Clone attempt failed, `extractOne: Interpret failed: Interpret: copy failed` about postgres-operator HOT 4 OPEN

We usually see these "extractOne: Interpret failed: Interpret: copy failed" messages when there's no disk space left and the specified volume is too small for the database cluster that is about to get cloned. To me the logs looks like the operator is working as expected - which is only setting the right environment variables for Spilo so wal-e will be called to pull a backup and apply WAL files.

Wal-e crashed ("wal-e backup-fetch exited with exit code -9"), hence clone is retried. Maybe specify more memory?

from postgres-operator.

I had provisioned the same size PV, and had given up with cloning by s3 and just manually restored by base backup.

from postgres-operator.

Had to redo test cluster for new tests. This time I attempted to restore to an exact replica, same pod specs and everything, failed twice.

acid-example-test6-0_postgres.log
acid-example-test7-0_postgres.log
postgres-operator-7469cf8c5b-7pbcf_postgres-operator.log

from postgres-operator.

Cloning from existing master works, but is greatly limited by the fact that it has to be the exact same namespace. We are only using it right now cause we are in a pinch and really want to get some tests done.

from postgres-operator.