HPC Cluster Backup and Recovery Scheme

Keeping the data of its Health Care HPC cluster safe at all times is imperative for Bayer. We produced a robust backup and recovery scheme for Bayer’s HPC stakeholders to insure that important data is always safeguarded and that in the event of a possible incident the system would recover rapidly and safely in a minimum of time.

Though sturdy the recovery scheme we designed allows retrieving important data quickly thus minimising downtime.

How we helped

• Creation of a backup and recovery scheme document, a companion technical note and a sample user manual
• The documents provided recommendations in a number of areas relating to HPC cluster backup and recovery
• A data classification approach
• Storage housekeeping policies
• Restore prioritisation policies
• Recovering from catastrophic failure
• Recovering accidentally deleted data
• Managing the Tivoli Storage Management (TSM) backup footprint
• Migration of inappropriate applications and data from HPC

Results

• Designed a complete backup and recovery plan for Bayer
• If implemented correctly it will prevent or minimise possible negative business impact in a possible system failure