When trying to run a function that uses a container:
db01=# select * from pylog100();
DEBUG: Message type Q received by from libpq, len = 34
LOG: statement: select * from pylog100();
DEBUG: GPORCA produced plan
DEBUG: plcontainer: Entering call handler with PLy_curr_procedure
DEBUG: plcontainer: Calling python proc @ address: 0x2cac068
DEBUG: plcontainer: fcinfo->flinfo->fn_retset: 0
DEBUG: plcontainer: docker created with id 564b39e38d392a5595f2f4a1c6fd43bc52c177a326a8f77fd21bdd223b032986.
DEBUG: plcontainer: container 564b39e38d392a5595f2f4a1c6fd43bc52c177a326a8f77fd21bdd223b032986 has started at Tue Jan 13 08:32:02 2026
DEBUG: plcontainer: PLContainer: Failed to connect to /tmp/plcontainer.2728091.0.0/unix.domain.socket.shared.file: No such file or directory
DEBUG: plcontainer: failed to connect to client. Maybe expected. dockerid: 564b39e38d392a5595f2f4a1c6fd43bc52c177a326a8f77fd21bdd223b032986
DEBUG: plcontainer: Waiting for 25 ms for before reconnecting
DEBUG: plcontainer: PLContainer: Failed to connect to /tmp/plcontainer.2728091.0.0/unix.domain.socket.shared.file: No such file or directory
DEBUG: plcontainer: failed to connect to client. Maybe expected. dockerid: 564b39e38d392a5595f2f4a1c6fd43bc52c177a326a8f77fd21bdd223b032986
plcontainer runtime configuration:
<?xml version="1.0" ?>
<configuration>
<backend name="podman" type="podman">
<exec_path>/usr/bin/podman</exec_path>
</backend>
<runtime>
<id>plc_python3_shared</id>
<image>docker.io/library/python311.ubi9_b:latest</image>
<command>/clientdir/py3client.sh</command>
<shared_directory host="/usr/local/greenplum-db/bin/plcontainer_clients" container="/clientdir" access="ro"/>
<setting use_container_logging="yes"/>
<setting enable_network="no"/>
<backend name="podman" />
</runtime>
</configuration>
This may be caused by some configuration issue.
There is a code bug in the plcontainer extension version 4.3.1 and older.
To use podman as container manager:
Upgrade Greenplum to 7.7.1 or above.
Ensure plcontainer is version 4.3.2 or above