If you see a problem on an Institute machine not listed here, please report it at www.msi.umn.edu/consult.html, send email to help@msi.umn.edu, or call 626-0802 (8:30am - 5:00pm, Monday through Friday).
Core Machines
There will be a downtime on Wed. May 7 beginning at 6:00am. This will affect all of the core machines (Altix, Calhoun, BladeCenter, Power4, and Netfinity). We expect the downtime to last approximately 12 hours. Duration of the downtime will vary depending on the system. During the downtime will be switching over to new LDAP servers. We will also be upgrading some of the networking in the machine room as part of our 10 Gb network upgrades. In addition, the following system specific maintenance will be performed: BladeCenter: update disk firmware and drivers on the nodes. Calhoun: apply latest OS patches and continue work on the new disk storage for scratch space. A reservation will be set so that jobs that do not complete before 6:00 am on May 7 will not be started. Accurate estimates of wall clock time will help schedule your jobs, especially as the downtime approaches.
The SP was retired on Monday, March 13 2006. Please see the announcement.
Altix Cluster
Wed May 7 09:32:07 CDT 2008 The Altix maintenance is complete. Of particular note, there is a MAJOR CHANGE in the max walltimes on many of the Altix nodes. * 6 of the 16-processor nodes have INCREASED to 300 hours. * The 256-processor node has REDUCED to 72 hours. * 2 of the 16-processor nodes have retained the maximum walltime of 24 hours. * The 42-processor node remains at a maximum walltime of 300 hours.
Calhoun (SGI Altix XE 1300 Linux Cluster)
Calhoun is available.
Thu May 8 15:07:22 CDT 2008 /scratch3 is now online, earlier than expected, bringing the total usable scratch space to 36TB.
Wed May 7 17:28:47 CDT 2008 The Calhoun maintenance is complete. There is currently 27 terabytes (TB) of usable scratch space, which is an increase from the 4TB of space available before the downtime. Data in the old /scratch1 has been moved to the new scratch storage. In addition, there are currently two additional new scratch filesystems called /scratch2 and /scratch4. All three of the available scratch filesystems are approximately 9TB. Another 9TB filesystem should appear at /scratch3 next week, for a total of 36TB of usable scratch space. Although all submission scripts on the system that use global scratch have been using /scratch1, please consider using some of the other new scratch spaces in order to balance performance.
BladeCenter Linux Cluster
The BladeCenter is available.
Netfinity Cluster
The Netfinity is available.
Power4 System
The Power4 System is available.
BSCL
All BSCL lab resources are available.
CGL
All CGL lab resources are available.
SDML
All SDML lab resources are available.
SDVL
All SDML lab resources are available.
BMSDL
All BMSDL lab resources are available.
Unisys ES7000-430
The Unisys ES7000-430 was retired December 31, 2007. Linux home directories were copied to regatta.msi.umn.edu:/homes/r60x/es7000. Please email help@msi.umn.edu if you have questions or concerns regarding this.
DTCCBL Compaq Alphas
The DTCCBL was retired in March 2007.
Altix:
Power4:
BladeCenter:
Calhoun:
Netfinity:
Downtime is required for regular maintenance of the machine.
Regular maintenance includes installation of the latest patches and
firmware upgrades. These installations require rebooting each node.
In general, a maintenance downtime will require one day.
However, this downtime may last longer. For example, if we
upgrade the operating system to a new version, a longer downtime
is necessary. As usual, all downtimes will be announced in the
message of the day shown when you login to the system, on this
page, and through email announcements.
Please plan your work accordingly.
This information is available in alternative formats upon request by individuals with disabilities. Please send email to alt-format@msi.umn.edu or call 612-624-0528. |
||||||||||||
|
HOME | QUESTIONS | FEEDBACK |
||||||||||||
|
||||||||||||