2017

  • 2017-02-20 Added nodes from HPC cluster and HYDRA cluster and introduced the queue "old"
  • 2017-02 Adding new Broadwell nodes and older nodes from Hydra and HPC cluster
  • 2017-01-02 Upgrade of all server and nodes and software modules (see here)

2016

  • 2016-06-16 - sccdata and scc had to be rebooted due to a network problem. Also the GPUs on scc116 and scc117 failed and made a reboot necessary.
  • 2016-06-08 - Added tips for data copy/move on Data Storage page
  • 2016-06-08 - Added separate queue "gpu" for GPU-Jobs. Use "-q gpu" instead of "-q scc -l gpu".
  • 2016-06-01 - Updated monitoring of the cluster (see here)
  • 2016-05-27 - Updated external LAN connections to 4x10 GE.
  • 2016-05-22/23 - (12:43-9:00) unplanned outage due to a kernel crash on the main file server sccdata. The incident is being investigated
  • 2016-05-05/06 - (18:55-10:05) unplanned outage due to a kernel crash on the main file server sccdata
  • 2016-04-01 - In March 2016 we had a new record regarding Core-h/month (ca. 1.57 Mio. Core-h/month)
  • 2016-02-24 - Added 2 Haswell nodes with each one Tesla K80 GPU
  • 2016-02-02 - Added 4 new Haswell nodes (24 cores, 512 GB each)
  • 2016-01-26 - Moved hydra, hydra05, hydra06 to N4. Moved scc011, scc012 and scc051 to N4.
  • 2016-01-22 - Added 8 new Haswell nodes (24 cores, 128 GB each)
  • 2016-01-20 - Added 8 dual Sandy Bridge nodes (16 cores each) (from SPHINX)

2015

  • 2015-12-01 - scc now uses RSA, ECDSA or DSA ssh host keys. Check the fingerprints on the login page.
  • 2015-11-20 (8:40-10:05) - file server outage due to heavy I/O was fixed.
  • 2015-10-26 - The cluster is up and running again
  • 2015-10-12 .. 2015-10-26 - The cooling system of the server room will be turned off for maintenance. All jobs will be suspended for this time. New jobs will be started and suspended jobs will continue after the maintenance.
  • 2015-08-20 .. 2015-08-23 - Restoring /data/scc from backup. Please see /work/README for progress informations.
  • 2015-08-20 - Installed a new backup server with 256 TB storage. All data spaces will be backuped to this server in the next weeks.
  • 2015-08-20 - The storage of sccdata (/data/scc and /home/scc) was optimized for better performance and stability.
  • 2015-08-17 - All jobs were suspended due to cooling system updates in the server room.
  • 2015-08-14 ... 2015-08-20 - The file system /work (/data/scc) was disabled due to storage system error. All data will be restored from backup.
  • 2015-07-21/22 - The power infrastructure in the server room was extended to be prepared for the new racks coming in August. Necessary short network outages may have harmed MPI jobs.
  • 2015-07-10 - Extended storage of sccdata2 to 120 TB
  • 2015-06-11 - The SCCKN Website moved to www.scc.uni-konstanz.de. There may be some broken links until everything is correctly setup.
  • 2015-01-15 - Improved statistics
  • 2015-01-09 - Added SPHINX Cluster to SCC queuing
  • 2015-01-07 - Moved Haswell nodes to the server room

2014

  • 2014-12-20 - Added new Haswell nodes, Updated website for phase 7
  • 2014-12-14 - Updated website for phase 6 (old Sandy Bridge nodes), Added HYDRA cluster
  • 2014-12-09/10 - Moved SCC to the new server room in N4 and adding old nodes, Updated Grid Engine to version 8.1.8
  • 2014-11-24 - Added resource complexes to allow users to limit the number of used cores (see here)
  • 2014-10-13 - Integrated NESSIE Cluster into SCC Queuing.
  • 2014-09-29 - Installed new Ivy Bridge nodes (4 x 256 GB, 12 x 64 GB, 2 x GPU)
  • 2014-06-25 - Improved job scheduling by adding resource options "cputype" and "exclusive"
  • 2014-05-14 - 8am: part of the nodes are suspended for a repair of the cooling system in P5
  • 2014-04-27 - Use starter method for all queues to set OMP_NUM_THREADS and Modules environment
  • 2014-04-17 - Added workstation queue "pc"
  • 2014-01-21 - Fixed bug in parallel environment "mpi" when using not integrated versions of MPI. Added "hybrid" parallel environment for multithreaded MPI applications.
  • 2014-01-17 - Updated usage statistics.
  • 2014-01-07 - SCC up and running again after the University power tests.

2013

  • 2013-12-13 - Fixed a memory reservation problem with MPI over Infiniband
  • 2013-12-12 - Fixed issue with x2go server
  • 2013-12-11 - Short Maintenance to add nodes and improve server performance (FINISHED)
  • 2013-12-04 - SCC up and running (phase 1)
  • 2013-10-17 - improved Monitoring page
  • 2013-10-04 - added Hardware section, Monitoring section and Simulation job documentation
  • 2013-09-24 - End of public tender (Ausschreibung)
  • 2013-09-16 - added Compression tutorial, Usage and Data storage page
  • 2013-09-13 - completed Home page
  • 2013-09-06 - improved Grid Engine tutorial
  • 2013-07-19 - finished Software section
  • 2013-07-10 - added more software packages
  • 2013-07-03 - started with Software section
  • 2013-06-28 - added Checkpointing tutorial
  • 2013-06-27 - added Grid Engine tutorial
  • 2013-06-25 - added Modules tutorial
  • 2013-06-25 - added website (Home+Contact)

see also Software Updates