From stability to agility Data centres have been around a long time during which various operational practices and best practice processes have been established to keep them running effectively, if not always as efficiently as possible. Data centres often rely on IT professionals specialised in a variety of skill sets; some are required to ensure high value systems and applications function smoothly and without interruption, while others are engaged in more routine operations. Until recently data centres had become islands of relative stability, frequently relying entirely on in-house staff. However, times are changing. In recent years, business demands have forced data centres to adjust, especially in terms of the speed at which they could react to rapidly altering requirements. This has been coupled with increasing demands to optimise operational costs often with the direct consequence of improving the utilisation rates of core compute, storage and networking components. Flexibility has thus become the new ‘Data Centre Black’, and this has impacted on the IT infrastructure used to deliver business services as well as the skills needed by IT staff. The promise and practicality of external services One manifestation of the drive to nurture data centre flexibility can be seen in the range of external and managed systems now routinely used alongside internal data centre systems to deliver IT services, both business critical and lower profile applications. Typical offerings used to supplement internal data centre operations may include private hosted services provided off-site, the use of public cloud based systems (e.g. IaaS, PaaS, SaaS) or having systems running inside the data centre but managed by a service provider. The addition of public cloud, private hosted cloud and other managed services into the mix of data centre operations can add great flexibility to previously fairly staid delivery capabilities. But adding such cloud and managed services into the mix creates new challenges. New context, but familiar challenges The nature of many of the challenges thrown up are not, fundamentally, new in nature. As might be expected, issues relating to the ‘fragmentation’ of services between those running locally in the data centre, systems managed on site via a managed service, and any solutions operating off-site in public or hosted private clouds feature often. The challenges include routine operational management issues such as looking after identity management between systems operated by different management teams and potentially running in widely dispersed locations. While such ‘integration’ issues are not new, identifying good practice in how to deal with them in the context of cloud, hosted and managed services can take considerable research. Similar challenges can arise when looking at implementing operational monitoring for services that run outside the data centre. The scale of the issues involved can escalate rapidly as the number of managed services, hosted and cloud systems expand, especially if these are provided by multiple suppliers. This is especially important for services with components that function across the various boundaries. The same applies to the very broad domain of how security and data protection are implemented across potentially very fragmented systems, especially when it can be difficult, sometimes impossible, to modify key aspects of systems run by such third parties. Maintaining those all-important service levels Finally, there is the over-riding matter of how to handle the in-house management of these third parties, especially in terms of keeping an eye on Service Level Agreements (SLAs) and the inevitable requirement for modifications and changes to be made to services. The issues thrown up usually require considerable work to establish timely, efficient and effective processes to keep third party provided systems working with changing business requirements. This is especially the case as internal users inevitably regard the in-house IT support teams responsible for all facets of IT service delivery, even when the service is provided or managed by a third party. After all, if anything goes wrong, IT will get the first support call. Which highlights the issue of support and how to integrate third party resources into established process and procedures. Effective and efficient IT service delivery requires that the support structure put in place for managed services, hosted and cloud solutions, usually needs the third party to have skilled staff working in support who really do ‘understand’ their customers, especially if these systems are core components in business operations with which users interact frequently. For some systems, this may even require the supplier to have local presence, something that may be supplied directly by the third party or by a local partner with tight integration into the support organisation. Ensuring these capabilities are in place is something that data centre professionals regard as ‘business as usual’, but can be problematic if IT is ‘handed’ a new solution that has been put together by the business users directly, something that can occur in organisations of any size. The bottom line Third party services are growing in importance in many data centres, but they do impose operational challenges that must be managed as part of standard processes if risk exposure is to be minimised, especially as the number of managed services, hosted and cloud systems used expands. Hoping things work as the contract describes is rarely a viable plan. Even more importantly, they may well require tangible modifications to long established operating procedures, and getting buy in from IT staff impacted can take significant effort.
Tony is an IT operations guru. As an ex-IT manager with an insatiable thirst for knowledge, his extensive vendor briefing agenda makes him one of the most well informed analysts in the industry, particularly on the diversity of solutions and approaches available to tackle key operational requirements. If you are a vendor talking about a new offering, be very careful about describing it to Tony as ‘unique’, because if it isn’t, he’ll probably know.