Data protection has long been thought of as somewhat of a necessary evil like life insurance. Throw money at the possibility that something will go wrong and pray that it never does. However, with cloud computing models and the concept of IT-as-a-service, data protection can now contribute directly to the value of the data center as well as to the bottom line.
Most organizations use numerous technologies to protect application data across primary, replica, and backup storage, as well as archive systems. Trying to manage a mixed data storage environment without an effective data protection management strategy makes applying a uniform chargeback to these systems almost impossible.
As IT transitions to a fully operational private cloud, effective ways to measure and meter data protection services across the environment need to be put into place. Data center managers need mechanisms to differentiate between service levels offered to the organization, and to justify and apply chargeback to business units.
Leaving Money on the Table
Earlier this year we looked at how data protection management should provide end-to-end visibility and intelligence across data storage, so you could know your virtual applications are protected. This post will examine how the right data protection management strategy can also get you to the money.
In the mixed infrastructure characteristic of today’s data center, data integrity is ensured by numerous different data protection element managers. This approach makes it difficult to apply a uniform set of rules and rates to the data for chargeback. The result is that you have lots of disparate information about different components, making it a challenge to measure the utilization and adherence to service levels across your infrastructure. Many organizations faced with this challenge are forced to either apply a one-size-fits-all approach to chargeback or not apply chargeback at all.
Without a means to manage chargeback levels, the business value of IT cannot be measured. Worse yet, you are effectively leaving money on the table.
Challenges to getting to data protection as a service include:
- Rapid data growth: Data growth is overtaxing traditional data protection approaches and technologies, leading to increased uncertainty about protection status. The inability to measure replication, backup, and archive activity per user means there is no way to gauge use or predict growth. As a result, infrastructure is either over- or under-utilized. Also, without uniform metrics, data center managers cannot determine and justify where investments must be made. Talk about missed opportunity!
- Heterogeneous systems: As mentioned earlier, in almost any kind of data center deployment, multiple protection methodologies are commonly applied to different technologies often from different vendors and to provide different service levels. Local backups or replication may be used for operational recovery, off-site replication used for disaster recovery, and archiving used for longer-term data retention. These islands of data protection make it difficult to apply a common method to charge for services rendered.
- Virtualization: The inevitable shift from dedicated physical to shared virtual server infrastructures in many data centers creates challenges because in a virtual world virtual machines (VM) are quickly spun up, then down, and moved around. The processes and technologies for chargeback that worked well in a physical environment may not be the most effective in a virtual environment with this rapidly changing VM topology. The result is uncertainty for applying uniform costs to protecting and recovering business critical data, as well as the added cost and complexity of sustaining multiple data protection systems.
Successful implementations of chargeback and show-back (i.e. the ability to measure and show use even in the absence of a cross-charge system) enable IT to apply a uniform set of metrics across all data protection components, including replicated, backup, and archive storage. Service plans then allow administrators to set rates and use the metrics as a way to apply chargeback and show-back for services provided.
Key data protection management capabilities needed for effective metering and chargeback include:
- Central management at scale: To get to a metering model at scale, the data protection management system needs to unify the chargeback policies applied to the usage data received across the entire replication, backup, and archive environment. Automated data collection needs to feed a central analytics engine and provide for setting different policies for different infrastructure. Regardless of the cloud model, the data protection management solution needs to scale out to support internal and external multi-tenant environments with more and more customers, including those customers with complex billing requirements (chargeback), while providing tailored views into data protection (show-back).
- Multiple rate support: The system must apply a real-time common cost methodology with the ability to effectively assign rates to the different services and use these rates for chargeback and show-back of services delivered to different classes of users. Administrators need control over establishing chargeback cost monitoring, with the ability to create separate fees for cost per replication, backup, and archive activity (single usage fee per activity), including cost per capacity retained (fee per capacity until the data expires) and cost per capacity restored (fee per capacity until the data gets restored).
- Historical views: Historical chargeback views are needed to provide insight into trends and analytics that would otherwise be difficult to measure using separate systems. Standard chargeback reports should automate data collection and analysis for administrators who would have to otherwise resort to pulling data together from different systems into spreadsheets—a time-consuming process prone to human error. Historical views and trending also provide insight into planning for expansion, including new or additional replicated, backup, and archive storage capacity.
Successful chargeback and show-back implementations in data protection management solutions such as EMC Data Protection Advisor enable administrators to apply a uniform set of metrics across all data protection components (e.g. replication, backup, archive), build service plans that allow administrators to apply rates to each of the service plans, and then use these metrics as a means to apply chargeback and show-back for services rendered. Imagine the value of having a single screen where administrators can view all components and all policies in a heterogeneous environment, and then give users self-service access at the flick of a switch to data protection as classes of services. Visualize having the ability to carry out chargeback on any number of metrics and to produce the output in a format suitable for billing systems. Then the data center becomes a profit center. Imagine that!