I’m amazed it’s even propping the quadrant up. What happened to that company. - apart from spending 30yrs doing stock buy-backs instead of innovating? Truly a fabulous example of how once great companies collapse and become insignificant.
Analyst outfit Gartner has published its annual Magic Quadrant assessing the world's leading Cloud Infrastructure and Platform Services providers, and in 2022 Oracle and Huawei were the big movers. Amazon Web Services, Microsoft Azure, and Google Cloud retain their places as the only three clouds in the prized Leaders box of …
Sovereignty belongs to those who have the biggest regional coverage and whilst Oracles is decent they are all well beaten by AWS.
Oracle whilst reasonable feature complete for a basic IaaS suffers from a big knowledge gap even among their own sales and consultancy teams.
You try pushing them to actually build terraform scripts for you.
Besides everyone knows half of Oracle’s services are just their old onprem service’s with a shiny cloud UI layer on top. They must have one of the lowest utilization densities of all the cloud players.
-> The contentious Chinese company
Why is it contentious? Why is Cisco with its litany of security holes not "contentious"? And why is Microsoft not labelled "ultra contentious"? Adobe too? Is the whole world supposed to suck up this American labelling of Chinese companies? From now on I would like to read in articles about Microsoft: the notoriously insecure Windows family of operating systems...
This is easy. If you visit any dictionary site, or even a paper version of something that defines words, you will see that this is effectively a derivative or similar to the word 'controversial.'
Controversial is something that Huawei has been for the past 5yrs or more across much of the western world -largely because it is suspected of being, basically, an operator of the Chinese ruling party with a stated aim of undermining the western hegemony.
That is why it is contentious.
Now, if you want to get into a debate about whether other companies have holes, back doors, insecurities or any such, then yes that's valid too. Indeed, if you've paid attention to the evolution of the enterprise technology landscape for the past 20yrs you'd note lots of market incumbents have had precisely this label.
Your example of Microsoft for instance has been hotly debated for a long time and was labelled all kinds of things during the years of it's strangleholds on browsers, desktops, Ballmer years etc.
However, as you may now note, these are very probably a distinctly different set of controversies.
The term "contentious" does not just apply where the politically correct want it to apply. In the context of ongoing business practices which impact cloud provision, or may do, every darn offering is highly contentious in one way or another. RedHat is possibly the only one that is contentious solely for religious reasons and not for any practical ones (dubious yes, contentious no). Much the same applies to all the other areas of IT. This universal contentiousness is after all what feeds the Vulture in the first place.
A propos of which, I think it high time the Reg website did an automatic find-and-replace when posting articles, to change "contentious" to "flaccidly normal".
For whatever it's worth: we use Oracle "Fusion" ERP, which lives in the OCL. Even given all the ways that I think Oracle sucks, I a still positively shocked at how quickly our workloads are decached and VMs apparently resource-neutered during even brief pauses in utilization. Worse, when you do return and need things to "perk up", it can take literally 10 or more minutes for things to return to normal speeds.
My overall impression is that Oracle knows f-all about building modern VM-based infrastructure. Unless they also know that most folks who would choose Oracle are already willing victims, and will continue to pay no matter how poor the performance. In which case, squeezing every possible penny/CPU cycle out of their stack, no matter how ugly the consequences, makes perfect sense.
Also, I have used the basic OCL compute resources "just for fun" by running odds and ends on Ubuntu servers. Initially they were significantly more performant than the competition. But something changed pretty much overnight about a year ago, and everything was scaled way back. Oracle's 1 vCPU went from being 150% vs most to being about 80%. Obviously, YMMV.