I am trying to dump a database which has a size of just over 100 GiB. I am using the following command (with
pv to give a progress bar to observe where this problem occurs):
pg_dump -h <host> -U <user> <db> | pv -c -s $(psql -h <host> -U <user> <db> -tc "SELECT pg_database_size('<db>')") -N dump | gzip > <db>.sql.gz
The dump chugs along happily until it reaches 56.8GiB raw passing through the pipeline (~25GiB of the compressed file), and then just dies. I am at a loss as to why this is happening. The filesystem is
ext4, so it's not a size limitation there.
I'm currently running it again, but splitting files into 10GiB chunks. Let's see how this goes.
Any ideas why, no matter what, the dump stops at 56.8GiB?
EDIT: The server I'm trying to get the dump from is a pod running on Google Kubernetes Engine. I have tried dumping both on the pod itself, and from another pod connecting through a service, and the problem persists.