• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 1535
  • Last Modified:

Increase Shared Buffer Size in Postgres

I have a MS 2003 Enterprise Windows Server with SP2, 2Ghz, 32 GB RAM sever which hosts both application (dotnet and java) and database (postgres 8.4) on the same m/c.
Following is its conf file , have to optimize it for better performance, kindly suggest the settings
# -----------------------------
# PostgreSQL configuration file
# -----------------------------
#
# This file consists of lines of the form:
#
#   name = value
#
# (The "=" is optional.)  Whitespace may be used.  Comments are introduced with
# "#" anywhere on a line.  The complete list of parameter names and allowed
# values can be found in the PostgreSQL documentation.
#
# The commented-out settings shown in this file represent the default values.
# Re-commenting a setting is NOT sufficient to revert it to the default value;
# you need to reload the server.
#
# This file is read on server startup and when the server receives a SIGHUP
# signal.  If you edit the file on a running system, you have to SIGHUP the
# server for the changes to take effect, or use "pg_ctl reload".  Some
# parameters, which are marked below, require a server shutdown and restart to
# take effect.
#
# Any parameter can also be given as a command-line option to the server, e.g.,
# "postgres -c log_connections=on".  Some parameters can be changed at run time
# with the "SET" SQL command.
#
# Memory units:  kB = kilobytes        Time units:  ms  = milliseconds
#                MB = megabytes                     s   = seconds
#                GB = gigabytes                     min = minutes
#                                                   h   = hours
#                                                   d   = days


#------------------------------------------------------------------------------
# FILE LOCATIONS
#------------------------------------------------------------------------------

# The default values of these variables are driven from the -D command-line
# option or PGDATA environment variable, represented here as ConfigDir.

#data_directory = 'ConfigDir'            # use data in another directory
                              # (change requires restart)
#hba_file = 'ConfigDir/pg_hba.conf'      # host-based authentication file
                              # (change requires restart)
#ident_file = 'ConfigDir/pg_ident.conf'      # ident configuration file
                              # (change requires restart)

# If external_pid_file is not explicitly set, no extra PID file is written.
#external_pid_file = '(none)'            # write an extra PID file
                              # (change requires restart)


#------------------------------------------------------------------------------
# CONNECTIONS AND AUTHENTICATION
#------------------------------------------------------------------------------

# - Connection Settings -

listen_addresses = '*'            # what IP address(es) to listen on;
                              # comma-separated list of addresses;
                              # defaults to 'localhost', '*' = all
                              # (change requires restart)
port = 5432                        # (change requires restart)
max_connections = 100                  # (change requires restart)
# Note:  Increasing max_connections costs ~400 bytes of shared memory per
# connection slot, plus lock space (see max_locks_per_transaction).
#superuser_reserved_connections = 3      # (change requires restart)
#unix_socket_directory = ''            # (change requires restart)
#unix_socket_group = ''                  # (change requires restart)
#unix_socket_permissions = 0777            # begin with 0 to use octal notation
                              # (change requires restart)
#bonjour_name = ''                  # defaults to the computer name
                              # (change requires restart)

# - Security and Authentication -

#authentication_timeout = 1min            # 1s-600s
#ssl = off                        # (change requires restart)
#ssl_ciphers = 'ALL:!ADH:!LOW:!EXP:!MD5:@STRENGTH'      # allowed SSL ciphers
                              # (change requires restart)
#password_encryption = on
#db_user_namespace = off

# Kerberos and GSSAPI
#krb_server_keyfile = ''
#krb_srvname = 'postgres'            # (Kerberos only)
#krb_caseins_users = off

# - TCP Keepalives -
# see "man 7 tcp" for details

#tcp_keepalives_idle = 0            # TCP_KEEPIDLE, in seconds;
                              # 0 selects the system default
#tcp_keepalives_interval = 0            # TCP_KEEPINTVL, in seconds;
                              # 0 selects the system default
#tcp_keepalives_count = 0            # TCP_KEEPCNT;
                              # 0 selects the system default


#------------------------------------------------------------------------------
# RESOURCE USAGE (except WAL)
#------------------------------------------------------------------------------

# - Memory -

shared_buffers = 1024MB                  # min 128kB
                              # (change requires restart)
#temp_buffers = 8MB                  # min 800kB
#max_prepared_transactions = 0            # zero disables the feature
                              # (change requires restart)
# Note:  Increasing max_prepared_transactions costs ~600 bytes of shared memory
# per transaction slot, plus lock space (see max_locks_per_transaction).
# It is not advisable to set max_prepared_transactions nonzero unless you
# actively intend to use prepared transactions.
#work_mem = 100MB                        # min 64kB
#maintenance_work_mem = 16MB            # min 1MB
#max_stack_depth = 2MB                  # min 100kB

# - Kernel Resource Usage -

#max_files_per_process = 1000            # min 25
                              # (change requires restart)
shared_preload_libraries = '$libdir/plugins/plugin_debugger.dll,$libdir/pg_stat_statements.dll'            # (change requires restart)

# - Cost-Based Vacuum Delay -

#vacuum_cost_delay = 0ms            # 0-100 milliseconds
#vacuum_cost_page_hit = 1            # 0-10000 credits
#vacuum_cost_page_miss = 10            # 0-10000 credits
#vacuum_cost_page_dirty = 20            # 0-10000 credits
#vacuum_cost_limit = 200            # 1-10000 credits

# - Background Writer -

#bgwriter_delay = 200ms                  # 10-10000ms between rounds
#bgwriter_lru_maxpages = 100            # 0-1000 max buffers written/round
#bgwriter_lru_multiplier = 2.0            # 0-10.0 multipler on buffers scanned/round

# - Asynchronous Behavior -

#effective_io_concurrency = 1            # 1-1000. 0 disables prefetching


#------------------------------------------------------------------------------
# WRITE AHEAD LOG
#------------------------------------------------------------------------------

# - Settings -

#fsync = on                        # turns forced synchronization on or off
#synchronous_commit = on            # immediate fsync at commit
#wal_sync_method = fsync            # the default is the first option
                              # supported by the operating system:
                              #   open_datasync
                              #   fdatasync
                              #   fsync
                              #   fsync_writethrough
                              #   open_sync
#full_page_writes = on                  # recover from partial page writes
#wal_buffers = 64kB                  # min 32kB
                              # (change requires restart)
#wal_writer_delay = 200ms            # 1-10000 milliseconds

#commit_delay = 0                  # range 0-100000, in microseconds
#commit_siblings = 5                  # range 1-1000

# - Checkpoints -

#checkpoint_segments = 3            # in logfile segments, min 1, 16MB each
#checkpoint_timeout = 5min            # range 30s-1h
#checkpoint_completion_target = 0.5      # checkpoint target duration, 0.0 - 1.0
#checkpoint_warning = 30s            # 0 disables

# - Archiving -

#archive_mode = off            # allows archiving to be done
                        # (change requires restart)
#archive_command = ''            # command to use to archive a logfile segment
#archive_timeout = 0            # force a logfile segment switch after this
                        # number of seconds; 0 disables


#------------------------------------------------------------------------------
# QUERY TUNING
#------------------------------------------------------------------------------

# - Planner Method Configuration -

#enable_bitmapscan = on
#enable_hashagg = on
#enable_hashjoin = on
#enable_indexscan = on
#enable_mergejoin = on
#enable_nestloop = on
#enable_seqscan = on
#enable_sort = on
#enable_tidscan = on

# - Planner Cost Constants -

#seq_page_cost = 1.0                  # measured on an arbitrary scale
#random_page_cost = 4.0                  # same scale as above
#cpu_tuple_cost = 0.01                  # same scale as above
#cpu_index_tuple_cost = 0.005            # same scale as above
#cpu_operator_cost = 0.0025            # same scale as above
#effective_cache_size = 128MB

# - Genetic Query Optimizer -

#geqo = on
#geqo_threshold = 12
#geqo_effort = 5                  # range 1-10
#geqo_pool_size = 0                  # selects default based on effort
#geqo_generations = 0                  # selects default based on effort
#geqo_selection_bias = 2.0            # range 1.5-2.0

# - Other Planner Options -

#default_statistics_target = 100      # range 1-10000
#constraint_exclusion = partition      # on, off, or partition
#cursor_tuple_fraction = 0.1            # range 0.0-1.0
#from_collapse_limit = 8
#join_collapse_limit = 8            # 1 disables collapsing of explicit
                              # JOIN clauses


#------------------------------------------------------------------------------
# ERROR REPORTING AND LOGGING
#------------------------------------------------------------------------------

# - Where to Log -

log_destination = 'stderr'            # Valid values are combinations of
                              # stderr, csvlog, syslog and eventlog,
                              # depending on platform.  csvlog
                              # requires logging_collector to be on.

# This is used when logging to stderr:
logging_collector = on            # Enable capturing of stderr and csvlog
                              # into log files. Required to be on for
                              # csvlogs.
                              # (change requires restart)

# These are only used if logging_collector is on:
#log_directory = 'pg_log'            # directory where log files are written,
                              # can be absolute or relative to PGDATA
#log_filename = 'postgresql-%Y-%m-%d_%H%M%S.log'      # log file name pattern,
                              # can include strftime() escapes
#log_truncate_on_rotation = off            # If on, an existing log file of the
                              # same name as the new log file will be
                              # truncated rather than appended to.
                              # But such truncation only occurs on
                              # time-driven rotation, not on restarts
                              # or size-driven rotation.  Default is
                              # off, meaning append to existing files
                              # in all cases.
#log_rotation_age = 1d                  # Automatic rotation of logfiles will
                              # happen after that time.  0 disables.
#log_rotation_size = 10MB            # Automatic rotation of logfiles will
                              # happen after that much log output.
                              # 0 disables.

# These are relevant when logging to syslog:
#syslog_facility = 'LOCAL0'
#syslog_ident = 'postgres'

#silent_mode = off                  # Run server silently.
                              # DO NOT USE without syslog or
                              # logging_collector
                              # (change requires restart)


# - When to Log -

#client_min_messages = notice            # values in order of decreasing detail:
                              #   debug5
                              #   debug4
                              #   debug3
                              #   debug2
                              #   debug1
                              #   log
                              #   notice
                              #   warning
                              #   error

#log_min_messages = warning            # values in order of decreasing detail:
                              #   debug5
                              #   debug4
                              #   debug3
                              #   debug2
                              #   debug1
                              #   info
                              #   notice
                              #   warning
                              #   error
                              #   log
                              #   fatal
                              #   panic

#log_error_verbosity = default            # terse, default, or verbose messages

#log_min_error_statement = error      # values in order of decreasing detail:
                               #   debug5
                              #   debug4
                              #   debug3
                              #   debug2
                              #   debug1
                               #   info
                              #   notice
                              #   warning
                              #   error
                              #   log
                              #   fatal
                              #   panic (effectively off)

#log_min_duration_statement = -1      # -1 is disabled, 0 logs all statements
                              # and their durations, > 0 logs only
                              # statements running at least this number
                              # of milliseconds


# - What to Log -

#debug_print_parse = off
#debug_print_rewritten = off
#debug_print_plan = off
#debug_pretty_print = on
#log_checkpoints = off
#log_connections = off
#log_disconnections = off
#log_duration = off
#log_hostname = off
log_line_prefix = '%t'                  # special values:
                              #   %u = user name
                              #   %d = database name
                              #   %r = remote host and port
                              #   %h = remote host
                              #   %p = process ID
                              #   %t = timestamp without milliseconds
                              #   %m = timestamp with milliseconds
                              #   %i = command tag
                              #   %c = session ID
                              #   %l = session line number
                              #   %s = session start timestamp
                              #   %v = virtual transaction ID
                              #   %x = transaction ID (0 if none)
                              #   %q = stop here in non-session
                              #        processes
                              #   %% = '%'
                              # e.g. '<%u%%%d> '
#log_lock_waits = off                  # log lock waits >= deadlock_timeout
#log_statement = 'none'                  # none, ddl, mod, all
#log_temp_files = -1                  # log temporary files equal or larger
                              # than the specified size in kilobytes;
                              # -1 disables, 0 logs all temp files
#log_timezone = unknown                  # actually, defaults to TZ environment
                              # setting


#------------------------------------------------------------------------------
# RUNTIME STATISTICS
#------------------------------------------------------------------------------

# - Query/Index Statistics Collector -

#track_activities = on
#track_counts = on
#track_functions = none                  # none, pl, all
track_activity_query_size = 6144
#update_process_title = on
#stats_temp_directory = 'pg_stat_tmp'


# - Statistics Monitoring -

#log_parser_stats = off
#log_planner_stats = off
#log_executor_stats = off
#log_statement_stats = off


#------------------------------------------------------------------------------
# AUTOVACUUM PARAMETERS
#------------------------------------------------------------------------------

#autovacuum = on                  # Enable autovacuum subprocess?  'on'
                              # requires track_counts to also be on.
#log_autovacuum_min_duration = -1      # -1 disables, 0 logs all actions and
                              # their durations, > 0 logs only
                              # actions running at least this number
                              # of milliseconds.
#autovacuum_max_workers = 3            # max number of autovacuum subprocesses
#autovacuum_naptime = 1min            # time between autovacuum runs
#autovacuum_vacuum_threshold = 50      # min number of row updates before
                              # vacuum
#autovacuum_analyze_threshold = 50      # min number of row updates before
                              # analyze
#autovacuum_vacuum_scale_factor = 0.2      # fraction of table size before vacuum
#autovacuum_analyze_scale_factor = 0.1      # fraction of table size before analyze
#autovacuum_freeze_max_age = 200000000      # maximum XID age before forced vacuum
                              # (change requires restart)
#autovacuum_vacuum_cost_delay = 20ms      # default vacuum cost delay for
                              # autovacuum, in milliseconds;
                              # -1 means use vacuum_cost_delay
#autovacuum_vacuum_cost_limit = -1      # default vacuum cost limit for
                              # autovacuum, -1 means use
                              # vacuum_cost_limit


#------------------------------------------------------------------------------
# CLIENT CONNECTION DEFAULTS
#------------------------------------------------------------------------------

# - Statement Behavior -

#search_path = '"$user",public'            # schema names
#default_tablespace = ''            # a tablespace name, '' uses the default
#temp_tablespaces = ''                  # a list of tablespace names, '' uses
                              # only default tablespace
#check_function_bodies = on
#default_transaction_isolation = 'read committed'
#default_transaction_read_only = off
#session_replication_role = 'origin'
#statement_timeout = 0                  # in milliseconds, 0 is disabled
#vacuum_freeze_min_age = 50000000
#vacuum_freeze_table_age = 150000000
#xmlbinary = 'base64'
#xmloption = 'content'

# - Locale and Formatting -

datestyle = 'iso, mdy'
#intervalstyle = 'postgres'
#timezone = unknown                  # actually, defaults to TZ environment
                              # setting
#timezone_abbreviations = 'Default'     # Select the set of available time zone
                              # abbreviations.  Currently, there are
                              #   Default
                              #   Australia
                              #   India
                              # You can create your own file in
                              # share/timezonesets/.
#extra_float_digits = 0                  # min -15, max 2
#client_encoding = sql_ascii            # actually, defaults to database
                              # encoding

# These settings are initialized by initdb, but they can be changed.
lc_messages = 'English_United States.1252'                  # locale for system error message
                              # strings
lc_monetary = 'English_United States.1252'                  # locale for monetary formatting
lc_numeric = 'English_United States.1252'                  # locale for number formatting
lc_time = 'English_United States.1252'                        # locale for time formatting

# default configuration for text search
default_text_search_config = 'pg_catalog.english'

# - Other Defaults -

#dynamic_library_path = '$libdir'
#local_preload_libraries = ''


#------------------------------------------------------------------------------
# LOCK MANAGEMENT
#------------------------------------------------------------------------------

#deadlock_timeout = 1s
#max_locks_per_transaction = 64            # min 10
                              # (change requires restart)
# Note:  Each lock table slot uses ~270 bytes of shared memory, and there are
# max_locks_per_transaction * (max_connections + max_prepared_transactions)
# lock table slots.


#------------------------------------------------------------------------------
# VERSION/PLATFORM COMPATIBILITY
#------------------------------------------------------------------------------

# - Previous PostgreSQL Versions -

#add_missing_from = off
#array_nulls = on
#backslash_quote = safe_encoding      # on, off, or safe_encoding
#default_with_oids = off
#escape_string_warning = on
#regex_flavor = advanced            # advanced, extended, or basic
#sql_inheritance = on
#standard_conforming_strings = off
#synchronize_seqscans = on

# - Other Platforms and Clients -

#transform_null_equals = off


#------------------------------------------------------------------------------
# CUSTOMIZED OPTIONS
#------------------------------------------------------------------------------

custom_variable_classes = 'pg_stat_statements'            # list of custom variable class names
pg_stat_statements.track = all

I wish to increase Shared Buffer Size to optimize CP usage
 
0
npglobal
Asked:
npglobal
  • 4
  • 3
1 Solution
 
gheistCommented:
32bit or 64 bit windows server? Standard or Enterprise?
0
 
npglobalAuthor Commented:
It's a 32 bit Windows Server Standarad Edition
0
 
npglobalAuthor Commented:
Clarifying

OS -32 bit Windows Server - Eneterprise Edition

Postgres DB - standard edition

Thanks
0
Free Backup Tool for VMware and Hyper-V

Restore full virtual machine or individual guest files from 19 common file systems directly from the backup file. Schedule VM backups with PowerShell scripts. Set desired time, lean back and let the script to notify you via email upon completion.  

 
gheistCommented:
Doe to heap fragmentation (by antiviruses, aslr etc) you cannot have significantly bigger shred buffers.
You might get lucky to use /3GB in boot.ini and get 1,5G
Shared buffers perform same functionality as disk cache, setting effective_cache_size does same.
say
shared_buffers=8MB
effective_cache_size=1024MB

will be equivalent to your configuration with less load on particual procese's address space

and will not reach memory space limitation

also your 100MB work_mem counts quickly towards 2/3/4GB address space.
0
 
gheistCommented:
100MB x 100 connections = 10GB which 32bit windows can not provide.
0
 
npglobalAuthor Commented:
Provides me a starting point
0
 
gheistCommented:
EnterpriseDB includes config optimizer which might be of a good help.

0

Featured Post

Free Tool: ZipGrep

ZipGrep is a utility that can list and search zip (.war, .ear, .jar, etc) archives for text patterns, without the need to extract the archive's contents.

One of a set of tools we're offering as a way to say thank you for being a part of the community.

  • 4
  • 3
Tackle projects and never again get stuck behind a technical roadblock.
Join Now