Running initdb
==============

TPA first creates ``postgres_data_dir`` if it doesn’t exist and ensures
it has the correct ownership, permissions, and SELinux context. Then,
unless the directory already contains a ``VERSION`` file, it runs
``initdb`` to initialize ``postgres_data_dir`` .

You can use the :ref:`pre-initdb hook <TPA hooks>` 

to execute tasks before ``postgres_data_dir`` is created and ``initdb``
is run. If the hook initializes ``postgres_data_dir`` , TPA finds the
``VERSION`` file and therefore doesn’t run ``initdb`` .

You can optionally set ``postgres_initdb_opts`` to a list of options to
pass to ``initdb`` :

.. code:: yaml

   cluster_vars:
     postgres_locale: de_DE.UTF-8
     postgres_initdb_opts:
     - --data-checksums

We recommend always including the ``--data-checksums`` option, which is
included by default.

When running ``initdb`` , TPA sets ``TZ=UTC`` in the environment and
sets ``LC_ALL`` to the ``postgres_locale`` you specify.

Separate configuration directory
--------------------------------

By default, ``postgres_conf_dir`` is equal to ``postgres_data_dir`` ,
and the Postgres configuration files (``postgresql.conf`` ,
``pg_ident.conf`` , ``pg_hba.conf`` , and the include files in
``conf.d`` ) are created in the data directory. If you change
``postgres_conf_dir`` , after running ``initdb`` , TPA moves the
generated configuration files to the new location.