Cluster Server 8.0 Agent for EMC SRDF/Star Configuration Guide - Windows

Last Published:
Product(s): InfoScale & Storage Foundation (8.0)
Platform: Windows

Typical EMC SRDF/Star setup using VCS clusters

Figure: Typical cluster setup in SRDF/Star shows a typical cluster setup in an EMC SRDF/Star configuration.

Figure: Typical cluster setup in SRDF/Star

Typical cluster setup in SRDF/Star

Clustering in an SRDF/Star environment typically consists of the following infrastructure:

The array at the workload site

The workload site comprises one or more workload hosts directly attached to a EMC Symmetrix array. The array contains R1 devices that are part of the SRDF/Star configuration.

The R1 devices must also be in an SRDF/S relationship with the regional sync target and in a concurrent SRDF/A relationship with the distant async target.

The array at the sync target

The sync target comprises one or more sync target hosts directly attached to a EMC Symmetrix array containing R2 devices that are part of the SRDF/Star configuration. The sync target array is paired with the workload array through a synchronous RDF link. The sync target hosts and array must be a sufficient distance away to survive a local disaster at the workload site, yet still be close enough to maintain synchronous replication.

The array at the async target

The async target comprises one or more async target hosts directly attached to a Symmetrix array containing R2 devices that are part of the SRDF/Star configuration. The async target array is paired with the workload array through an asynchronous RDF link. The sync target hosts and array must be a sufficient distance away to survive a regional disaster across the workload and sync target sites.

A global cluster configuration

All three clusters are members of a VCS global cluster relationship. The sync target hosts and the async target hosts are paired with the workload hosts in a VCS global cluster relationship. In VCS parlance, each cluster has two remote clusters.

The cluster (preference) list in the cluster configuration would look like the following for each cluster:

  • Workload site cluster A

    ClusterList=[Cluster_A=0, Cluster_B=1, Cluster_C=2]

  • Sync target site cluster B

    ClusterList=[Cluster_A, Cluster_B, Cluster_C]

  • Async target site cluster C

    ClusterList=[Cluster_A, Cluster_B, Cluster_C]

Note:

In a global cluster environment, you must attach all hosts in a cluster to the same EMC Symmetrix array.

A replicated data cluster configuration

The workload hosts and sync hosts belong to one cluster. The async hosts belong to a second cluster. The workload hosts belong to one system zone of replicated data cluster, and sync hosts belong to the second system zone of the replicated data cluster. All workload hosts are connected to the same Symmetrix array. All sync site hosts are connected to the same Symmetrix array.

The cluster (preference) list in the cluster configuration would look like the following for each cluster:

  • Workload site cluster A

    ClusterList=[Cluster_A=0, Cluster_B=1]

    System Zones=[Workload_Host=0; Sync_Host1=1]

Network heartbeats

You must configure network heartbeats among the three data centers to monitor their health.

See About cluster heartbeats.