MeltDBWiki/DeveloperDocumentation/PersonalWebServer

Setting up a personal MeltDB Web Server at the CeBiTec

Here we describe how to set up a personal MeltDB Webserver within the CeBiTec environment.

The following documentation was written for the head version of the MeltDB CVS repository.

If you have not done so already, create a CVS directory in your home directory (e.g. `CVS_HEAD`) where you will store the current MeltDB sources:

cd
mkdir CVS_HEAD
cd CVS_HEAD

Check out the following modules from the CVS repository (make sure your `CVS_ROOT` and `CVS_RSH` environment variables are set, see the internal CVS documentation for further details):

cvs co bioinfo/common
cvs co bioinfo/MeltDB
cvs co bioinfo/MeltDB_web
cvs co bioinfo/projectmanagement
cvs co bioinfo/GOPArc
cvs co bioinfo/BRIDGE

Create a `user_conf.sh` file in the `CVS_HEAD/bioinfo/gendb_web/share/www/` directory. This file must contain the following lines:

PORT=9000
CVS_SUBDIR=CVS_HEAD 

The `PORT` variable defines the port where your web server process will wait for requests (here `9000` is an example which has to be substituted by a port number on your machine). Make sure this port is not used yet. The `CVS_SUBDIR` sets the directory relative to your home-directory where the checked out CVS `bioinfo` subdirectory can be found. In this example it is `CVS_HEAD`.

Open a remote shell on a host like bane, hook or trillian:

rsh bane

Change to the `MeltDB_web` directory on that host and run the `apache2` script with parameter `start`:

cd ~/CVS_HEAD/bioinfo/MeltDB_web
./apache2 start

This will start an httpd process on the current machine that listens on the port you provided in the `user_conf.sh` file.

If you now direct your browser to the corresponding address that you have configured for your web server, e.g.

http://bane:9000

you should see the login dialog of MeltDB.

Executing

cd ~/CVS_HEAD/bioinfo/MeltDB_web
./apache2 stop

will stop the web server. Do not forget to restart the web server if Perl modules were modified or updated. Changes in CGI scripts and templates will get updated automatically.

In order to read error messages that were produced by your web server, check the error logs in

~/CVS_HEAD/bioinfo/MeltDB_web/share/logs

A convenient way to analyze current errors is the use of

tail -f <logfile>

Make sure that you execute the `tail` command on the same host that the httpd process is running on (otherwise the results will be displayed with delay):

rsh bane
cd ~/CVS_HEAD/bioinfo/MeltDB_web/share/logs
tail -f MeltDB.errors

If you encounter problems, send an email to the gendb mailing list (MailTo(meltdb AT cebitec DOT uni DASH bielefeld DOT de)) or contact one of the MeltDB developers directly.