To Cloud or not to - No longer a question!

There is no doubt that most are going to the cloud

However, the cloud definition and understanding is different now. So, the cloud is no longer a location. Cloud is an operating model

Do you remember the conversations on the subject years ago?

Public cloud providers took a bold stand to say that everything is going to the cloud? This was when the cloud meant a "location" somewhere away from your sites and installations. Furthermore, this meant no servers or compute for your IT management team to touch and physically manage. Cloud was an abstraction synonymous with "away from on-premise." Cloudy

Well, hyper-scale Cloud Providers such as Google, Amazon, and Azure are all providing on-premise solutions such as the following:

Google with Anthos

Amazon with Outpost

Azure with Azure Stack and Hub

The above examples prove that the very well-known cloud providers have started to respond to our reality well known to most of us since the beginning of the cloud. The reality has always been. Not all workloads belong away from on-premise and into a cloudy location. However, all workloads do deserve some of the benefits of a Cloud Operating model

While moving away from on-premise was a trend, there is the realization that it costs too much to run and some unforeseen overhead. Examples are: 

-Public cloud is a shared infrastructure publicly!

-High availability does not come from hardware any more

-To add higher availability, there is a need to re-architect applications

-Lift and shifts are possible, moving Virtual Machines to Public Cloud. However, this adds complexity to the overall environment with more locations to manage and new skills required

-Adding load balancers, adding replications, adding more Virtual machines will help add redundancy across cloud provider regions

The above is just a quick list of the impact I have seen companies experience with over 20 years of experience in the field

To add more, if you move an application workload to the cloud "location" remember, to always add redundancy to your application. Here is a quick link where issues/outages are published. Notice the number of services in a particular cloud provider: https://status.azure.com/en-us/status  look for similar sites for your cloud provider of choice, such as AWS https://status.aws.amazon.com/ 

This shows that these providers are ready to manage massive outages and have placed the mechanism on publishing services massively impacted. As clients, it is our responsibility to understand our new software architectures' impact to compensate for lower levels of availability

As a result, there is another trend, repatriation! These are companies choosing to move back to on-premise or build their private clouds

Look at further proof, documented by Slashdot.org Resource links found at the end of this blog: Quote: "Bank of America has bucked the Wall Street trend by building its own private cloud software rather than outsourcing to companies like Amazon, Microsoft, and Google. From a report:The investment, including a $350 million charge in 2017, hasn't been cheap, but it has had a striking payoff, CEO Brian Moynihan said during the company's third-quarter earnings call. He said the decision helped reduce the firm's servers to 70,000 from 200,000 and its data centers to 23 from 60, and it has resulted in $2 billion in annual infrastructure savings."

So, what are the key takeaways?

1. Look at the workload! Not all workload belongs "away"
2. Check into latency on applications and interdependencies
3. Ensure reliability is ensured by understanding what will need to change in the application before migrating away
4. Find partners that offer the most comprehensive set of services, software, hardware to give you a cloud operation model
5. Do not pay to learn, meaning. investing heavily in moving to cloud "the place" only to repatriate a year or two later (repatriate means bringing the applications back from the cloud "place" back to on-premise)
6. Only migrate what needs to burst. Here is a quick example to help you think about it. If you manage a large environment that is idle at 10 % for most of the year and hits 100% during a major holiday, you have a burst in usage. Another example: if you architect for a peak time that happens a few times a year rather than architect for daily average, you have a bursty workload

The "cloud" benefits come in the Operation model. If you can achieve a cloud-like operation model across your entire infrastructure, then you will see benefits in your business and operations. Here is a link where you can find information on the subject. https://www.delltechnologies.com/en-us/services/consulting-services/operating-model.htm  in my opinion, if you can achieve bursting a workload with public cloud, consume on-demand and pay for what you consume on-premise, have a single data protection solution across on-premises and public cloud assets and not loose availability on your applications, you are benefiting from the cloud experience

Would be great to get your feedback, send questions and comments

Have a great rest of the week; remember to register and join in Dell Technology world: https://www.delltechnologiesworld.com/index.htm?cid=dtwe20_paid_brand&dgc=st&lid=paidsearch&AffiliateData=dtwe20_paid_brand&gacd=9643275-1040-5761040-266682520-0&dgc=st&&gclid=Cj0KCQjwtsv7BRCmARIsANu-CQf1_M7ryQ1uwQPRV1wNrc64Lpj6DHiNVbRAPUT7TMbJAsmUanZvR_IaAmChEALw_wcB&gclsrc=aw.ds

Regards,

Julio Calderon - @juliocus #iwork4dell

Resources used:

https://slashdot.org/story/19/10/21/197242/bank-of-america-says-it-saves-2-billion-per-year-by-ignoring-amazon-and-microsoft-and-building-its-own-cloud-instead

https://www.datacenterknowledge.com/colocation/google-anthos-starting-make-inroads-colocation-data-centers

https://azure.microsoft.com/en-us/blog/bring-innovation-anywhere-with-azures-multicloud-multiedge-hybrid-capabilities/

https://www.crn.com/slide-shows/cloud/aws-outposts-gm-off-the-charts-interest-in-hybrid-cloud-service

https://www.hashicorp.com/cloud-operating-model