Skip to content

SinaraML Known issues

Jump to bottom
pestovid edited this page Jun 7, 2024 · 8 revisions

Apache Spark UI can't be opened in first instance of the running SinaraML Server

You can get following warnings while running pipeline steps which is not affected pipeline functionality:

  • Spark unable to delete temporary files
ERROR: ShutdownHookManager: Exception while deleting temp dir ...
  • Pandas version
UserWarning: Pandas requires version '2.8.0' or newer of 'numexpr'
  • Python version for the BentoML

Error message "Port is already allocated"

If you encounter the following error while starting a sinara server:

ERROR: Docker daemon failed, Run CLI with sinara --verbose flag to see details
ERROR: 500 Server Error for http+docker://localhost/v1.44/containers/2211908376b1fcdb547b32ea4edef3d57fb4b30dbc0eab6bb98a559603d0ffd6/start: 
Internal Server Error ("driver failed programming external connectivity on endpoint test1 (3bb75ad51c3b132ad2b008ef5546b9c778145cc2818c923ee4ab9200d26d828b): 
Bind for 0.0.0.0:8888 failed: port is already allocated")

This means, that you are trying to start two sinara servers on the same port. It could happen if you created the first server and immediately created the second server instance without starting the first one.

To avoid this error, do the following:

  • create the first server
  • start the first server
  • create the second server
  • start the second server

CLI will correctly manage port allocation in this case and both servers will be able to start.

Error message "Your OS is not supported"

SinaraML CLI can only be run on certain OS platforms, such as:

  • WSL
  • MacOS

Running CLI via PowerShell or cmd in Windows is not supported.

Storing files with name containing ':' symbol in name not supported.

Because of issue in Hadoop library https://issues.apache.org/jira/browse/HADOOP-3257 you can get error when using pack_files_from_tmp_to_store:

Caused by: java.net.URISyntaxException: Relative path in absolute URI: .my_file_2022-01-19_13:26:05.521206.jpg.crc

You can use following code in your notebook to replace ':' symbol

import glob
for file in [x for x in glob.glob(tmp_entity_dir + "/**", recursive=True) if ':' in x]:
    os.rename(file, file.replace(':', '_'))

Storing files with name starting with '.' (dot) or '_' (underscore) symbol in name not supported.

These files are considered hidden and not stored using pack_files_from_tmp_to_store method. Examples:

.this_file_is_hidden
_this_file_is_hidden_too
Clone this wiki locally