Libreville, Gabon - Sept 8th, 2021
pg_dumpbinary is a program used to dump a PostgreSQL database with
data dumped in binary format. The resulting dump must be restored
pg_restorebinary that is provided with this tool.
pg_dumpbinary 2.5 was released today, this new release adds support to data restoration in tables with altered structure.
pg_dumpbinary will now create a file
table that stores the tables columns list at export time. At restore
time pg_restore binary will read this file if present, or get the target
columns list from the pre-data section dumped. The target list will be
appended to the
COPY FROM command to avoid error messages like
"ERROR: row field count is 28, expected 29". This is necessary for tables
from upgraded extensions with more columns or if you try to import data
only into an existent table with more columns.
pg_dumpbinary is useful in some situations:
pg_dump because the
total size of the escape/hex output exceed 1Gb.
\0 internally in bytea but data
are returned as char/varchar/text which truncate data after
the '\0'. In this case pg_dump will export data in the output
type which will result in data lost.
If you are in this case
pg_dumpbinary will help you by dumping the
PostgreSQL database in binary format. In all other cases you must
use the pg_dump/pg_restore commands distributed with PostgreSQL.
See documentation for a complete description of the features.
pg_dumpbinary is an open project from LzLabs GmbH. Any contribution to build a better tool is welcome. You just have to send your ideas, features requests or patches using the GitHub tools.