It is particularly in the cloud market that managed service providers (MSPs) are increasingly relied upon to bridge the gap between IT vendors and enterprises as enterprises grapple with data that is growing in both volume and complexity, says CommVault Systems, which works with more than 20% of 200 MSP partners in the Asia Pacific region.
 |
| Source: CommVault. Bentkower. |
Mark Bentkower, Director of Enterprise Solutions APAC, CommVault Systems, points out that with policies that focus on standardisation, cataloguing, automation and budgets, MSPs can enable enterprises to achieve increases in efficiency, reductions in management time and significant cuts in costs. He has identified three MSP processes that can be applied to internal data management strategies:
Internal service level cataloguing
"Most companies have little knowledge of the data they generate and store. As a result, to err on the side of caution, many companies retain unnecessary data, which means redundant, long-term storage costs to the organisation," he said. "As a matter of standard practice, most MSPs will look to categorise and catalogue data based on what data should be retained, and how long it should be retained before it is deleted. MSPs categorise data in a hierarchy system, thus ensuring that data is tiered and stored at the most cost-efficient and effective platform for pre-established lengths of time."
MSPs, he said, typically catalogue datasets into different categories, ‘Gold, Silver, Bronze’ for example. ‘Bronze’ processes are applied to less critical data sets, and ‘Gold’ processes, for more sensitive data. "This cataloging must take into account both the costs of actually storing the data, and also the accessibility of the data. Enterprises need to assess how frequently various types of stored data will need to be accessed," he said.
"A robust cataloguing and archiving system should consist of multiple variables, such as file type, size, ownership, and last accessed and creation dates, among others. Additionally, good MSPs recognise that the value of data is subject to change, for instance, the ‘gold’ category of data today may drop in ‘value’ depending on factors, such as creation date."
Regular cost and data analysis
The cost of storing increasingly large data volumes is becoming significant, so much so that a recent IDC report reveals that 87% of companies are keen to reduce storage costs*, Bentkower says. "Even at the proof of concept stage, many MSPs promise to deliver cost reduction. This motivation to reduce costs and MSPs’ typically eagle-eyed focus on budgets enable them to deliver massive cost savings. It is common, for example, for CommVault’s MSP partners to be able to help companies slash costs by as much as half, thus, doubling staff productivity and accelerating time-to-revenue in the process," he shared.
An IT department could encourage up-to-date data analyses by billing divisions within an enterprise for storage, for example. "This encourages better visibility in terms of cost and utilisation of storage space by each business division or function while ensuring that departments are made accountable for their consumption of data services," Bentkower explained.
"Expensive disk space might have once been warranted for data which need to be stored for compliance purposes. However, it might not justifiable at the same tier following expiration of the legally dictated timeframe. Reducing data management costs with a more efficient data infrastructure will free up budgets – enabling companies to pursue opportunities that can power business growth**," he said.
Automation is key
When making storage decisions around corporate data, organisations require a degree of foresight, said Bentkower. "An IDC survey reveals that Singapore-based enterprises are actually twice as likely as the average Asia Pacific organisation to anticipate year-on-year data growth of more than 51%*. This leads even more enterprises to leverage on public cloud for ‘pay as you grow’ models, preventing budgets from being locked up in under-utilised hardware."
Such public cloud platforms have to be scalable and flexible. "In the retail industry for example, data levels fluctuate. We have seen that there are certain times in a month or year, where retailers’ data levels rise significantly, only to fall drastically, where they could remain until the next ‘peak’ period. In these cases, if an enterprise is to run an efficient data management system – the cloud model that the enterprise opts for must offer some degree of flexibility, as well as scalability," Bentkower noted.
Data trends, particularly BYOD, should be supported too. "Companies must consider the additional processes and resources required to support endpoint data back-up, and control and access from remote devices, such as smart phones. The complexity of data environments means that it makes more sense to automate data management processes, based on the pre-determined policies set through the service level cataloguing," Bentkower said.
"When data policies are reliably automated, this not only ensures that the policy is seamlessly adhered to, it also reduces management costs by simplifying data management operations. Automating the archiving of data, both structured and unstructured, according to internal cataloguing processes, results in massive cost reduction and improved IT department efficiency."
Companies put themselves at a disadvantage when they neglect to run their own IT departments by the same high standards that they expect from a third-party service provider, Bentkower said. "Generally, there is not enough standardisation in IT departments - particularly, in large enterprises. Unless an organisation has full visibility of its data, it is impossible to make informed and effective data-related decisions on cataloguing, cost analysis and automation. By learning from MSPs’ engagement and delivery model, enterprises will be better equipped to ensure that their data is stored both effectively and efficiently."
*IDC/CommVault (2013) ‘Smart data in the big data era’ survey, Jimenez
**IDC (2013) APEJ C-Suite Barometer Study
