I wonder how this would compare to just punching a hole in the already applied section of the WAL (1) file. It seems that this should allow the filesystem to fr

I wonder how this would compare to just punching a hole in the already applied section of the WAL (1) file. It seems that this should allow the filesystem to free the disk space similarly to if the file had been deleted while allowing the writers to keep writing. I guess eventually it would become a problem as you hit max logical file size but it seems that you can likely bet on getting a chance to truncate it before hitting exabytes.