note, I tried both with "#log_rotation_size = 10MB " - commented/default, and then enabled it as per documentation page both having similar result - no truncation. there is example on that documentation page that suggests that both size and time based rotation could be possible while truncation is enabled - which seem to contradict configuration's file comments:
log_truncate_on_rotation = on           # 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.

Also, in my original post there was a typo: I'm watching pg_log/ dir " , not pg_logs/ dir

Thanks!

vk wrote:
Hello everyone!
I'm trying to implement circular stderr logging:
postgresql 8.3.4 compiled from src
fedora 6 x86_64, Intel Core 2 Duo, 2GB RAM
/usr/local/pgsql/ partition mounted noatime

based on the following document managing circular/logging should be possible:
http://www.postgresql.org/docs/current/static/runtime-config-logging.html

however, I'm unable to implement it the way it is described on that page, pg_logs/ logs just keep on appending and not truncating as needed:

here are my logging settings (grep "^log_" and grep "^#log_" postgresql.conf ):

#log_destination = 'stderr' # Valid values are combinations of #log_directory = 'pg_log' # directory where log files are written, #log_filename = 'postgresql-%Y-%m-%d_%H%M%S.log' # log file name pattern, #log_min_messages = notice # values in order of decreasing detail: #log_error_verbosity = default # terse, default, or verbose messages #log_min_error_statement = error # values in order of decreasing detail:
#log_checkpoints = off
#log_connections = off
#log_disconnections = off
#log_duration = off
#log_hostname = off
#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 #log_timezone = unknown # actually, defaults to TZ environment
#log_parser_stats = off
#log_planner_stats = off
#log_executor_stats = off
#log_statement_stats = off
log_filename = 'postgresql.log.%S'
log_truncate_on_rotation = on # If on, an existing log file of the log_rotation_age = 1min # Automatic rotation of logfiles will log_rotation_size = 10MB # Automatic rotation of logfiles will
log_min_duration_statement = 0  # -1 is disabled, 0 logs all statements
log_line_prefix = '%t [%p]: [%l-1] %h '                 # special values:
log_autovacuum_min_duration = 0 # -1 disables, 0 logs all actions and

note the above config is prof of concept only, I'm not going to rotate logs every second on production.
for production I intend to have
"log_rotation_age = 1 h"
"log_filename = postgresql.log.%M"

I'm using pgbench to generate db/logs traffic:

while (true); do date; pgbench -s 5000 -c 10 -t 5000 PGBENCH; done

meanwhile watching pg_log dir:

after 50 minutes or so of this testing I never saw any log file to drop in size. after looking through individual log files, I confirm that old data is still there, and new data is just appended at the bottom.

how can I get truncate to work properly in this setup ? could you point me in the right direction ?
or could this be a bug ..

Thanks !
Vlad




--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general

Reply via email to