Be advised, at least my experience, that many linux builds require certain "undocumented" configuration options. At least some of them are hard to find, and configure --help doesn't always show them. Without those, things do get put in different locations than where the rpm's do. Also, make sure you remove the current rpms. You might just want to do the remove, and reinstall the binaries anyway at first.
Sent from my iPad
I've used Ubuntu Repositories to be on the
safe side. If this is really the source of the problem I'll build
from source.
Sounds maybe like a botched
installation. The only pg_hba.conf that should be present is the
one in $PGDATA. Also, depending on your configuration and how
long it's been running, you may not have anything the pg_xlog
directory other than a subdirectory. I think, if I'm recalling
correctly, that the subdirectory is named archive. WAL segments
get written out here based on the setting for wal_keep_segments
in the postgresql.conf file.
Not really certain when pg_clogs get
written out,
--
Jay
Sent from my iPad
Hi everyone,
First of all it's appears that there was something wrong
with my Upstart conf file (Ubuntu is using Upstart instead
of SysV init. Next year they'll also move to Systemd). Jay
and Raghu: That solved my connection problem and /var/run
problem.
So now pg is up (according to ps -aux |grep post). However
pg_ctl status says no server
is running. What's wrong?
About the logs: I have pg_xlog and pg_clog under PFDATA -
nothing there.
Remote connection: When I try to connect to the server I
get an error that my user is not present at pg_hba.conf.
However, I see that I have 2 such files: One in PGDATA and
one in /etc/postgresql/9.3/main/. What's the difference
between them?
Thanks
Hi Shay,
Per your questions:
1) from my experience, /var/run
is only used to write a PID file, but this file is usually
only created by your system's service commands. I don't
use Ubuntu, only CentOS, so my command is "service
postgresql-9.3 [start | stop | status| ... ] and the
script writes the PID file. The CentOS script also writes
out a /var/subsys/lock file for postgresql.
2) your logfiles, unless you've
butchered the postgresql.conf file, live in the
$PGDATA/pg_log directory. Wherever that is on your system.
3) i probably can't help much here, but first if you're
running selinux, I'd probably turn that off first. Next,
check if iptables is running and possibly blocking your
response. Either turn that off or add a rule to permit
this port.
Hope this helps,
Jay
Sent from my iPad
Hello everybody,
I've just installed PostgreSQL and I'm unable to
connect to it remotely from my main desktop (I'm
trying to connect with pgAdmin 3).
My setup:
Ubuntu Server 14.04 x64 (Azure Cloud)
PostgreSQL v9.3.5-0ubuntu0.14.04.1
contrib package is also installed.
PostgreSQL is started on boot using Upstart. I've
modified this script: https://gist.github.com/haad/6020401
(just version numbers)
My Problems:
1) For some reason, when I've installed pg,
/var/run/postgresql was missing. I've manually created
this folder and gave my pg user write permissions. Is
that a bug?
2) I've tried to change the log config in
/etc/postgresql/9.3/main/postgresql.conf but it won't
work. I think that pg is ignoring that file. I also
have no idea where is pg log file (I tried
/var/log/postgresql - not there). I tried starting the
DB manually with pg_ctl -l but the log does say much:
LOG: database system is ready to accept
connections
LOG: autovacuum launcher started
LOG: received SIGHUP, reloading configuration
files
LOG: received smart shutdown request
LOG: autovacuum launcher shutting down
LOG: shutting down
LOG: database system is shut down
LOG: database system was shut down at 2014-10-17
21:22:11 UTC
LOG: database system is ready to accept
connections
LOG: autovacuum launcher started
This log contains a single DB run.
3) I'm unable to connect from my main desktop. Server
is listening according to netstat. There is no
firewall on the server and I've created an Azure
endpoint for port 5432. Ideas?
Can you help me?
Thanks,
|
