Timeout seems changed in stopping mounts from 10 secs to 1 sec.

[tomgreen66]

Version of Apptainer

What version of Apptainer (or Singularity) are you using? Run

apptainer --version (or singularity --version).

apptainer version 1.3.0

Expected behavior

When exiting from a container I expect to not see an error message.

Actual behavior

What actually happened? Why was it incorrect?

exit
DEBUG   [U=1170613,P=145401]CleanupContainer()            Cleanup container
DEBUG   [U=1170613,P=145401]umount()                      Umount /var/apptainer/mnt/session/final
DEBUG   [U=1170613,P=145401]umount()                      Umount /var/apptainer/mnt/session/rootfs
DEBUG   [U=1170613,P=145401]stop()                        Waiting for squashfuse_ll pid 145432 to exit
DEBUG   [U=1170613,P=145401]stop()                        Terminating pid 145432 after wait timeout
DEBUG   [U=1170613,P=145401]waitInstance()                squashfuse_ll pid 145432 has exited with status 241
DEBUG   [U=1170613,P=145401]filterMsg()                   fuse: failed to clone device fd: Inappropriate ioctl for device
DEBUG   [U=1170613,P=145401]filterMsg()                   fuse: trying to continue without -o clone_fd.
INFO    [U=1170613,P=145401]CleanupContainer()            Cleanup error: while stopping driver for /var/apptainer/mnt/session/rootfs: squashfuse_ll exited
DEBUG   [U=1170613,P=145401]Master()                      Child exited with exit status 137

Looks like in:

apptainer/internal/pkg/image/driver/imagedriver.go

Line 507 in 3608761

waitTimeout := 1 * time.Second

The timeout got changed from 10 seconds to 1 second looking at the previous code. Maybe misreading the change though so might be just taking a little longer to unmount.

Steps to reproduce this behavior

How can others reproduce this issue/problem?

In my case I would run a miniconda container, load an environment, run python and load a module and then exit Python and the container.

What OS/distro are you running

$ cat /etc/os-release
NAME="CentOS Linux"
VERSION="7 (Core)"
ID="centos"
ID_LIKE="rhel fedora"
VERSION_ID="7"
PRETTY_NAME="CentOS Linux 7 (Core)"
ANSI_COLOR="0;31"
CPE_NAME="cpe:/o:centos:centos:7"
HOME_URL="https://www.centos.org/"
BUG_REPORT_URL="https://bugs.centos.org/"

CENTOS_MANTISBT_PROJECT="CentOS-7"
CENTOS_MANTISBT_PROJECT_VERSION="7"
REDHAT_SUPPORT_PRODUCT="centos"
REDHAT_SUPPORT_PRODUCT_VERSION="7"

How did you install Apptainer

Write here how you installed Apptainer (or Singularity). Eg. RPM, source.

From source.

[DrDaveD]

Please try it with 1.3.1. This could be a duplicate of #2104. There's still a message when it times out now, but it's more helpful and doesn't indicate that it was an error.

If you don't actually have any background processes holding a reference to the container, I'd like to know more details of exactly how to reproduce the problem.

[tomgreen66]

Thanks for that. Having run ps aux after exiting Python it seems the python module (in this case the ArcGIS python module arcpy) can leave a process running Xvfb. So I assume it is entirely the same as #2104 - apologies for duplicate. Will try 1.3.1.