Cloud Migration Questions

Подробнее о видео

The End Game for Cloud Becomes Clear
A Bit of Security for August 7, 2024
Is it cheaper to own a home or to rent?
Cloud computing is approaching the end of its exponential growth phase, as the curve’s slope drops. Every phenomenon that involves growth is subject to Stein’s Law: “If something cannot go on forever, it will stop.” (1986) Cloud adoption is slowing as could vendors increase their rates, undercutting the trade-off between the capital expense of acquiring and running one’s own computers and the operating expense of renting compute power from others. Accelerating this trend is the continuing march of Moore’s Law – the cost of computing continues to drop, lowering the CapEx side.
Cloud vendors today are looking for “growth drivers” – as IBM did in the 1980s – applications that will drive utilization and increase demand. Currently the biggest cloud growth driver is AI, which is now delivering a litany of failed efforts and underwhelming results. Further eroding the appeal of cloud is the lack of what vendors call “stickiness” – feature exclusive to one cloud vendor that impede migration from one cloud platform to another. (Cloud consumers call this “vendor lock-in.”)
The shape of the end game is now becoming clear. Current revenue projections assumed accelerated demand due to AI, which is not materializing. As that demand continues to ebb, cloud vendors will begin cutting prices to drive remaining marginal migrations. This red ocean will be extremely difficult to escape.
Cloud customers and those considering migrating additional workloads to cloud should carefully review their financial assumptions.
One assumption is the cost of running on-prem equipment.
Another is the relative cost of migration.
A third is the variability of that workload.
Highly variable workloads can find some value in cloud even now; but those environments that run a fairly steady level of utilization will only incur the extra costs of off-premise workloads: network costs, remote debugging complexity, failure to understand the actual utilization patterns which might allow for more targeted buying and configuration management, and skills retention. If your goal is to get to cloud regardless, consider building your own on-premise cloud and migrate there, using canonical APIs. You’ll have the best of both worlds – you bake in the lower cost of newer compute resources, you’ll develop your infrastructure and operations skill base to better tune and grow your installation following the strategic architectural direction and pace of your organization’s growth, and you’ll eliminate the security and privacy risks of third-party data residency.
A Bit of Security for August 7, 2024
Cloud migration is slowing as the demand from AI is failing to materialize, as vendors raise prices, and as Moore’s Law inexorably drives down the TCO of on-premise computing. Listen to this -
Let me know what you think in the comments below or at wjmalik@noc.social
#cybersecuritytips #cloudcost #cloudTCO #migation #BitofSec