GIMI Documentation

Services

For each service we need documentation for installation and administration.

Labwiki

[Jack]

Job Service

[Thierry]

OML server

[Olivier]

AMQP

[Thierry]

GIMI Experiment Service

[Jack or Thierry?]

iRODS

iRODS, which stands for integrated Rule-Oriented Digital System, is an open-source data management software in use at research organizations and government agencies worldwide. iRODS acts as middleware sitting on top of the distributed storage systems and underneath the user applications that require archiving, processing and disseminating data in a transparent and automated fashion. For more detailed and comprehensive descriptions, please refer to ​http://irods.org/.

In GIMI, we use iRODS as the back-end system to provide archive and automated processing functions to experimenters. We configured two iRODS instances, one as the production data system, the other as the environment for development purposes. The production system has the iCAT running on geni-gimi.renci.org, and two data servers running on geni-gimi.renci.org and emmy9.casa.umass.edu. geni-gimi.renci.org is currently allocated with ~2TB disk space, which can be further increased depending on its usage. To interact with iRODS, various user interfaces are provided. For example, iCommand (​https://wiki.irods.org/index.php/icommands) provides a UNIX-like command line interface to users who are already familiar with UNIX commands (such as ls, cp, rm, etc). iRODS also provides a rich set of APIs. For example, Jargon, a JAVA client API for iRODS, provides the REST functions so that GENI experimenters can easily create their iRODS accounts via the GENI portal. iRODS is equipped with a powerful rule language and rule engine. We developed iRODS policies and rules that recognize metadata files created based on specific schemas and automate data ingestion to help experimenters more efficiently organize and share their data.

Currently our iRODS instances are running release 3.3.

Images

[Divya]

InstaGENI Ubuntu 12.04 Image

Image URN: urn:publicid:IDN+instageni.illinois.edu+image+ch-geni-net:gimidev2
Image URL: https://www.instageni.illinois.edu/image_metadata.php?uuid=df9ce62b-ec1a-11e3-bc1a-000000000000

Installation Procedure for Ubuntu 12.04

The detailed procedure for installing OMF6 RC maybe found ​at this link and for installing OML ​at this link

In brief:

1. Boot IG VM with Default Ubuntu 12.04 image
   
2. Login to VM and continue as root
   
3. Install OMF RC
   

#apt-get install ruby1.9.1-dev ruby1.9.1
#apt-get install ruby1.9.3
#gem install omf_rc --pre
#gem install file-tail (Optional*)
#install_omf_rc -i -c

4. Test if OMF RC is installed

#install_omf_rc -i -c
#echo "---
:uid: samplehostname2
:uri: amqp://gimi3.casa.umass.edu
:environment: production
:debug: false" > /etc/omf_rc/config.yml
#start omf_rc
tail /var/log/omf_rc.log

After running the above you should see

[2014-06-04 22:29:50 +0000] INFO  OmfRc::Runner: Starting OMF Resource Controller version '6.1.1'
[2014-06-04 22:29:50 +0000] INFO  OmfRc::Runner: Connected using {:proto=>:amqp, :user=>"guest", :domain=>"127.0.0.1"}

5. Install OML Client

$sudo sh -c "echo 'deb http://download.opensuse.org/repositories/home:/cdwertmann:/oml/xUbuntu_12.04/ /' >> /etc/apt/sources.list.d/oml2.list"
$wget http://download.opensuse.org/repositories/home:cdwertmann:oml/xUbuntu_12.04/Release.key
$sudo apt-key add - < Release.key 
$sudo apt-get update
$sudo apt-get install oml2
$sudo apt-get install oml2-apps

*Only required if you want to parse data from a continuously growing file as shown [ here]

Postboot scripts

[Divya]

What is done in the postboot script? What may need to change over time?

Tool Usage

[Divya/Jeanne]

• Links to tutorials
• Links to instructions on how to write OEDL scripts
• Links to instructions on how to add OML capability to applications