In 2013, enterprise IT teams in Asia-Pacific and elsewhere were told to optimise network infrastructure by cutting data centre cost and complexity rather than adding capacity. Five-tip guidance from that period started from a familiar problem. Data volumes were growing, budgets were tight, and chief information officers were expected to show a return while limiting capital spending.
The advice below follows the points made in vendor-written guidance published by Network World, which the publication said it had edited to remove product promotion. The original five-tip list is not reproduced here.
Why were enterprises told to use lower-tier data centres?
The first point was cost. Network World’s primer described lower-tier data centres as one of the most effective ways to cut spending, because each step up in tier adds to the bill.
Many organisations running high-availability applications kept redundant high-tier facilities. The primer argued that only some workloads justify that, and it named certain government operations and financial services as the exceptions. Non-critical applications and archival storage could sit in cheaper sites.
It added a caution: recovery-time requirements should be checked before any move. It said recovery time is independent of a facility’s availability or uptime. That gives some flexibility, but the business still has to define what it can tolerate.
What did consolidation promise?
The second tip was to consolidate internal data centre operations. Network World’s primer said that shrinking the footprint could cut costs significantly and earn “green” credits. It put the savings from such a transformation at between 20% and 60%. That figure was the vendor-written primer’s own claim, and no study was cited for it.
Other guidance published in this period said consolidation also reduced the carbon footprint. It warned that the transformation takes substantial time and money before the savings arrive.
How did virtualisation and cloud fit in?
Virtualisation and cloud computing were presented as ways to use available resources and applications better. Guidance of the period said the rapid adoption of cloud computing was reducing the need for physical infrastructure, server hardware and support.
It also pointed to time saved on routine provisioning. Teams no longer had to hunt for rack space, allocate power and cooling, or locate network addresses for each new workload.
What were the alternatives to building more capacity?
Another tip was to reduce the data centre footprint. Network World’s primer noted that information growth creates a constant demand for more space. It said access to capital would stay difficult for some time, so firms needed to plan alternatives to building or leasing an extra data centre.
Those alternatives ranged from shared facilities to colocation and cloud. The primer also said some options displace capital expenditure.
What was the common thread?
The tips shared a few ideas:
- Match the resilience of a facility to the criticality of the workload.
- Reduce the number and size of sites that have to be run.
- Use virtualised and cloud resources to avoid buying and housing more hardware.
- Treat recovery-time requirements, not just uptime ratings, as a design input.
The guidance was written by vendors, and Network World itself said such primers tend to favour the submitter’s approach. Teams were therefore expected to test claims such as the 20% to 60% savings range against their own costs before committing to a consolidation programme.