Cheatography
https://cheatography.com
Quick reference for most notable PostgreSQL settings and config options.
This is a draft cheat sheet. It is a work in progress and is not finished yet.
The config file(s)
Main config file is postgresql.conf in instance directory (or specified explicitly as startup argument).
Settings changes (incl. ALTER SYSTEM) are often not applied immediately: see the context property for how/when setting changes are applied.
Official doc: 19.1. Setting Parameters / 52.24. pg_settings |
|
Connection/network
listen_addresses |
interfaces to listen on (ipv4 and v6) |
port |
TCP port |
max_connections |
Max simultaneous sessions |
ssl |
Enable encrypted transport |
Logging
log_destination |
Log format |
log_filename |
Log filename pattern |
log_min_messages |
Minimum event severity to be logged |
log_min_duration_statement |
Log all statements with exec time > value |
log_checkpoint |
Log checkpoints |
log_line_prefix |
Prefix for all log lines |
log_lock_waits |
Log lock waits longer than value |
log_statement |
Log SQL statements (DML only, DDL only, ...) |
log_temp_files |
Log temp files larger than value |
WAL/checkpoints
wal_level |
Transaction log level: minimal (no log), replica, logical |
fsync |
Force flushing OS disk cache to physical storage |
wal_sync_method |
Method used for forcing WAL updates out to disk when fsync is on |
synchronous_commit |
Wait for writes to be committed to disk before sending commit to client |
full_page_writes |
Do not write partial pages to WAL |
wal_buffers |
Size of WAL buffers in RAM |
checkpoint_timeout |
Max delay between checkpoints |
checkpoint_completion_target |
Spread checkpoint over n times checkpoint_timeout |
min/max_wal_size |
Soft limits (targets) for WAL size |
Replication
archive_mode |
Enable WAL archiving |
archive_command |
Command used to archive WAL files |
max_wal_senders |
Max concurrent replication threads |
wal_keep_size |
Size of preserved WAL segments on the primary |
max_replication_slots |
Number of replication slots (max concurrent standby connections) |
hot_standby |
When in standby (in recovery), accept connections |
hot_standby_feedback |
When in standby, give status feedback to primary |
|
|
Resources
shared_buffers |
Instance-wide shared memory (see oracle sga) |
temp_buffers |
Dedicated memory per session (see oracle pga) |
work_mem |
Dedicated memory per query |
maintenance_work_mem |
Memory reserved for maintenance ops (vacuum, index creation, etc) |
autovacuum_work_mem |
Memory dedicated to autovacuum processes |
Autovacuum controls
autovacuum |
Enable/disable autovacuum |
autovacuum_max_workers |
Max parallel autovac jobs |
autovacuum_naptime |
Max delay before autovac |
autovacuum_vacuum_cost_limit |
Accumulated I/O cost of autovac before pausing |
autovacuum_vacuum_cost_delay |
Duration of autovacuum pause (see previous) |
autovacuum_vacuum_threshold |
Insert/update threshold to trigger autovac (absolute) |
autovacuum_vacuum_scale_factor |
Insert/update threshold to trigger autovac (relative) |
Autovacuum keeps count of I/O (as "cost"). Once cost has reached autovacuum_vacuum_cost_limit, it pauses for autovacuum_vacuum_cost_delay.
Vacuum threshold is the sum of vacuum_threshold and vacuum_scale_factor.
Autovacuum trigger thresholds have separate, identical parameters for inserts.
Planner/optimizer
seq_page_cost |
Sequential page read cost |
random_page_cost |
Random page read cost |
cpu_operator_cost |
Operator call cost |
effective_cache_size |
Estimated system cache size |
default_statistics_target |
"Target quality" of the optimizer: higher values takes longer to calculate plan but yields better results |
from_collapse_limit |
Merge nested FROM (to avoid "reorder experimentation" from the planner) if sub-FROM would yield less than this setting |
join_collapse_limit |
See from_collapse_limit but with JOINs instead of FROMs |
|
