Обсуждение: Warm Standby log filling up with "FATAL: the database system is starting up" entries
Warm Standby log filling up with "FATAL: the database system is starting up" entries
От
"Greg Swisher"
Дата:
Hi Everybody,
I am not a linux expert, but am able to follow well enough documentation and such that I have been able to get postgresql 9.0.1 up and running on Debian Lenny in a primary + warm standby configuration, for the most part.
I am successfully sending WAL files to a standby server temp location, and the standby is able to read the log files after they arrive and process them. All seems to be working, however I am seeing 2 remaining issues for which I cannot find an answer. I turn to you!
On the standby server the postgresql.conf file is at pure default. The recovery.conf file has only the restore_command configured as:
restore_command = '/usr/local/pgsql/bin/pg_standby –d -s 2 -t /home/postgres/pgsql/pgsql.trigger /home/postgres/pgsql/backupdata %f %p %r 2>>/usr/local/pgsql/data/standby.log'
The logfile gets these “FATAL: the database system is starting up” entires on a regular basis. They are not logged at nearly the frequency that pg_standby is checking for the WAL files though, according to a comparison of the log files. I have not found any info on searches about this phenomena. (probably just don’t know where to look!)
LOG: database system was shut down in recovery at 2010-11-30 20:53:00 PST
LOG: starting archive recovery
LOG: restored log file "00000001000000000000001A" from archive
LOG: redo starts at 0/1A0000C0
LOG: consistent recovery state reached at 0/1B000000
FATAL: the database system is starting up
FATAL: the database system is starting up
FATAL: the database system is starting up
FATAL: the database system is starting up
FATAL: the database system is starting up
FATAL: the database system is starting up
And on...
And on...
And on...
Forever...
Any suggestions? I don’t want to just hide them unless they are harmless. Once I get things stable for awhile I will muck about with the log rotation and keeping them weeded down to size.
The second issue has to do with pg_standby logging.
If I use the –d switch, I get lots of good info in the standby.log file, but I also get a “WAL file not present yet. Checking for trigger file...” every time pg_standby checks. I tried removing the –d but then I get nothing at all. The documentation doesn’t indicate what is logged without the –d switch. I like getting the detail when a WAL file is found and acted upon, but I really don’t want to see pages and pages of status check messages.
Any suggestions here?
Thanks so much!!!!
Greg
I am not a linux expert, but am able to follow well enough documentation and such that I have been able to get postgresql 9.0.1 up and running on Debian Lenny in a primary + warm standby configuration, for the most part.
I am successfully sending WAL files to a standby server temp location, and the standby is able to read the log files after they arrive and process them. All seems to be working, however I am seeing 2 remaining issues for which I cannot find an answer. I turn to you!
On the standby server the postgresql.conf file is at pure default. The recovery.conf file has only the restore_command configured as:
restore_command = '/usr/local/pgsql/bin/pg_standby –d -s 2 -t /home/postgres/pgsql/pgsql.trigger /home/postgres/pgsql/backupdata %f %p %r 2>>/usr/local/pgsql/data/standby.log'
The logfile gets these “FATAL: the database system is starting up” entires on a regular basis. They are not logged at nearly the frequency that pg_standby is checking for the WAL files though, according to a comparison of the log files. I have not found any info on searches about this phenomena. (probably just don’t know where to look!)
LOG: database system was shut down in recovery at 2010-11-30 20:53:00 PST
LOG: starting archive recovery
LOG: restored log file "00000001000000000000001A" from archive
LOG: redo starts at 0/1A0000C0
LOG: consistent recovery state reached at 0/1B000000
FATAL: the database system is starting up
FATAL: the database system is starting up
FATAL: the database system is starting up
FATAL: the database system is starting up
FATAL: the database system is starting up
FATAL: the database system is starting up
And on...
And on...
And on...
Forever...
Any suggestions? I don’t want to just hide them unless they are harmless. Once I get things stable for awhile I will muck about with the log rotation and keeping them weeded down to size.
The second issue has to do with pg_standby logging.
If I use the –d switch, I get lots of good info in the standby.log file, but I also get a “WAL file not present yet. Checking for trigger file...” every time pg_standby checks. I tried removing the –d but then I get nothing at all. The documentation doesn’t indicate what is logged without the –d switch. I like getting the detail when a WAL file is found and acted upon, but I really don’t want to see pages and pages of status check messages.
Any suggestions here?
Thanks so much!!!!
Greg
On Mittwoch, 1. Dezember 2010, Greg Swisher wrote: > FATAL: the database system is starting up This is what you get on 8.4, when you try to connect to a standby. Rgds, Jens