Exit code =137?

Hi all – I’m running rclone on a Centos7 box to sync between a Dell ECS and Wasabi for lukewarm backup. The script we use runs nightly and iterates through an array of bucket names and rclone syncs them one at a time.

For the most part, it’s successful (though Wasabi has its own bugs). To try and get some additional error checking, I decided to capture the exit code and write that to the log every time the rclone sync finishes an iteration, so that I have error_code = # throughout my logs.

Most buckets finish with “0”, but with no other indication that there are errors occuring, sometimes the exit code will be “137”. Might there be something happening that the INFO log level isn’t capturing, and I need to crank that up? Or is this something innocuous that I can ignore?

The actual exit codes are documented here:

http://rclone.org/docs/#exit-code

Example:

notebook2:~$ rclone delete robgd:go1.12.linux-amd64.tar.gz -q 
2019/03/01 14:16:23 ERROR : : error listing: directory not found
2019/03/01 14:16:23 ERROR : Attempt 1/3 failed with 1 errors and: directory not found
2019/03/01 14:16:23 ERROR : : error listing: directory not found
2019/03/01 14:16:23 ERROR : Attempt 2/3 failed with 1 errors and: directory not found
2019/03/01 14:16:23 ERROR : : error listing: directory not found
2019/03/01 14:16:23 ERROR : Attempt 3/3 failed with 1 errors and: directory not found
2019/03/01 14:16:23 Failed to delete: directory not found
notebook2:~$ echo $?
3
notebook2:~$ rclone delete robgd:go1.12.linux-amd64.tar.gz -q -v
2019/03/01 14:16:47 Can't set -v and -q
notebook2:~$ echo $?
1

Are you using an older version?

I just pulled it down a month ago, so I don’t think the codes have changed that recently.

If I just needed the documentation, I wouldn’t be here. It’s the non-standard exit code that’s baffling.

logfile with that error code? and what does rclone version show?

rclone v1.42

  • os/arch: linux/amd64
  • go version: go1.9.4

1.42 is pretty old from June 2018. rclone is supposed to return codes documented in the documentation. You have a few options as you could probably guess. Turn up the logging to see what the issue is and if it is even in there or upgrade to 1.45 or the current beta and see if you still get those return codes. If you do, then open a issue with the log file as something isn’t right.

I’m limited to corporate repos. I can pull down 1.44 at best. I’ll update to that and see if that makes a difference.