When you talk about Oracle in combination with the cloud, the thoughts of many OpenWorld visitors will immediately go to Larry Ellison’s keynote, in which he is mainly trying to indicate that Oracle is better, farther, bigger and so on is then AWS. As far as we are concerned, that is unfortunate (even though it regularly produces witty moments, when Ellison makes another joke) and it detracts from the unique story that Oracle can certainly bring when it comes to the cloud. The latter is therefore what we want to focus primarily on in this article. First, however, we will briefly comment on some of the cloud claims that Ellison makes on stage during OpenWorld, because we cannot ignore them altogether. Bold statements The claims made by Ellison to prove that the Oracle Cloud is better than that of AWS can, in our opinion, be criticized. For example, he continues to compare Oracle’s Gen2 cloud, including better separation in the basic architecture that contributes to security, with Gen1 from other parties. Oracle is of course the winner in that comparison, but other public cloud providers have not sat still and have already reached Gen2.
Another statement that raises eyebrows with us is that he states that the Oracle cloud will have more regions than AWS next year at the time of OpenWorld. That seems to us more of a definition discussion between what is called a region in one and the other. It certainly does not mean that the Oracle Cloud will be bigger than AWS’s next year. Of course he does not say that either, but it is clearly the connotation that is sought with it. In addition, the link with Microsoft’s data centers (see below) also seems to play a role in this. Some of Ellison’s claims regarding the database types that are both serverless and elastic also seem a bit far-fetched. For example, he states that only the not-too-often-used Aurora is with AWS, but there are two more that we think they know. Anyway, we find the claim that the fact that there are so many different types of databases at AWS is a weakness, not enormously strong. The one does not necessarily have anything to do with the other.
Where Ellison does have a point for us, is in the field of Autonomous. An autonomously operating cloud will respond more predictably than one that is not. Provided that Oracle initially built it properly, it should reduce the number of errors and at least exclude human failure. That would ultimately lead to a cloud that is safer in the basics. The announcement of Autonomous Linux during OpenWorld also ensures that the OS running on top of the architecture is given the same autonomy. Mind you, when it comes to Autonomous, you naturally shift the problem partially, for example towards the patch servers. These will then become targets for malicious parties. But below the line we certainly agree with the overarching point that Ellison wants to make: fewer people who interfere with the configuration of environments means fewer vulnerabilities. Finally, the ability to run Exadata databases in the cloud is certainly a distinguishing factor for OCI (Oracle Cloud Infrastructure). This was also demonstrated in an Accenture keynote, in which the difference was certainly not small and cannot be explained as a benchmark margin. The new on-prem Exadata systems (X8) are also very well received, if we just listen to that, so that story seems to go well anyway.
Unique position But enough about the substantial differences with the competition. The proposition of Oracle is of course certainly interesting enough (you could say ‘unique’) to take a closer look. Whereas the competition briefly approaches the cloud from IaaS and attaches more and more value to it with PaaS and SaaS, Oracle enters it from a different perspective, or actually several perspectives. There is of course the OCI story, but that is far from the only approach one chooses at Oracle. Basically, a full-stack approach to the cloud story was chosen. As a party that has traditionally been active on the on-prem piece, people also come from that side. See for example the section about running Exadata databases in the cloud above.
Finally: wealth or necessity? If you are a bit cynical in life (which we are regularly accused of), you can say that collaborations such as those with Microsoft and VMware, but also the somewhat multi-headed approach to the cloud, are a necessity for Oracle. Head-to-head with the IaaS offer from AWS, both Oracle and Microsoft will have a tough job winning it, also in the longer term. Collaboration is then the only way to make a fist. All partners are of course also very welcome for VMware to serve the hybrid future. On the other hand, you can also interpret it as a substantially different approach, which is more in line with where the market as a whole now seems to be going. Namely a collection of environments, packed together in a distributed architecture. That seems to play a much less important role at AWS. Earlier this year we spoke to Ian Massingham of AWS, who stated that only CERN cannot run on AWS, otherwise it can handle everything. From that it seems that one has a much less distributed vision there. If you look at it this way, the partner-driven full-stack story from Oracle (in the longer term) might well cut more wood than you would say at first glance. Time will tell. We will continue to follow it well anyway.