In my previous blog I outlined the four fundamentals every business needs to consider when addressing cloud connectivity – bandwidth, latency, availability and cost control – as well as the three maturity phases for data centres. To recap, these are:
- Phase 1 – construction and management of internally owned facilities
- Phase 2 – connection to co-located / outsourced data centres
- Phase 3 – connection to cloud providers with their own data centres
Regardless of where the business is at on the spectrum, every organisation needs a specific level of connectivity, especially at the application level. Every application has its own requirements and while this can overwhelm some due to the complexity involved, it is always a more effective strategy than considering connectivity from a broad business perspective.
Conducting A Zero Assessment
This is why it is always advisable to conduct a Zero Assessment. It will help highlight how aligned connectivity is with the business strategy and its specific processes/technology. Once application differences have been established, the business can begin to make an informed decision on a connectivity choice that not only meets current requirements, but one that will scale in the future.
At this stage also consider what outside pressures might influence connectivity. This could mean regulator considerations, compliance requirements or customer expectations. Regulation is particularly important for sectors like finance, trading, insurance, law and other security heavy verticals. The connectivity must accommodate these sector-specific demands.
The second criteria driving a Zero Assessment is application development/application rollout at multiple sites. To assess this, begin by determining the associated risks and costs for each application category. For example, storage replication needs a different connectivity choice to simple file sharing. Is Microsoft Office availability more important than business platforms like CRM, ERP and process control?
The average latency for transatlantic connections is 60 – 80ms. This is fine for general functionality, but not suitable for the synchronous replication of data. If we assess time-sensitive sectors like trading, banking and the media, even faster connections are necessary otherwise there can be a huge impact to potential profitability and competitive advantage.
The Zero Assessment Fundamentals
Next assess each application category and align it with a potential connectivity choice. A full zero assessment should include:
- Overall business requirements and any possible regulatory considerations
- Core back-office infrastructure and applications (i.e. Office)
- Business control applications (e.g. ERP)
- Unified Communications and collaborative technologies
- Interconnects and links between data, servers, storage and each corporate network
- Data volume per unit of time (actual/anticipated)
- Required bandwidth (dimensioned/peak load)
Once those have been outlined, the minimum/maximum bandwidth requirements for each category should become apparent. It should also highlight the potential effect of latency from a user perspective and an IT management point of view. Following that, it will be possible to assess which maturity stage the business is at, which is important for overcoming the challenges (and benefits) of maximum availability.
In my next blog I will look at the ‘five nines’ environment (99.999% availability) and the relationship between connectivity providers and data centres.