Go to: Main Content
Minnesota Supercomputer Institute
- Home
- >
- Core Hardware
- >
- Policies and Procedures
Core Hardware Policies and Procedures
These Policies and Procedures are common to every MSI core resource. The pages for each core resource include any policies specific to each machine.
Ethical Guidelines
All users must follow the Supercomputing Institute's ethical guidelines. In particular, passwords must not be shared with anyone. Every individual with access to the facility should have his/her own user account and password.
Privacy
The default file permissions for the new user is read/write only by the user, but this may be changed. A principal investigator has the right to access data and files in his/her research group. In the event that the P.I. has difficulty doing this, he or she may contact the Research Programs Administrator or the Head of Technical Support.
Technical support staff do not read a user's email or files or alter them without the user's permission except in cases of emergency and only within the scope of administrative responsibilities.
Passwords
Users will have an MSI account and an associated password that will grant them access to various systems, labs, and some software that MSI provides. This single ID and password can then be used on all systems that the user has been granted access to. The user's password will automatically expire six months after the last password change or from initially setting the password for new accounts. See the full MSI password policy for more information.
Fairshare Scheduling
MSI uses a fairshare job scheduler to try to ensure a mix of jobs from all users can utilize any given core resource efficiently and fairly. The goal of fairshare is to increase the priority when scheduling jobs of the groups who are below their fairshare targets, and decrease the priority of jobs belonging to those groups whose usage exceeds their fairshare targets. In general, this means that when a group has recently used a large amount of resources, the priorities of their waiting jobs will be negatively affected until their usage decreases to their fairshare target once again.
For detailed information about fairshare scheduling, see the full user support information regarding fairshare scheduling.
Backups
Home directories are backed up every night. The nightly backups are kept for one month. Directories in which special arrangements have been made with the support staff (for example, /project directories) are also backed up nightly. Once a month, a copy of the nightly backup tapes is moved off site. This copy is kept for one month and is meant for disaster recovery (flood, fire). These tapes cannot be used for normal file recovery as this would require several days to restore and would therefore force us to suspend our regular nightly backups during this time.
The scratch filesystems (for example, /scratch1, /scratch.i12) and temporary directories (like /tmp) are not backed up.
Suggestions for backing up your data
- Keep two separate copies of anything you back up.
- Keep detailed records of what you've backed up, including date, file names, and device used.
- Verify your archive after you've written it.
How do I have data restored from systems backups?
Please send a request with the following information via email to user support. Include:
- name of system
- name of file(s) or directories you need to have restored
- approximate date that the file(s) existed on our systems
Service Unit Allocations
Service Units (SUs) are charged for computer time and large disk usage on core resources and for large disk storage at the SDML. A SU can be either a Computing Service Unit (CSU) or a Disk Service Units (DSU). The sum of CSU charges and DSU charges are added together as the total SU charge.
One SU will provide a fixed number of hours of CPU time as follows:
| Core Resource | CPU hours/SU |
|---|---|
| Regatta | 3.5 |
| Altix | 3.5 |
| Blade | 3.5 |
| Calhoun | 3.5 |
DSUs will be charged when a group's usage has exceeded 500 GB of disk usage on a system for more than an hour. There will not be a charge for disk usage on scratch spaces. DSUs will be charged as follows:
| Storage Location | GB hours/DSU |
|---|---|
| Core Resources | 2400 |
| SDML | 7200 |
There is no service unit charge for software usage, technical support, or for other laboratory resources.
Eligibility
Allocations are available to faculty members at the University of Minnesota and other faculty researchers at accredited institutions of higher education in the state of Minnesota for their work and that of their students and research groups. All proposals requesting an allocation must be sponsored and signed by a faculty member of the University of Minnesota or other accredited institution of post-secondary education in the state of Minnesota.
Supercomputing resource allocations are awarded based upon a competitive review (review criteria) by the Core Resource Allocation Subcommittee.
The deadline to apply for resources for a January to June allocation period is October 15 and the deadline to apply for a July to December allocation period is April 15.
July–December 2008 Allocation Summary
A total of 4,619,867 SUs were allocated for the July 1 through December 31, 2008 period. The following is a summary of allocations, showing a range of SUs and the number of groups receiving an allocation within that range.
| Number of SUs | Number of Groups |
|---|---|
| 1–500 | 50 |
| 501–1,000 | 28 |
| 1,001–5,000 | 36 |
| 5,001–10,000 | 16 |
| 10,001–50,000 | 22 |
| 50,001–100,000 | 6 |
| 100,001–200,000 | 5 |
| 200,001–300,000 | 3 |
| 300,001–400,000 | 3 |
| 401,000–500,000 | 1 |
Service Unit Account and Monitoring Usage
The Supercomputing Institute keeps track of machine usage using Service Units. Service Units awarded to a group can be used on any of the MSI core hardware. There are two allocation types to help a PI distribute SUs to their group, either make them available to all members of the group (Group Allocation) or assign certain amounts to individual users (User Allocation). The default setting is for group allocation. If you wish to change your allocation type, or have questions about accounting and monitoring usage, send email to user support.
You can monitor your usage by using the command acctinfo. Run by a member of a research group, the command acctinfo summarizes usage by that member on various machines. We can set up acctinfo so that the PI, or the PI and a named group member, can view usage by every member of the entire group. You can also get usage information from the previous allocation period.
Further information about SU accounts and monitoring SU usage is available in detail.