IBM's Red Hat gobble: Storage will be a test of Big Blue's commitment to open-source software So we all know by now that IBM is buying Red Hat for $34bn – the largest software acquisition ever. Of course, that includes all Red Hat's storage products. There are the two big ones, Red Hat Ceph (object storage with file and block access) and Red Hat Gluster (Scale-out NAS). Then there are the stragglers – the attempt at …
>> In other words, IBM's proprietary storage software will still
>> effectively see Red Hat software as competition.
Remember when MySQL was the one database everyone was flocking to? What has Oracle done with it? Right. It doesn't make sense to support development of a lower (or zero) cost product eating into your high margin one.
"Open source is the default choice for modern IT solutions."
The statement is true, if you pay attention to the word MODERN. Unfortunately, modern solutions are not so prevalent. Look at government. Look at just about any large organization that has been around for a while (long enough to be infiltrated by PHBs, dancing around in a macabre mockery of what that place used to be). Those are not using open source as they are far from being receptive to the concept "modern".
Those are not using open source as they are far from being receptive to the concept "modern".
And yet there's often a lot of heated bandwagon jumping when a IT new buzzword comes to their attention - probably in the hopes that they can fire everyone under senior management and still remain functional and profitable.
Look at government? Amazon and Microsoft are both tossing fat wads of cash at the "public sector". Similarly, the "public sector" is trying figure out how to "do that cloud thing". Hell, you look at DoD - one of the more staid sectors - even they're trying to make a sea-change towards cloud (look for stories about JEDI, "C2S" and GovCloud).
While definitely not as nimble as start-ups, there's a general move to try to shed the slowness that typically plagues them.
My dear - you assumption regarding Government's not being Open (and to Open Source and to the modern world is not correct). Best example is the Belgian Government who have made a choice long time ago to go Open Source first. The Belgian Government Cloud as what they call a Green Shift in which all modern workloads and new applications are built in containers on an RedHat's Openstack - including Openshift, container native storage, JBoss MW... Even for their Database they have taken a drastic shift towards EDB instead of Oracle...
You are right though when stating that this is only the beginning for Open Source and and modern, Cloud native technologies to only gradually starting to replace the traditional ones out there... The future looks very bright for Open Source and for RedHat - IF and only IF IBM doesn't kill this wonderful company...
Open source is just cheaper. It's far from being more "modern". Actually, often it has a lot of cruft accumulated in the past fifty years which graybeards can't get rid off - just like the basements of many old people.
That's why in storage - where every bit, every cycle and every microsecond count, and the hardware itself has a eye-watering price, some cheap or free software stuff doesn't make the difference. Of course when you have to install ten of thousands machines of commodity hardware and sell their services to others, the cheaper the software the higher the profits - and you don't really care to sell state-of-the-art systems, just good enough.
I know that this article is trying to look at this thing from a storage perspective, but that's really not what this deal is about.
IBM's storage sales have been flagging, in part, because so many people are getting storage from their cloud provider. As more customers move to to AWS, Azure, Alphabet and Alibaba, there's less organizations looking to directly buy storage. And, while the various public cloud providers plow through gobs of storage, they're generally not putting it on the type of storage that IBM sells (especially since the got out of the business of selling bare disks).
You look at what IBMs been trying - and failing - to do the past several years, it's been trying not to lose relevance as significant chunks of their customers make the move to cloud. IBM's made a number of cloud-oriented acquisitions ...none of which made the difference. The Red Hat buy is IBM's way of saying, "we know we've lost the public CSP game, but, with a unit like Red Hat in our portfolio, we establish a good beachhead in *every* CSPs environment against we can sell consulting services and management and other software".
Sorry but truth hurts. Red Hat storage is just not very good storage. Businesses buy storage because there is nothing quite as important as their data. They don't take risks with products that are incomplete and don't have the needed ecosystem of products for data management, virtualization integration, and so on. Red Hat is up against industry giants who invest heavily in their products and have been for decades. It's probably an unfair comparison given the monetization they are able to put into storage. But the fact remains...just not as good as the competition...open or not.
Opensource storage just isn't very good at what exactly, traditional workloads? Ceph will run VMware, but that's another story.
If you're still thinking virtualisation is the way to modernise then it shows that you're behind the times and should stop reading now/ you're about to be made irrelevant! The world is moving on to containers/cloud native apps and webscale storage is required for this and other mode2 workloads. Opensource storage (Red Hat included) allows for these and when you look at the announcements from some of the anaytics companies, moving from HDFS to containers and S3, you understand why traditional storage just won't work in the future. In my opinion (and I have no insider info) IBM are most likely buying Red Hat as they see their market share reducing with Azure, AWS, Google etc eating it away. By investing in a PaaS solution, which Opensource storage is the default, IBM will probably feel they can influence the many customers who need either private or hybrid clouds. Remember these public cloud guys want to make money, they entice you with deals and how to migrate workloads in, getting them and your data back out isn't so easy and certainly not cheap! But if you can transparently move between private and public, or from one public cloud to another then it allows you to reap cost benefits without lock-in.
So you can ask your traditional storage vendor to make fasters horses... or you can embrace the future in the same way you probably did not do with x86 virtualisation 15years ago. Remember F1 cars are the fastest, but if you need to move a wardrobe then a Transit van will be more suitable for the workload!!
Just my 2 cents.
*I'm not an employee of Ford Motor company and my views are entirely personal
Over the past 25 years there have been a series of purchases made by IBM. All of these fail. Starting with Lotus, moving to DataMirror, TeaLeaf, XIV (which did well by eating IBM's DS8k business), Resilient....The model is the same: buy, increase prices, reduce dev and support, resource alignment, sell to top 100 clients under massively discounted ELAs, resource action, reduce dev and support, rinse and repeat.
There will be a raft of financial engineering to recover these costs but it smacks of desperation.
$42bn for RH vs $63bn for EMC & VMWare. I know where I'd rather be.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
