Open fridim opened 1 month ago
If you are seeing multiple errors like that, aws-nuke isn't exiting because of those, it's just indicating it can't do the query against the region, but it does sounds like you said the pod got killed due to OOM?
Any idea as to the number of actual resources found in the account?
@fridim exclude all the resource types listed in your stderr and try again. They shouldn't have any impact. Those errors trigger when the lister for the resource is unable to list, and it then exits the specific lister, the rest of the tool continues to operator. Also make sure S3Object
is excluded.
I can exclude them but what if i want aws-nuke to remove them if they exist (and work in other regions)?
For troubleshooting purposes. I'll look into handling the other errors but they aren't interfering with the tool running.
On one of our accounts, aws-nuke fails with
dial tcp
andno such host
errors in stderr.command:
aws-nuke nuke --profile ... -c "nuke-config.yml" --quiet --log-level error --no-dry-run --force
duration: ~4min RC: -9We notice the memory usage goes up to 2Go (limit of the pod) before being killed. Could be the indication of a memory leak as none of the other account reach that amount of memory.