NIGERIA: Why We Need Data Centers (II)
This is a sequel to the first piece on why we need standard and globally acceptable data centers in Nigeria.
As shared previously, there are four(4) Tier-III certified data centers in Nigeria. If one assumes that each of the four(4) Tier-III data centers in Nigeria have a capacity of 255 Racks, we estimate that they have less than 30% occupancy currently. If the other four major data centers that are currently not certified are added, the rack capacity would increase to over 2000 racks currently available in the country. This would imply that there is currently an estimate of less than 15% data center occupancy. Half of this capacity is owned by government agencies and telcos, and as such, do not have as much pressure from shareholders as the commercial data centers for the ROI to be realized from these investments current capacity. If taken up, it would save many businesses a great deal of expense and contribute to developing the co-location market for better services and availability.
The following benefits would be derived from outsourcing or co-locating their infrastructure: Co-locating with a Tier-III certified data center will achieve better stability and service delivery than is attained from their present situation. Moving from an out-of-the-country data center to a local data center will achieve the following:
- Better latency
- Better and closer support
- Stability in cost – No foreign exchange component.
- Cheaper TCO – Savings will be achieved from cutting out Cap Ex expenses on; Power – generators, inverters and UPS-es, as well as batteries and other component replacements; Cooling – specialized precision cooling systems are more expensive than traditional cooling systems and they have a very high maintenance cost due to their specialized nature.
- Better connectivity – This is especially relevant in the disaster recovery settings. A multi-branch organisation would benefit from a data center location as a fail-over to an alternative location. This is a simpler topology and is cheaper.
- Taking advantage of 24/7 NOC (Network Operations) engineers – This effectively outsources the environmental and resource monitoring. This also contributes to lower staff cost and takes advantage of specialized data center staff.
- Managed services – A more granular support than just the NOC staff. Faster reaction to system failures and restarts. Again, staff costs are lowered and staff with managed services skills do not have to be in-house. Overall the shared services model adds to a drastic lowering of costs as services are delivered at a shared cost as opposed to an individual organisations cost.
Over the last seven years, there have been talks in the press of Tier-IV data centers to be built locally. In an environment where there is not a single Tier-III operational data center – it is an ambitious target. Currently, there are thirty-eight(38) Tier-IV constructed data centers and only about nine(9) Tier-IV operational data centers worldwide.
We are still trying to understand the benefits to be achieved with building a Tier-IV data centre locally, perhaps this is the reason why it has not seen light of day so far. We will achieve much more by encouraging the other data centers to achieve Tier-III operational status to give the community better stability and reliability for their existing infrastructure. This is evident by the number of disruptions we currently see in vital organisations like banks and telcos. This is also supported by the fact that the only Tier-III data centre certified by construction has not had a single minute of downtime since inception three years ago. Currently in Nigeria we have four data centers that are Tier-III certified by design and only one of them is by design and construction.
There is an interesting discussion raging at the moment about the Uptime Institute’s 3-levels of Tier certification – Design, Construction and Operations.
It is deemed unfair to offer a design certification as the data center is not bound to construct to the design specification. In some quarters, this is misleading as it really does not translate to the stability or reliability of the data center.
In other circles, the Uptime Institute’s Tier standards is seen as confusing and complex. Our view is that any adherence to the Uptime Institute standards is a good place to start in ensuring that the data center is able to host business critical applications and offers a greatly reduced risk. Outsourcing infrastructure and hosting will save organisations 40% – 70% of their total IT spend.
The reduction in Total Cost of Ownership (TCO) is achieved by:
- Conversion of CAPEX to OPEX.
- Reduction of overhead costs
- Lowering operational costs.
- Elimination of hardware and replacement costs.
- Freeing up real estate for other uses.
It is important for organisations to seriously consider the outsource model. They need to stop the CAPEX model which is not sustainable, and start to drive the OPEX strategy. It will provide more agility and achieve more in a climate of reduced revenue and increased competition.
This article has been concluded.
Remi Adejumo is the Chief Technology Officer of CloudFlex.
Cloudflex is an IaaS company that provides private, public