Home > Redbook, SVC, Uncategorized > Stretching SVC to meet new business availability needs.

Stretching SVC to meet new business availability needs.

IBM has been secretly working on a technology that is changing the way administrators will deploy their VMware environments.  SAN Volume Controller (SVC) Split Cluster solution allows live application mobility across data centers based on VMware Metro vMotion. This solution can now alleviate unplanned outages as well as provide additional flexibility.

SVC can now support a split cluster configuration where nodes can be separated by a distance up to 300 KM.  This type of configuration does require clustering software at the application and server layer in order to failover to a server at the corresponding site in order to continue business and resume access to the disks. SVC will keep both copies of the storage in sync and mirror the cache between both nodes so the loss of one location does not disrupt data access at the other site.

The next big advantage SVC has is a small quorum disk that can allievate any split-brain issues.  Split-brain problems occur when nodes are no longer abale to communicate with each other and they start allowing writes to their own data.  SVC creates a tie break from a third site in order to ensure the survival of at least one location.

The SVC Split Cluster configuration uses a term called failure domain.  This allows the SVC cluster to know which components are in a certain boundary where any failure may occur (ie power, fire, flood). The entire SVC Split Cluster must comprise of three such boundaries, two for the storage controllers containing the customer data and a third for the active quorum disk.

Using VMware’s VMFS file system, the SVC can supply access to all of the vSphere hosts in both locations. During vMotion, the virtual machine can switch to a new physical host while keeping the network identity. By using the Metro vMotion to migrate VMs, customers can now plan for disaster avoidance, load balancing as well as data center optimization of power and cooling.

This solution is also for customers looking to add high availability with in their single datacenter. Imagine two servers in different parts of the data center. They could be on different power feeds and different SAN fabrics. You can now provide continuous availability to these servers with out having an administrator called in the middle of the night.

For more information about SVC Split Cluster and even step by step instructions on how to setup the SVC, VMware and Layer2 network check out this Redbook.

  1. No comments yet.
  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: