June Update on 2019 Predictions from the Wasabis
6 months ago in January, Dave made some predictions on what was going to unfold in the market in 2019. Rather than wait until the end of the year, we thought we’d do a mid-year update and see where things stand.
1. Cloud Standards – De Facto or Vendor Neutral?
The Amazon S3 API will become accepted as the de facto standard interface for cloud object storage. Microsoft will not use that API and will attempt to keep a standards war alive.
June 2019 Update:
This prediction is more about whether there is “one standard to rule them all” as opposed to multiple, competing standards, particularly standards that are vendor-specific.
With more than 200 technology partners that work with Wasabi, the vast majority had barely any work to do to vet that their technology worked with us, as the S3 API is a well known standard, and clear winner for the vast majority of 3rd parties.
Amazon, Microsoft and Google still fighting it out.
Google isn’t all in on the S3 API as their only API, which serves to fracture the market. They had originally created their own S3-like API, and clearly intend it to be their preferred API, with some underlying differences that appear intended to lock-in customers to google-specific cloud implementations, despite beginning to offer direct S3-compatibility as an alternative.
Microsoft Azure and Azure Stack – appear to have no interest in supporting the S3 API.
Even on-prem vendors are using the S3 API, for the same reasons that we do. It’s already well established, and more and more vendors of all kinds (on-prem, cloud, hardware, software, open source) are freshly adopting the S3 API. There are other things we could all spend our resources on compared to creating yet another storage API.
2019 will be the year of multi-cloud services. No more walled gardens.
June 2019 Update:
Yes to the first part, no to the second part. Multi-cloud is certainly expanding, being led by many companies honing in on offering the next-generation of specific functionality, whether it’s pure infrastructure like Wasabi for cloud storage, Packet for compute, Cloudflare for CDN and security, etc..
And for all of us outside of the Big 3, not only are we continuing to focus on being disruptive improvements in our technologies, but we’ve banded together to reduce or remove egress fees between us to make the choice to pick best-of-breed, multi-vendor cloud solutions, rather than settling for the convenience, yet high price of doing everything with a single provider.
Amazon, Google and Microsoft continue to have no interest in tearing down their walls, and continue to build more and more services, more tiers, more, more, more, in an attempt to keep customers within their walls.
3. Fewer storage tiers
The industry – both vendors and buyers – will finally get it that having infinite choices for storage doesn’t do anyone any good.
June 2019 Update:
This hasn’t happened yet broadly. So many people are addicted to the idea that more tiers are better. While other vendors continue to add more tiers, our customers tell us that the brainpower they’ve freed up from no longer caring about how to implement many different tiers and reconcile the financial as well as operational details, has let them focus on more value adding areas.
For example, Amazon Reduced Redundancy Storage (RRS) is still offered – why offer that when the stakes for lost data are so high? 99.99% durability = as many as 12 million files lost per year per petabyte. Compared to Wasabi with 11 9s – one file lost every 8 years per petabyte. (And BTW, RRS is still more expensive than Wasabi, even if it is slightly less expensive than S3)
Sidenote: For more on durability – See our blog post on this…What Does 11 Nines of Durability Really Mean? https://wasabi.com/blog/11-nines-durability/
4. The end of egress fees
No more paying to use your own data – slowly, the industry is going to realize that customers hate egress fees more than anything else about cloud storage. Unpredictable monthly pricing from unpredictable usage makes customers rightfully angry.
All the independent vendors like us see the benefits of no egress fees – the price for connectivity and compute power tied up in egress costs is radically less expensive now compared to when public cloud providers first came online. Why shouldn’t those savings be passed on to you?
Cloudflare – Last year the Bandwidth Alliance led by Cloudflare showed that we weren’t alone in thinking egress was universally hated, and we are proud members.
Media Innovation Cloud Alliance – And this year at NAB in Vegas, we announced our Media Innovation Cloud Alliance, with 14 partners who have also either eliminated egress charges, or radically cut them for mutual customers.
Internet2 – Our integration within the Internet2 world also benefits from reduced egress charges between cloud partners by virtue of the high-speed network that Internet2 members take advantage of, avoiding the public internet. We’re the only pure play cloud storage vendor on that network, and researchers love us for enabling them to focus their spend on research work, not bandwidth.
Big 3 – The big 3 still aren’t touching this – they’re busy battling each other for mindshare, and frankly, they’re making a killing with the margin on egress.
Amazon – If anything, Amazon is teasing customers with low costs for Glacier Deep Archive, but 9 cents per GB for access at small volumes turns into roughly $100 per terabyte including API calls, storage costs, retrieval costs and transfer fees out of AWS to anywhere else. So much for the $1/TB advertised cost.
5. Ever lower prices and faster speeds for storage
In the brief time that Wasabi has been on the market, we have seen the capacity of hard drives increase from 8TB to 15TB. We predict a 20-30% annual reduction in the cost of disk storage for at least the next 7 years.
We’ve moved from 14TB to 15TB this year, and just this week, Seagate and other manufacturers announced they’re beginning to ship 16TB drives, specifically for hyperscalers and cloud users. 20TB drives are around the corner.
6. Special Bonus: Object Storage will dominate from a size perspective
The vast majority of all data will be in object storage because the volumes are in video (media & entertainment, consumer video, surveillance, etc), genomics (could become the largest consumer of cloud storage in the world), still images (consumer, industrial, defense-related, scientific), and IoT (everything these days is generating data).
We’re definitely on track here. Block storage is the old standard, but today’s world is all about objects – and BIG objects at that. Object storage has proven itself, and is far more flexible in nearly anyway you can conceive of, even masquerading as a file system, block storage, tape library, etc..
Old infrastructure choices continue to haunt companies. Legacy implementations are a ticking time bomb.
All told, Dave’s predictions are holding up pretty well. How will the rest of the year pan out?
We’ll check in again at the end of the year and no doubt we’ll have some fresh predictions for 2020 as we roll into the new year as well.