If you’ve found a bug in StarCluster’s code, please submit a bug report to the github issue tracker. In the case of an unrecoverable error, or crash, StarCluster will create a crash file containing debugging logs useful for diagnosing the issue. Below is an example of a crash due to a bug in the code:
% starcluster start -s 2 -v mycluster StarCluster - (http://star.mit.edu/cluster) Software Tools for Academics and Researchers (STAR) Please submit bug reports to email@example.com >>> Using default cluster template: smallcluster Traceback (most recent call last): File "/workspace/starcluster/starcluster/cli.py", line 155, in main sc.execute(args) File "/workspace/starcluster/starcluster/commands/start.py", line 180, in execute scluster = self.cm.get_cluster_template(template, tag) File "/workspace/starcluster/starcluster/cluster.py", line 82, in get_cluster_template ec2_conn=self.ec2) File "/workspace/starcluster/starcluster/config.py", line 589, in get_cluster_template clust = Cluster(ec2_conn, **kwargs) File "/workspace/starcluster/starcluster/cluster.py", line 315, in __init__ blah = blah UnboundLocalError: local variable 'blah' referenced before assignment !!! ERROR - Oops! Looks like you've found a bug in StarCluster !!! ERROR - Crash report written to: /home/myuser/.starcluster/logs/crash-report-6029.txt !!! ERROR - Please remove any sensitive data from the crash report !!! ERROR - and submit it to firstname.lastname@example.org
Each time StarCluster encounters a crash, as in the example above, a new crash report will be written to $HOME/.starcluster/logs/crash-report-$PID.txt where $PID is the process id of the buggy StarCluster session. When a crash report is generated users should first check the crash report for any sensitive data that might need to be removed and then submit a bug report with the crash report attached.
For all other issues, questions, feature requests, etc. you can either:
It’s highly preferred that bug reports are submitted to the github issue tracker if possible.