FPS Shell And MTG Configuration

Shell Configuration

MTG_ROOT

Set MTG_ROOT to point to the directory that contains the FPS code.



MTG_JOB
MTG_TOOLSET

These values are used if no command line options were supplied.



PYTHONPATH

Add the FPS Python modules to the Python path. The directory will be $MTG_ROOT/python/modules



Add $MTG_ROOT/python/bin to the PATH environment variable.



For developers, add $MTG_ROOT/postgres/bin to the PATH environment variable.




MTG Configuration

Pick a job name. The job name should be all lowercase and only a few characters in length.

Database Admin

Create admin users for the database server. The admin users need the ability to create and delete databases.



Create the Databases

For a new job, create the following databases using the job name. The "dev", "test", and "prod" represent the tool set type.

dev_
<job_name>

Development database, used by developers to write tools.



test_<job_name>

Test database, used to test new tools, migrations to new schemas, etc.



prod_<job_name>

Production database.




There is one setup tool called pdm_tool. It is located in mtg_pub/trunk/sbin.

The tool uses the -job and the -dev, -test, -prod flags on the command line.

The bonus is that if you setup the config file correctly and then you don't have to remember any of that stuff. You just need to know the job name and the release.

To setup everything new:

pdm_tool -job <job_name> -dev setup_db


pdm_tool -job <job_name> -test setup_db


pdm_tool -job <job_name> -prod setup db



Be sure to read the prompts when you are doing the setup. You will be prompted for the password a couple of times. Use the password you have in your config file.




To delete your current schema and load test data:

pdm_tool -job <job_name> -dev dismantle_schema


pdm_tool -job <job_name> -dev load_test_data



To create the schema fresh, ready to be loaded with pdm_test:

pdm_tool -job <job_name> -dev dismantle_schema


pdm_tool -job <job_name> -dev setup_schema


run pdm_test



To dump the schema as test data:

pdm_tool -job <job_name> -dev dump_test_data



The dump_test_data function can be followed with either 'sm', 'md', or 'lg' to dump the small, medium, or large test data set. The default (nothing specified) is small. For example:

pdm_tool -job <job_name> -dev dump_test_data lg



would dump the schema into the large data set file.



Database Users

All users of a database will need username/password access to the database.




Create the MTG Config Files

Config files go in

$MTG_ROOT/config



There should be the following config files for a job:

job_dev_<job_name>.cfg


job_test_<job_name>.cfg


job_prod_<job_name>.cfg



Use job_template.cfg as a starting point. Change the values in the file as necessary.

Create the MTG Env Files

Env files go in

$MTG_ROOT/lib/tcsh/



There should be the following config files for a job:

fps_dev_<job_name>.env


fps_test_<job_name>.env


fps_prod_<job_name>.env



Use navajo_dev_example.env as a starting point. Change the values in the file as necessary. Source the file in users .tcshrc file.