILESERVER_CLIENT_PORT=8888 - The port on which you launch Jupyter
• LOCALTILESERVER_CLIENT_PREFIX='proxy/{port}' - Same for everyone using jupyter-server-proxy

[MATRIX4284]

I will test after setting the variables accordingly

[banesullivan]

This was further improved in #67 such that only setting:

LOCALTILESERVER_CLIENT_PREFIX='proxy/{port}' should be sufficient for you, @MATRIX4284

I will release 0.4.2 shortly

[banesullivan]

@MATRIX4284, I'm going to assume your issue has been resolved and close this issue. Please reach out if you are still having issues

[MATRIX4284]

@MATRIX4284, I'm going to assume your issue has been resolved and close this issue. Please reach out if you are still having issues

Hi @banesullivan @giswqs thanks for you support .Now i can see my tiles in jupyter .Thanks a ton