Jetstream allows you to quickly spin up new science resources in a managed cloud. You can even leverage other XSEDE resources, such as Wrangler, to retrieve or store additional data sets.
- To get started, choose one of the following featured images as source for a new Jetstream node:
- Ubuntu 14.04.3 Development
- Ubuntu 14.04.3 Development GUI
- CentOS 6 (6.8) Development
- CentOS 6 (6.8) Development GUI
- Centos 7 (7.2) Development
- Centos 7 (7.2) Development GUI
- Intel Development (CentOS 7)
- Click on the
Launchbutton to create a new instance based on this image.
- After the node finishes deployment, a compatible client package for iRODS will already be installed.
At the prompt, enter the username used by Wrangler for the collection. There is no need to match usernames within the Jetstream node and iRODS usernames. Multiple Jetstream usernames can access the same iRODS collection.
Next, you will be prompted to choose either the IU or TACC iRODS instance. Although both iRODS zones are linked, they contain separate data sets. It is best to communicate directly with the instance where the data collection is stored.
The "isetup" script will create the initial iRODS environment file, “irods_environment.json” in the .irods folder in your home directory and ask to overwrite it, if it already exists.
Once the proper environment is set, initialize a session using “iinit” using the Wrangler account credentials, which will use the Wrangler password for the account you specified. This creates a temporary credential which will be valid for two weeks. Once connected, the file system can be navigated and the standard iRODS commands are available.
The command “ihelp” lists the common iRODS commands, such as ils, ipwd, iget, iput, and imeta, which are used to transfer files to iRODS or perform tagging, queries, or other operations. Detailed instructions on using these commands can be found at the iRODS site.
Any files submitted into iRODS are given a checksum, and access is logged for reads and writes. These files will then also be available via the Wrangler data analysis system on XSEDE. It is possible to prototype program flow using Jetstream, and then move to Wrangler to perform larger calculations on the same data collection.