On the weekend on 15/16th July there will be essential maintenance being carried out on the 5 West data-centre. For this work to be carried out, Balena will need to powered down for that weekend. Following on from this planned service interruption we are going to keep Balena offline for the following week to carry out our own essential maintenance on Balena, see details below.

Duration: 4pm, 14th of July to 21st July 2017

ClusterVision and the HPC Team will be working on the system during this period.

To note: -

  • All jobs will need to be dequeued before the Balena shutdown - we will put in a reservation to make sure there are no jobs in running state when the system in shutdown.
  • Balena will not be accessible, including your data, during this maintenance period.

Maintenance activities: -

  • Test emergency shutdown of Balena.
  • Upgrade OS (minor kernel update and security patches) on all nodes
  • Upgrade Bright Cluster Manager
  • Upgrade OFED (Infiniband driver + libraries) on all nodes and reinstall MPI libraries
  • Upgrade MySQL database on master nodes
  • BeeGFS - Parallel File-system
    • Firmware update
    • Enable user quota on inodes and data usage (informational only)
    • Re-tune meta-data partition

Please contact us at hpc-support@bath.ac.uk if you have any issues.

Team HPC

Posted in: Advancing Research Computing, High Performance Computing (HPC)

Respond

  • (we won't publish this)

Write a response