For some explanation I’m on the listing of each PR company with a consumer that wants to have its predictions for the approaching calendar year heard. Generally their guesses are noticeable and unhelpful, these as “cloud computing will carry on to grow” or “cloud protection will carry on to be a worry.” People today are genuinely going out on a limb.
On the other hand, I see a few predictions in excess of and in excess of once again. I suspect that lots of out there in the cloud computing marketplace are beginning to imagine them, so I need to have to position out some realities that are often missed or misunderstood.
The increase of multicloud to grow to be “cloud agnostic” is the most common—and most concerning—prediction. I in particular get worried if the perceived benefit is that becoming cloud agnostic refers to environments that are able of working with any community cloud supplier with small disruptions to a organization.
The idea appears to be compelling, contemplating that you can leverage finest-of-breed community cloud computing solutions, these as picking out storage from two or extra community cloud suppliers. However, the reality is that purposes and data have to be localized to a specific cloud supplier to be of true worth to the organization. This signifies that you are going to have to lock in to a cloud supplier to choose true advantage of that cloud provider’s capabilities.
So if you leverage multicloud, and at the identical time attempt to be cloud agnostic, you are going to have to choose the least typical denominator approach to setting up purposes and connecting data storage. This signifies that you won’t be optimized for any community cloud platform, or that you won’t run perfectly wherever.
I’ll also point out that you are missing crucial indigenous functions these as protection, governance, management and monitoring, etcetera. If you do use these functions, that software and data won’t be cloud agnostic considering the fact that they’re no extended portable between community cloud platforms because of the coupling of cloud solutions optimized for those people platforms. No great cloud architect will choose this approach given the destructive trade-offs. I push again on them weekly.
Of study course, rising technologies these as containers and container orchestration engines do deliver cross-cloud portability. The compromise right here is that you are going to have to port your purposes and data to containers, which provides cost and chance. But it is an alternative.
A second prediction considerations the setting up of “cloud-native” purposes. Merely set this signifies that you have refactored or objective-developed purposes and data storage to choose advantage of indigenous or proprietary functions of specific community cloud platforms. The purposes have been coded to choose advantage of the indigenous solutions provided by a one community cloud supplier, these as indigenous protection, indigenous management, or indigenous databases.
Cloud indigenous is a great matter in that it permits developer and architects to enhance purposes and data storage working with solutions for the cloud supplier they run on. You are setting up and deploying purposes that are almost one hundred p.c optimized, and consequently deliver greater effectiveness, protection, and dependability.
The value is community cloud supplier lock-in. Port the purposes and data to yet another community cloud supplier and you are going to need to have to do some major surgical procedures to make them cloud indigenous to the new supplier, these as AWS or Google.
I really don’t necessarily mean to be the selected buzzkill right here. It’s just my reaction to the sum of misinformation out there right now and the need to have to push again on some of it. I would somewhat you move to multicloud or cloud indigenous with your eyes open, ready to make some hard alternatives.
Copyright © 2021 IDG Communications, Inc.