rishabh6788
commented
7 months ago As per subprocess documentation to ignore UnicodeDecodeError exception, we need to add errors=ignore in the subprocess.call command. To verify this I spun up a windows host and ran the windows container.
I ran for opensearch-observability without error=ignore flag, and I got the same UnicodeDecodeError exception , then i add the flag in the command in execute method and it ran successfully. To confirm it works, in the same container session I removed the flag again and got the decode error again and after adding it back it is gone.
I can see the files in the test-results directory.
We can better handle it by catching this exception.
try:
result = subprocess.run(command, cwd=dir, shell=True, capture_output=capture, text=True, encoding='utf-8')
except UnicodeDecodeError as e:
print(f"Error decoding stdout: {e}")
result = subprocess.run(command, cwd=dir, shell=True, capture_output=capture, text=True, encoding='utf-8', errors='ignore')
peterzhuamazon
Hi, We are currently seeing this issue in Windows Zip, for OS only.
UnicodeDecodeError: ‘utf-8’ codec can’t decode byte 0xbb in position 523169: invalid start byte2024-03-23 01:13:49 INFO Executing "bash C:\Users\ContainerAdministrator\opensearch-build\scripts\default\integtest.sh -b localhost -p 9200 -s true -v 2.13.0" in C:\Users\ContainerAdministrator\tmpwbgpq28i\<plugin name>until timeoutmainto2.12.0opensearch-build code does not fix the issuesubprocess.rundoes not receive a properexit/returnfrombash integtest.sh, thus keep waiting for the end of the subroutine until time outThanks.