jbrowse-contrib

Configuration and miscellaneous support files for the AIP project JBrowse installation

Installing JBrowse

For the AIP project, please follow the steps outlined below for the setup and installation of the app:

### server: columbia
$ cd /opt/www/araport/htdocs

### clone the jbrowse-contrib repo
### remember to clone `--recursive` to get the submodule as well
### submodule `jbrowse-latest-release` points to latest release (https://github.com/GMOD/jbrowse/releases)
$ git clone --recursive https://github.com/Arabidopsis-Information-Portal/jbrowse-contrib.git .

### make a symlink and change into the release directory
### set up jbrowse by running the `setup.sh` script
$ ln -sf jbrowse-latest-release jbrowse
$ cd jbrowse
$ ./setup.sh

Point the web browser to http://araport-dev.jcvi.org/jbrowse to check that the app is functioning properly.

Loading data into JBrowse

Before data can be loaded into JBrowse, please following the instructions for setting up the data area from here: araport-contrib

Once the JBrowse setup has completed successfully and data has been downloaded to the right location, refer to the following steps to load data:

### server: columbia
$ cd /opt/www/araport/htdocs/jbrowse

### create symlink to shared data area
$ ln -sf ../../data/jbrowse data

### copy config file(s), modified css/html and loading script to JBrowse installation directory
$ cp ../jbrowse-contrib/jbrowse-load.sh .

### run the loading script
$ ./jbrowse-load.sh > jbrowse-load.log 2> jbrowse-load.err

### after loading, replace the trackList.json config with the customized copy from the repo
$ cp ../jbrowse-contrib/data/json/arabiodopsis/trackList.json data/json/arabidopsis/trackList.json

JCVI specific code/data migration workflow to production

Migrating to test

This involves making use of a command line mechanism to push the app from dev --> test.

### prepare to push from dev --> test

### edit `tarinclude` file which lists relative paths to code that needs to be migrated
### Example line(s) in `tarinclude`:

    ./htdocs/jbrowse-latest-release
    ./htdocs/jbrowse

$ cd /opt/www/araport
$ vim tarinclude

### invoke push from dev --> test, this step only copies the app code
### data area is already shared between the dev and test servers
$ make

Point the web browser to http://araport-test.jcvi.org/jbrowse to check that the app is functioning properly.

Synching with production

This mechanism is regulated by rdist, a remote file distribution client, which is controlled by a configuration file, called a distfile.

### prepare to synch test with prod
### to check the areas that are set up to be synched, invoke the following command:
$ cd /opt/www/araport
$ make getdistfile

### this will retrieve a file called `distfile.araport`
### contains a list of directories to be mirrored (and a list of excluded directories as well)

### synch the application
$ make test2pro

### synch the data (if this is the first time, process will take a while: ~0.5 hrs)
$ make test2pro_data

Once the synch is complete, point the web browser to http://apps.araport.org/jbrowse to check that the app is functioning properly.