PGBOUNCER(5)

NAME

pgbouncer - Lightweight connection pooler for PostgreSQL.

SYNOPSIS

[databases]
db = ...

[pgbouncer]
...

DESCRIPTION

Config file is in "ini" format. Section names are between " and ".
Lines starting with ";" or "" are taken as comments and ignored. The
characters ";" and "" are not recognized when they appear later in the line.

SECTION [PGBOUNCER]

Generic settings: logfile
Specifies log file. Log file is kept open so after rotation kill
-HUP or on console RELOAD; should be done. Note: On Windows
machines, the service must be stopped and started.

Default: not set.; pidfile
Specifies the pid file. Without a pidfile, daemonization is not
allowed.

Default: not set.; listen_addr
Specifies IPv4 address, where to listen for TCP connections. You
may also use * meaning "listen on all addresses". When not set,
only Unix socket connections are allowed.

Default: not set; listen_port
Which port to listen on. Applies to both TCP and Unix sockets.

Default: 6432; unix_socket_dir
Specifies location for Unix sockets. Applies to both listening
socket and server connections. If set to an empty string, Unix
sockets are disabled. Note: Not supported on Windows machines.

Default: /tmp; user
If set, specifies the Unix user to change to after startup. Works
only if PgBouncer is started as root or if user is the same as the current user. Note: Not supported on Windows machines.

Default: not set; auth_file
The name of the file to load user names and passwords from. The
file format is the same as the PostgreSQL pg_auth/pg_pwd file, so
this setting can be pointed directly to one of those backend files.

Default: not set.; auth_type
How to authenticate users.

md5: Use MD5-based password check. auth_file may contain both
MD5-encrypted or plain-text passwords. This is the default
authentication method.

crypt
Use crypt(3) based password check. auth_file must contain
plain-text passwords.

plain

Clear-text password is sent over wire.

trust

No authentication is done. Username must still exist in
auth_file.

any

Like the trust method, but the username given is ignored.
Requires that all databases are configured to log in as
specific user. Additionally, the console database allows any
user to log in as admin.
pool_mode: Specifies when a server connection can be reused by other clients.; session
Server is released back to pool after client disconnects.
Default.; transaction
Server is released back to pool after transaction finishes.; statement
Server is released back to pool after query finishes. Long
transactions spanning multiple statements are disallowed in
this mode.
max_client_conn: Maximum number of client connections allowed. When increased then
the file descriptor limits should also be increased. Note that
actual number of file descriptors used is more than
max_client_conn. Theoretical maximum used is:

max_client_conn + (max_pool_size * total_databases * total_users); if each user connects under its own username to server. If a
database user is specified in connect string (all users connect
under same username), the theoretical maximum is:

max_client_conn + (max_pool_size * total_databases); The theoretical maximum should be never reached, unless somebody
deliberately crafts special load for it. Still, it means you should set the number of file descriptors to a safely high number.; Search for ulimit in your favourite shell man page. Note: ulimit
does not apply in a Windows environment.; Default: 100
default_pool_size: How many server connections to allow per user/database pair. Can be overriden in the per-database configuration.; Default: 20
reserve_pool_size: How many additional connections to allow to a pool. 0 disables.; Default: 0 (disabled)
reserve_pool_timeout: If a client has not been serviced in this many seconds, pgbouncer
enables use of additional connections from reserve pool. 0
disables.; Default: 5
server_round_robin: By default, pgbouncer reuses server connections in LIFO (last-in,
first-out) manner, so that few connections get the most load. This gives best performance if you have a single server serving a
database. But if there is TCP round-robin behind a database IP,
then it is better if pgbouncer also uses connections in that
manner, thus achieving uniform load.; Default: 0
ignore_startup_parameters: By default, PgBouncer allows only parameters it can keep track of
in startup packets - client_encoding, datestyle, timezone and
standard_conforming_strings.; All others parameters will raise an error. To allow others
parameters, they can be specified here, so that pgbouncer knows
that they are handled by admin and it can ignore them.; Default: empty
Log settings: syslog
Toggles syslog on/off As for windows environment, eventlog is used instead.

Default: 0; syslog_facility
Under what facility to send logs to syslog. Possibilities: auth,
authpriv, daemon, user, local0-7

Default: daemon; log_connections
Log successful logins.

Default: 1; log_disconnections
Log disconnections with reasons.

Default: 1; log_pooler_errors
Log error messages pooler sends to clients.

Default: 1
Console access control: admin_users
Comma-separated list of database users that are allowed to connect and run all commands on console. Ignored when auth_mode=any, in
which case any username is allowed in as admin.

Default: empty; stats_users
Comma-separated list of database users that are allowed to connect and run read-only queries on console. Thats means all SHOW commands except SHOW FDS.

Default: empty.
Connection sanity checks, timeouts: server_reset_query
Query sent to server on connection release, before making it
available to other clients. At that moment no transaction is in
progress so it should not include ABORT or ROLLBACK.

Good choice for Postgres 8.2 and below is:

server_reset_query = RESET ALL; SET SESSION AUTHORIZATION DEFAULT;

for 8.3 and above its enough to do:

server_reset_query = DISCARD ALL;
server_check_delay: How long to keep released connections available for immidiate
re-use, without running sanity-check queries on it. If 0 then the
query is ran always.; Default: 30
server_check_query: Simple do-nothing query to check if the server connection is alive.; If an empty string, then sanity checking is disabled.; Default: SELECT 1;
server_lifetime: The pooler will try to close server connections that have been
connected longer than this. Setting it to 0 means the connection is to be used only once, then closed. [seconds]; Default: 3600
server_idle_timeout: If a server connection has been idle more than this many seconds,
and there are too many connections in the pool, this one can be
dropped. [seconds]; Default: 600
server_connect_timeout: If connection and login won't finish in this amount of time, the
connection will be closed. [seconds]; Default: 15
server_login_retry: If login failed, because of failure from connect() or
authentication that pooler waits this much before retrying to
connect. [seconds]; Default: 15
client_login_timeout: If a client connects but does not manage to login in this amount of time, it will be disconnected. Mainly needed to avoid dead
connections stalling SUSPEND and thus online restart. [seconds]; Default: 60
autodb_idle_timeout: If the automatically created (via "*") database pools have been
unused this many seconds, they are freed. The negative aspect of
that is that their statistics are also forgotten. [seconds]; Default: 3600
Dangerous timeouts: Setting following timeouts cause unexpected errors.; query_timeout
Queries running longer than that are canceled. This should be used only with slightly smaller server-side statement_timeout, to apply only for network problems. [seconds]

Default: 0 (disabled); query_wait_timeout
Maximum time queries are allowed to spend waiting for execution. If the query is not assigned to a server during that time, the client is disconnected. This is used to prevent unresponsive servers from grabbing up connections. [seconds]

Default: 0 (disabled); client_idle_timeout
Client connections idling longer than this many seconds are closed. This should be larger than the client-side connection lifetime
settings, and only used for network problems. [seconds]

Default: 0 (disabled)
Low-level network settings: pkt_buf
Internal buffer size for packets. Affects size of TCP packets sent and general memory usage. Actual libpq packets can be larger than
this so, no need to set it large.

Default: 2048; listen_backlog
Backlog argument for listen(2). Determines how many new unanswered connection attempts are kept in queue. When queue is full, futher
new connections are dropped.

Default: 128; sbuf_loopcnt
How many times to process data on one connection, before
proceeding. Without this limit, one connection with a big resultset can stall PgBouncer for a long time. One loop processes one pkt_buf amount of data. 0 means no limit.

Default: 5; tcp_defer_accept
For details on this and other tcp options, please see man 7 tcp.

Default: 45 on Linux, otherwise 0; tcp_socket_buffer
Default: not set; tcp_keepalive
Default: not set; tcp_keepcnt
Default: not set; tcp_keepidle
Default: not set; tcp_keepintvl
Default: not set

SECTION [DATABASES]

This contains key=value pairs where key will be taken as a database
name and value as a libpq connect-string style list of key=value pairs. As actual libpq is not used, so not all features from libpq can be used (service=, .pgpass).

Database name can contain characters [0-9A-Za-z_.-] without quoting.
Names that contain other chars need to be quoted with standard SQL
ident quoting: double quotes where "" is taken as single quote.

"\*" acts as fallback database: if the exact name does not exist, its
value is taken as connect string for requested database. Such
automatically created database entries are cleaned up if they stay idle longer then the time specified in autodb_idle_timeout parameter.
Location parameters: dbname
Destination database name.

Default: same as client-side database name.; host
IP address to connect to.

Default: not set, meaning to use a Unix socket.; port
Default: 5432; user, password
If user= is set, all connections to the destination database will
be done with the specified user, meaning that there will be only
one pool for this database.

Otherwise PgBouncer tries to log into the destination database with client username, meaning that there will be one pool per user.
Pool configuration: pool_size
Set maximum size of pools for this database. If not set, the
default_pool_size is used.; connect_query
Query to be executed after a connection is established, but before allowingf the connection to be used by any clients. If the query
raises errors, they are logged but ignored otherwise.
Extra parameters: They allow setting default parameters on server connection.; Note that since version 1.1 PgBouncer tracks client changes for their
values, so their use in pgbouncer.ini is deprecated now.; client_encoding
Ask specific client_encoding from server.; datestyle
Ask specific datestyle from server.; timezone
Ask specific timezone from server.

AUTHENTICATION FILE FORMAT

PgBouncer needs its own user database. The users are loaded from a text file that should be in same format as PostgreSQL's pg_auth/pg_pwd file.: "username1" "password" ...
"username2" "md5abcdef012342345" ...
There shoud be at least 2 fields, surrounded by double quotes. The first field is the username and the second is either a plain-text or a MD5-hashed password. PgBouncer ignores the rest of the line.
This file format allows you to directly use the existing PostgreSQL authentication files in the Postgres data directory.

EXAMPLE

Minimal config: [databases]
template1 = host=127.0.0.1 dbname=template1; [pgbouncer]
pool_mode = session
listen_port = 6543
listen_addr = 127.0.0.1
auth_type = md5
auth_file = users.txt
logfile = pgbouncer.log
pidfile = pgbouncer.pid
admin_users = someuser
stats_users = stat_collector
Database defaults: [databases]; ; foodb over unix socket
foodb =; ; redirect bardb to bazdb on localhost
bardb = host=127.0.0.1 dbname=bazdb; ; access to destination database will go with single user
forcedb = host=127.0.0.1 port=300 user=baz password=foo client_encoding=UNICODE datestyle=ISO

docs.sk

comprehensive documentation repository

In section

See also