Clustered Data ONTAP is fantastic technology and helps our customers design their multi-site data center infrastructure in a way that maximizes IT skills and controls cost. With clustered Data ONTAP, you can setup your primary and secondary data centers to optimize your performance, capacity, disaster recovery and backup requirements.
For this post, I want to point out a few ways in which clustered Data ONTAP can optimize your multi-site environment to address performance, cost and management needs. Consider the diagram below. What I’m showing is two data centers, a primary and a secondary data center. This is not an atypical configuration for our customers.
The primary data center is optimized for performance. It is supported by a 6-node cluster of high performing FAS6200 storage systems with high performance SAS or SSD drives or a hybrid configuration. The secondary data center is optimized for capacity and cost. It is supported with a two-node configuration using a FAS3200 storage system with high capacity SATA disks. The secondary data center has scaled up to achieve the capacity requirements, while the primary data center has scaled out to achieve greater I/O performance. To take this scenario further, in order to support SAN workloads, the primary data center can be deployed with FCoE or traditional Fibre Channel while the secondary data center may simply be piped with Gigabit or 10 Gigabit Ethernet supporting iSCSI. In either site both SAN and NAS protocols can be supported on the same infrastructure.
Backup and disaster recovery are supported with SnapVault and SnapMirror features, respectively, which are configured by volume, allowing you to select all or a subset of volumes to protect at the secondary site. This added flexibility allows you to further address cost pressures of maintaining a secondary data center. And you can leverage NetApp’s FlexClone feature to space efficiently create clones of the mirrored volumes which can be mounted to validate that the mirror is successful, or to test Disaster Recovery failover without breaking the mirror – so you can test during the day on a weekday instead of 2AM on a weekend. This is a simple way to validate your DR environment without a painful failover event.
The architecture above represents only one possible configuration. SnapMirror and SnapVault can be configured in a myriad number of ways including local site SnapVault for backup, then SnapMirror to the remote site; SnapMirror to remote site and then SnapVault from the mirror. NetApp also supports fan in, fan out and daisy chaining. One unique capability is if SnapVault is used in the primary data center to backup from one node to another in the same cluster, you can change source and destination volumes on the fly – even during replication!
Your secondary site doesn’t have to be a cost burden simply waiting around for a disaster. Many of our customers, BBM Canada is an example, actually use secondary sites for development work. Since the ongoing mirroring and backup of data is a light weight workload and requires little bandwidth, you can create clones of the mirrored production data and use these copies for test and development of new applications without impacting production systems. Along the same lines, instead of idle stand by servers, you have active servers for dev/test, and then in the event of a real disaster you can convert them to handle the production load.
Clustered Data ONTAP is extremely powerful and designed to accelerate your business. Do you see opportunities in your data center where clustered ONTAP can increase utilization of your valuable IT assets?