FreeBSD 7.0 and PostGreSQL 8.3 in Jails
Since the release of FreeBSD 7.0 many things have been changed to allow PostGres to run better than any other database under FreeBSD. To take advantage of this, I’ll show how to setup PostGreSQL in FreeBSD jails. Of course, the same method is used to install on your main system as well.
First off, we need to compile our kernel with a few different options.
# Shared Memory options SHMMAXPGS=400000 # Maximum size of shared memory segment (in pages) options SHMSEC=512 # Maximum number of shared memory segments per process options SEMMNI=512 # Maximum number of semaphore identifiers options SEMMNS=1024 # Maximum number of semaphores system-wide options SEMUME=200 # Maximum number of undo entries per process options SEMMNU=512 # number of undo structures in system
You may notice later that after installing PostGreSQL, it will mention the same thing, but with lesser amounts. Using the amounts given by the port won’t allow as many connections, and is optimized for a system with roughly around 256M of memory. The system I’m building this how-to around has over 2 gigabytes of RAM, and allocating 1.6 gigabytes for PostGreSQL in shared memory. If you are using less, you might want to divide SHMMAXPGS accordingly. The others, if you are using one gigabyte total, you might cut those in half. Keep in mind, this setup is fairly light on the server, considering it’s not meant to be a full-fledged PostGreSQL server.
For use of shared memory for jails, be sure to add this to your /etc/sysctl.conf:
Compile your new kernel, reboot, and then on to ports. If you want to check your allocated shared memory, you can check with:
sysctl -a | grep shmmax
The settings above will show ‘kern.ipc.shmmax: 1638400000’.
If you are planning to compile the database in a jail, be sure to mount the nullfs ports to the ports directory in the jail (refer to the FreeBSD 7.0 Jail How-To in order to see how this is done, if you have forgotten). Next, go to /usr/ports/databases/postgresql83-server, and do the usual, ‘make install clean’. After it is done installing, edit your /etc/rc.conf, and add:
If you don’t add this, the rc.d script will NOT work.
Next, you need to initialize the database. To do so, simply run:
With this done, it’s time to change the setup of the database. First, we go to /usr/local/pgsql/data, and edit the postgresql.conf. In here, we’ll set the bind-address, port, and whatever else we need. If you plan to have other jails, or the base system access the database, set the bind-address to the IP of the jail itself.
Next we move to the pg_hba.conf file. Here we set what hosts, and users we want to allow access to the database.
To start, we’ll add some sort of remote access, say, via VPN (10.211.0.0/16).
host all all 10.211.0.0/16 password md5
With that in place, we restart the server so that the rules are initialized. Now to create a user.
First we log into the pgsql account.
$ su pgsql $ createuser Enter name of role to add: testing Shall the new role be a superuser? (y/n) n Shall the new role be allowed to create databases? (y/n) y Shall the new role be allowed to create more new roles? (y/n) n
Now, let’s use the new user, create a table, and view it.
$ psql -d template1 -U testing template1=> create database testing; CREATE DATBASE template1=> \l List of databases Name | Owner | Encoding -----------+---------+----------- postgres | pgsql | SQL_ASCII template0 | pgsql | SQL_ASCII template1 | pgsql | SQL_ASCII testing | testing | SQL_ASCII (4 rows) template1=> \c testing; You are now connected to database "testing". testing=> create table public.test_table ( id char(2), name varchar(40)); CREATE TABLE testing=> \d List of relations Schema | Name | Type | Owner --------+------------+-------+--------- public | test_table | table | testing (1 row) testing=>
Now, for future reference, let’s see how to see users, and change the password of users. The database template1 is the main default database for PostGreSQL, and holds the user configurations.
testing=> \c template1 You are now connected to database "template1". template1=> alter user testing with password 'testme'; ALTER ROLE
Since you added a role in the pg_hba.conf file, you can connect remotely if needed with the following syntax:
So, that’s it for this tutorial. You should now have a basic understanding of PostGreSQL, and be able to install it with full optimization in FreeBSD, and even in a jail if needed. Enjoy!