Delivering high availability and disaster protection for business critical applications in cloud and hybrid cloud environments
Written by Jerry MelnickOctober 15, 2014
Enterprise data centers are moving more and more applications to innovative cloud and hybrid cloud configurations to save money, speed time to deployment, and meet demanding RTOs and RPOs. Despite these benefits however, many companies are reluctant to move their business-critical applications to these environments because of their concerns about high availability and disaster protection. Unlike physical environments where providing HA and DR protection is often a straightforward matter of creating a traditional shared storage (SAN-based) cluster, your choices may not be as obvious in cloud or hybrid cloud environments. There is also confusion about what levels of application protection are currently offered in the public cloud; how much added protection is needed in the cloud for business-critical applications; and how best to deliver HA and DR protection without adding prohibitive cost or complexity.
In a cloud, you need to protect important applications from downtime in the event that the cloud instance or the public cloud provider fails. Most cloud providers offer separate and redundant data centers or computing resources. These resources are simply redundant; they do not include any type of clustering or application failover capabilities. More importantly, they do not offer shared storage (i.e., a SAN), which is required to support traditional clustering solutions.
HA and DR in the Cloud
The easiest way to move business-critical applications to a cloud without sacrificing HA or DR protection is to use a SANless cluster. SANless clusters use efficient block level replication to keep storage on the primary and standby nodes of a cluster synchronized. Instead of sharing the same storage, in a SANless cluster, the cluster nodes each have access to identical copies of the storage.
This process has a variety of benefits. It eliminates the single point of failure risk posed by shared storage and it enables you to use failover clustering in environments, like the cloud, where shared storage is not offered. It also eliminates the configuration limitations of SAN storage.
SANless clusters can be built within a cloud or used in combination with traditional SAN-based clusters to provide HA in a wide range of configurations that mix physical, virtual, public and private cloud. They can also be used with local high performance SSD storage for fast application response times without the cost of a SAN alternative. You get the failover protection of a traditional cluster and the configuration flexibility benefits of cloud computing.
There are several SANless clustering options to choose from. One of the most common SANless deployments is in Windows environments in conjunction with Windows Server Failover Clustering. You can build a two-node Windows Server Failover Clustering environment in a public or private cloud and simply add SANless cluster software as an ingredient to enable operations of the cluster in the cloud. You can configure nodes in different cloud availability zones or domains for DR protection. The SANless software synchronizes storage in the primary and standby nodes, making it appear to the WSFC as identical to a SAN. These clusters can be setup quickly and managed as a traditional WSFC without changes to your processes.
In Linux environments or Windows environments where WSFC is not a requirement, SANless clustering software can also be used to monitor the entire application environment and to manage the failover process as well as to synchronize storage. This software includes configuration wizards that eliminate the complexity of typical Linux clustering.
In some application environments this SANless configuration can save thousands of dollars in licensing fees by providing HA and DR for applications. You can also use SANless software as a highly efficient replication-only solution for DR.
'Cloud First' Data Centers
SANless clusters are playing an essential role for companies that are moving their entire data center to the cloud in so-called “cloud first” infrastructures. These companies have two challenges: to meet their project deadlines for moving applications from existing on-premises environments to the cloud; and ensuring the application environment in the cloud is protected from downtime and data loss.
They enable companies to create a cluster and move critical application environments to the cloud quickly and easily. They can also add a failover node to an existing cloud-based application environment for HA and DR as they need it
Protecting Hybrid Clouds
Companies are using new hybrid cloud configurations that let you add computing resources and storage from the public cloud during peak demand and to reduce resource allocation during low demand. You can configure a SANless cluster with a node in an on-premises private cloud with failover (and failback) to a node in a public cloud.
Extending Clusters to the Cloud for DR
For many years, Windows Server Failover Clustering (WSFC) has been used to provide HA and DR protection for SQL applications in traditional on-site environments. Until recently, managing site failures has been very complex and expensive, requiring large investments in specialized hardware and software as well as the availability of a second data center site.
The cloud offers an extremely attractive and cost-effective second site to locate a cluster member and use it to handle failover when the local site fails, SAN-based clusters require a shared SAN between all cluster members precluding their use when building a cluster that includes both on-premise members and members located in the cloud. However, SANless clusters you are completely free to build a fully operational WSFC cluster combining on-premise cluster members with cluster members located in the cloud. With a SANless cluster, you gain scalable disaster recovery protection without the cost or complexity of building out or renting an off-site disaster recovery facility.
The use of innovative cloud configurations will continue to grow as IT departments look for new ways to save money, improve service levels, and adapt quickly to changing IT requirements. SANless clusters are playing an essential role by providing the availability protection and flexibility needed for business critical applications in these environments.
About the Author:
Jerry Melnick (jmelnick@us.sios.com) is responsible for defining corporate strategy and operations at SIOS Technology Corp. (www.us.sios.com), maker of SIOS SAN and #SANLess cluster software (www.clustersyourway.com). He has more than 25 years of experience in the enterprise and high availability software industries. He holds a bachelor of science degree from Beloit College with graduate work in computer engineering and computer science at Boston University.
