How to Understand the Problems with Oracle Cloud

Executive Summary

  • Oracle has a highly uncompetitive cloud.
  • The major problems with Oracle Cloud are undermining Oracle’s account control.

Introduction

We could go through the Oracle Cloud in more detail, however, concerning usability, the use of a brochureware cloud offering that is just like SAP Cloud more designed to serve as “cloud propaganda” rather than be used. Oracle has made many claims proposing equivalence between Oracle Cloud and AWS that we think is incorrect. Oracle had a study published by Storage Review that promoted the idea of Oracle Cloud having better performance than competing offerings.

“There are a wide variety of cloud offerings at this point and even several large cloud offerings with AWS, Google Cloud Platform, and Microsoft Azure at the top of that list. While these cloud service providers offer many great products and services, one thing they typically don’t have is performance. When comparing cloud to on-premises, on-prem always beats cloud hands down. Oracle is looking to change this view with its cloud infrastructure offerings.”

Let us go to the last part of this quote.

“Oracle is looking to change this view…”

What Oracle is really trying to do is compare its bare metal against visualized environments. This is Oracle’s “go to” argument against the cloud, to promote the performance benefits of bare metal.

It is not a surprise, and known by everyone that has studied the area, that on-premises bare metal can outperform a virtualized environment. However, it does not follow that performance is not available from the public cloud.

Cloud Has Poor Performance?

The performance depends upon the resources applied against the load. We have covered how scaled resources can be selected in the cloud, how auto scaling can be enabled, and how “serverless” has autoscaling built in as a natural feature. Therefore the phrasing in this quotation is inaccurate. It is not that performance is lacking in the cloud. It is that virtualized environments require more resources than do bare metal environments. And it also misunderstand the bottleneck faced by databases. This is explained in excellent research paper The Data Center Evolution from Mainframe to Cloud by Nikola Zlatanov.

“The mainframe / terminal model was founded on a fundamental scarcity of compute resources. At the time, computers were so large that they required highly specialized rooms, HVAC, operators-the works. If you wanted to use them, you often stood in line, made your case, and were eventually rewarded with some of the scarce and highly valuable cycles via a terminal.

We now have so many powerful computers that we are willing to greatly sacrifice raw performance for manageability. You have so much compute out there, it is not about raw performance but simplicity and convenience.”

This is the issue. Oracle is presenting a view where the primary bottleneck is performance, when it isn’t.

Bare Metal Versus Cloud?

However, Oracle are not comparing their bare metal service against AWS’s bare metal service. Instead, Oracle seeks to compare Oracle bare metal against AWS/GCP/Azure cloud. But why? The comparison should be bare metal vs. bare metal and virtual machine vs. virtual machine. The problem is in this quote above, where Oracle sets the assumption before getting into the testing results. In this quote, public cloud providers like AWS, Google Cloud and Azure are compared against Oracle Cloud. However, Oracle Cloud is a hosting service. Yet, hosting and cloud are discussed as if they are interchangeable terms.

Gartner, an entirely profit-oriented firm which is extremely “big commercial vendor friendly” not prone to calling out a major funder like Oracle, calls out Oracle’s cloud in the following way.

“Tier 2 bare bones minimally viable product that’s not ready for production workloads” and “Customers need to have a high tolerance for risk”

But on the other hand, on other occasions, Gartner has sung a different tune regarding Oracle Cloud.

“These announcements are more cloud announcements than database announcements,” Gartner analyst Adam Ronthal told Techrunch. “They are Oracle coming out to the world with products that are built and architected for cloud and everything that implies — scalability, elasticity and a low operational footprint. Make no mistake, Oracle still has to prove themselves in the cloud. They are behind AWS and Azure and even GCP in breadth and scope of offerings. ATP [Oracle’s Autonomous Transaction Processing] helps close that gap, at least in the data management space.”

This quote from Gartner begins by describing well known features of cloud and then soft peddles Oracle Cloud’s vast inferiority to cloud with the following sub-quote.

“Make no mistake; Oracle still has to prove themselves in the cloud. They are behind AWS and Azure and even GCP in breadth and scope of offerings.”

Who is making a mistake?

How Good is Oracle Cloud?

Oracle Cloud is horrid and does not exist in the same universe as AWS, Google Cloud and Azure. And is why is having such an ineffectual history in something rephrased as needing to “prove oneself?” Why is Gartner assuming Oracle will?

Furthermore, if this is the end of the quote, then why does the quote begin with…

“They are Oracle coming out to the world with products that are built and architected for cloud and everything that implies.”

One would hope that Oracle built the products that they announce. Because if Oracle begins coming out to the world with products built by Salesforce or SAP, then there will obviously be trouble. So yes…..let us agree with Gartner that Oracle is announcing products that Oracle built. Furthermore, not only products that it built, but that Oracle both “built and architected.”

This quotation is what you get when someone is paid but has to dance around a bit to obscure an obvious truth, and a truth which the payee would prefer the reader to think of differently than the reality would show. We will leave it the reader to guess how much Oracle pays Gartner on a yearly basis. For that type of dancing, let us hope it is a lot.

A review on Quora was a bit more colorful.

“Oracle Cloud is an absolute nightmare. All of their IaaS and PaaS solutions are complete sh*t. Nothing works. Support does not exist. It is a leach that should be purged from the industry ASAP. I am a CCNP / RHEL / AWS Architect / Python 3 Developer / DataCenter & Cloud Architect / Senior Engineer with 15+ years of experience. If you do not have my qualifications don’t even think about launching your platform in Oracle Cloud. It’s a fledgling cloud running alpha code and at best in it’s infancy.

You have been warned. Also shame on Gartner for reporting that Oracle Cloud was the most innovative, well executed cloud. Gartner has lost all credibility for me at this point. Oracle Cloud will sh*t on you, your reputation, your company, cost you millions, and all the while their “cloud success” engineers will belittle your team. Stay the F*CK away. RackSpace, AWS, Azure, and Google are all viable alternatives.”

How Many Four Letter Words Are Required to Describe Experiences with Oracle Cloud?

We don’t mean to violate the virgin ears of our readers, but this type of quotation is not unusual for online comments about Oracle. If you don’t like swear words, stay away from Oracle forums as the comments seem to have far more use of expletives than any other vendor that we research. In terms of the number of expletives applied to them, Oracle is the clear industry leader.

All of this makes an Oracle cloud sale nearly impossible for any mature organization. CIOs have to explain why they went with a tier 2 immature provider when they could have chosen a proven and safe leader like AWS, Google, or Microsoft. The only viable solution for Oracle is to start targeting small and medium businesses who use the Oracle database. However, this means they have to compete with Amazon and Google on price and let go of their 80% margin business model.

That will be a problem for Oracle. Discussions around reducing the margin are not responded to well from the top of Oracle.

Disagreement on Oracle’s Strategy with Cloud

In September 2018, Thomas Kurian took a leave of absence as head of Oracle Cloud and appeared to be leaving the company.

The reason?

“Kurian wants Oracle to make more of its software available to run on public clouds from chief rivals Amazon.com Inc. and Microsoft Corp. as a way to diversify from its own struggling infrastructure, a view opposed by Ellison, one of the people said.”

Larry Ellison has a history of pushing executives out of Oracle (if they disagree with him, or he gets jealous of them), and Thomas Kurian appears to the latest casualty. Kurian should have learned from Safra Catz who has survived by never contradicting Larry. However, Kurian was acknowledging (internally at least, never publicly of course) what people who are objective and have studied the Oracle Cloud know. Oracle Cloud is not a competitive offering. Oracle Cloud is a significant liability for customers that use it and is entirely about locking companies into Oracle products. Oracle’s cloud strategy is to “own the entire stack, top to bottom.” It wants to run its databases, run its apps, and run its own analytics/BI. Just as with SAP, the business model is not open to other major software providers. Both vendors want to offer IaaS/PaaS to the market, but neither vendor has any capability in the area. In contrast to Oracle and SAP, AWS and Google Cloud are open to all software providers. Even Microsoft, which sells competing software products say in the ERP space, has Azure open to all software providers.

AWS’s Lead

A testament to AWS’s growth is that more Oracle databases are running on AWS than on Oracle cloud. This is due, in part, to the substantial lead AWS has over all other cloud providers. However, more importantly, it’s because AWS is an open platform that supports all vendors, including open source. Customers can run any product imaginable on AWS, including Microsoft databases and apps, Oracle database and apps, and SAP databases and apps. AWS is not designed or optimized for AWS. AWS is designed and “tuned for everyone.” These same things apply to Google Cloud, except the number of options is smaller, but the emphasis on openness is the same. Kurian knew how little chance Oracle Cloud had to create a viable IaaS, Larry Ellison disagreed, and could not take it. One can work as a top executive at Oracle as long as you agree with Larry but disagree and your days are numbered. Acceptance at Oracle is going to be a tricky thing.

We interpret this as a significant error on the part of Oracle and a loss for Oracle customers. This is because while Oracle has endeavored mightily, they have managed to coerce very few companies into using their cloud. It is the view of one of the authors, Mark Dalton, that at some point Oracle will come up with a functional cloud offering. However, by the time that arrives, how much further advanced will AWS and Google Cloud be? Right now is a perfect time to be a technical employee at AWS or Google Cloud because Oracle is desperate to improve their cloud, has a lot of money that it is willing to throw at AWS and Google Cloud Engineers however, with all that money, Oracle has shown no ability to develop cloud competency.

Improvement will have to come from hiring employees from AWS or Google or other places, rather than anything novel coming from inside Oracle. However, they have already hired quite a few AWS and Google Cloud engineers. If the current Oracle Cloud is all that they can muster for all those high priced hires, the future of cloud at Oracle does not look good.

Faking Cloud Revenues

Oracle’s stock declined when they made an accounting change in how they reported cloud revenues that chose to obscure their cloud revenues. They had to, as the actual story is so different from the projected story. The story Mark Hurd called a “nothing burger” is a “huge juicy In and Out Burger.”

Wall Street is finally beginning to figure out that SAP and Oracle are cloud poseurs.

For Oracle, cloud means the following:

  1. No upfront license revenue
  2. Recurring revenue is not guaranteed since IaaS/PaaS are consumption based
  3.  Massive CAPEX spending on data centers, cloud DevOps resources, and customer success resources. Hiring engineers from AWS, Microsoft, and Google Cloud is very difficult and very expensive.
  4. A massive spending to transition partners to cloud technology
  5. An enormous loss in service revenue both for Oracle and their partners since cloud SaaS is not customizable and IaaS/PaaS is automated.

Oracle’s cloud strategy is about customer churn containment. Doubling the cost of licensing on competitor clouds is one example. And hiding cloud revenues is another.

Financial Disclosure

Financial Bias Disclosure

Neither this article nor any other article on the Brightwork website is paid for by a software vendor, including Oracle, SAP or their competitors. As part of our commitment to publishing independent, unbiased research; no paid media placements, commissions or incentives of any nature are allowed.

Search Our Oracle Cloud Content

References

AWS and Google Cloud Book

How to Leverage AWS and Google Cloud for SAP and Oracle Environments

Interested in how to use AWS and Google Cloud for on-premises environments, and why this is one of the primary ways to obtain more value from SAP and Oracle? See the link for an explanation of the book. This is a book that provides an overview that no one interested in the cloud for SAP and Oracle should go without reading.