It doesn’t really work like that, as PostgreSQL needs to be tuned according to your specific setup. I’ve written about this in past topics.
So after looking into this problem again, I suddenly realized something which I had completely forgotten. We’ve also run into this exact problem quite some time ago. It’s not PostgreSQL being slow, it’s gzip compressing the backup files afterwards. We’ve replaced it with pigz, which is multi-threaded replacement for gzip. It is a slight change to the backup script, but a huge improvement.
I thought we had internally discussed filing a PR to ask to incorporate our changes into the zammad_backup.sh script, but I can’t find anything at the moment. I guess more pressing matters came in between at the time and we kind of forgot about it eventually. I’ll ask my coworker who came up with the idea to file a PR.