11-Node-Cluster with redundant RAID

  • total of eleven nodes, 7 nodes in main data centre 4 nodes in backup data centre .
  • one FC switch and one EMC-Symmetrix per data centre
  • cluster nodes connected redundant to RAID over separate FC switches
  • SRDF mirroring to backup data centre

This configuration guarantees

  • global availability of all disk resources on all cluster nodes
  • easy to use administration, no Symmetrix reconfiguration needed for reallocation of applications.
  • fast error recovery after RAID breakdown through SRDF, full assisted by OSL Storage Cluster (no reconfiguration required
  • OSL Multipath offers load distribution on all available paths and bypasses single path failures as well as switch failures transparently (e.g. online switch replacement)
  • on Node failure application is taking over by another cluster node (ACO - Application Control Option)
  • easy to use handling of EMC Timefinder data copies (e.g. backups, development environments) (XDM - Extended Data Management)

This way of setup (running SAP) is quite a typical case scenario for the usage of OSL Storage Cluster. Due to this setup the impact of complexity for daily administration is reduced as far as possible. With OSL Storage Cluster the administrator is able to simply "forget" about RAID and SAN configuration and focuses on using storage resources on the hosts. In this dimension the advantages of a global storage pool in regard to performance, flexibility and storage usage as effective as possible, becomes seriously noticeable for administrators as well as merchants.