Too many failed attempts to log into the web interface will result in your account being locked. You will get a message saying that your account has been frozen if you trigger it. If you are a student, please contact your TA. Otherwise, please contact us.
Although you can log into the web interface using your email address, you must use your actual account name to log into
users.isi.deterlab.net . Also, too many failed attempts to log into
users.isi.deterlab.net will result in an IP address ban. We automatically whitelist all IP addresses that have successfully logged into the web interface and this list is synchronized every 15 minutes. So if you find yourself banned from connecting to
users.isi.deterlab.net please log into the web interface and then wait.
Your home directory from
users is available on the nodes in your experiment. Copy your files to
users.isi.deterlab.net using scp or sftp to make them available on your nodes.
The reverse of the previous question: copy the files you want to your home directory, then download them from
users.isi.deterlab.net using scp or sftp.
The currently supported operating system images (see the Recommended list for currently supported images) images have access to full package repositories on a local mirror. Depending on your OS you can use
apt-get , or
pkg_add to install software that's been pre-packaged for each OS.
If there is no package for the software you wish to install, you can install from source. Copy the source tarball to the testbed (see How do I copy files from my workstation to a node in an experiment? or use
users ), then follow the package's installation instructions.
While we will do everything we can to assist any issues you face, we do not have the resources to help individual users install software.
Make sure you've followed the instructions How to set up an experiment with SEER.
This error is most frequently seen because you have not included a node named
control which is not connected to anything else in the topology.
Also make sure you've set the start command to run on every node, including
If you are a class user: the maximum number of nodes that a class can allocated is limited (see ClassResourceLimits for details). Wait for some of your classmates to free up resources before trying to swap in again.
You are less likely to encounter this during non-peak hours (late night and early morning) and when deadlines are distant.
If you are not a class user: please file a ticket because something is broken.
This may not be so frequent an error, but can arise when the experiment deliberately brings an interface down without bringing it back up prior to swap out. Try scheduling the link back up before the end of experiment.