pg_staging - Prepare a staging environment from http accessible backups
Configuration file is a .ini containing several sections: first a
GLOBAL section containing common options to the following sections,
then a section per target database.
The [DEFAULT] section allows to provide default values for any setting
describe here, that you can override in any of the specific sections.
Each specific section represent a target database to restore. The
options will allow to say from which backup and on which host. The host
knob is used in commands such as pg_restore, scp and ssh, so can be
either an ip address or a hostname (resolv.conf applies).
Which host to connect to to fetch the backup file. HTTP only.
Path leading to the backup for the section’s database. The database
dump file will get expanded to dbname.date -I.dump, this is not
(yet?) configurable as of version 0.5.
Path to the cluster pg_dumpall -g accompanying file, e.g.
Target where to restore the dump file and manage pgbouncer setup.
Database name, will get appended by the backup date to form a
Owner of the database to get restored.
Maintenance database role used in the connections to create or drop
Maintenance database to connect to, usually postgres.
Database encoding to use in the following statement:
CREATE DATABASE ... WITH OWNER "..." ENCODING ´...´;
PostgreSQL port to use in pgbouncer configuration file.
Major version of postgresql, allows to choose the pg_restore binary
with the default settings.
pgbouncer port to connect to, used for all pg_staging operations.
Configuration file name for pgbouncer, e.g.
boolean (True or False) to control if pg_staging removes the dump
file after restore.
How many databases to keep arround, pg_staging drops less recent
databases. Not Yet Implemented as of 0.5.
boolean to control if pg_staging switches pgbouncer configuration
at restore time
boolean, used when invoking staging-client.sh
boolean, use --single-statement option (-1) in pg_restore.
boolean, when true the restore command will VACUUM ANALYZE the
Number of working process to use at the same time while restoring.
Will set the -j option to pg_restore and will not check for you
that you’re using a pg_restore client version 8.4 or newer.
configuration filename where to setup the replication options.
Directory where to fetch backup files, defaults to /tmp. Command
line argument -t will override .ini setting.
Directory where to dispose SQL files to be sourced with the psql
utility, in alphabetical order. This directory may contain a pre
and a post subdirectories containing files following the shell
*.sql global pattern. No subdirectory of pre and post will get
walked into. The name of the section using the files will always
get expanded to the given sql_path.
List of schema (comma separated) to restore. If present, any schema
not listed here nor in schemas_nodata will get filtered out from
the pg_restore catalog.
List of schemas (comma separated) to add to the database
search_path, using the SQL command +ALTER DATABASE self.dbname SET
search_path TO +.
List of schema (comma separated) to restore without content. The
pg_restore catalog TABLE DATA sections will get filtered out.
backup_host = backup1.backbone
dbuser = postgres
maintdb = postgres
postgres_port = 5432
postgres_major = 8.3
pgbouncer_port = 6432
pgbouncer_conf = /etc/pgbouncer/pgbouncer.ini
remove_dump = True
keep_bases = 2
auto_switch = True
use_sudo = True
pg_restore = /usr/lib/postgresql/%(postgres_major)s/bin/pg_restore
pg_restore_st = True
restore_vacuum = True
restore_jobs = 1
replication = replication.ini
backup_base_url = /PostgreSQL/allopass_db/nightly/
dumpall_url = /PostgreSQL/clusters/bdd.allopass/8.2-main.globals.sql
host = bdd.allopass.dev
postgres_major = 8.2
dbname = allopass_db
dbowner = root
schemas = public, payment, utils, jdb
pg_staging is written by Dimitri Fontaine <firstname.lastname@example.org>.
[FIXME: source] 11/24/2009