The blockchain era is here but big biz, like most folk, hasn't a clue what to do with it
Now Oracle and others are wading in, like a BaaS
As far as database giant Oracle is concerned, the October announcement of its Blockchain Cloud Service (BCS) was timed perfectly.
After dabbling with blockchain internally for a year before joining the Hyperledger consortium in August, it had reached the point where its customers – and a market intrigued by what looks like another money-making tech upheaval – were ready to hear the sales pitch.
What they got with this shiny hunk of Hyperledger "blockchain-as-a-service" (BaaS) were smart contracts, permissioned secure access through Oracle's Identity Cloud Service, and integration with big IT through REST APIs. Organisations could use Oracle as a Platform-as-a-Service (PaaS) solving the difficult storage provisioning, scaling and latency problems while they build distributed ledger projects on top at their own speed.
"Oracle has been researching and evaluating blockchain technology for some time, monitoring the maturity of the technology and adoption by customers across various industries," group vice president of blockchain product development Frank Xiong told The Register.
Using Hyperledger to avoid vendor lock-in, BaaS was a perfect model sectors such as banking, payment, supply chain, healthcare, government, and tamper-proof Internet of Things, he said, offering up the early-deployment examples every vendor mentions.
Thus configured, BaaS appears to solve the biggest uncertainties organisations have about where to start with blockchains, namely what they run it on, how it integrates with what they are already have, and how they scale something that by its nature involves shared infrastructure. Cleverly, Oracle's pricing model is transaction-based, allowing organisations to experiment without having to pay for expensive licences upfront.
Blockchain bros' London powwow: Regulation, education, oversaturationREAD MORE
Predictably, Oracle was accused of jumping on the blockchain bandwagon, which misses the point about how alien this still is to customers looking to invest. If blockchains are to fulfil expectations, organisations must move beyond proof-of-concept trials. But doing this for something as complicated as blockchains requires help from service providers, developers, startups with new ideas and, of course, familiar names like Oracle. If this really is a revolution, it will be built on infrastructure that not every blockchain wannabe has.
As of early 2018, BCS is not yet officially on sale, but its announcement is still a boost for a market brought up on IBM's Bluemix Hyperledger Fabric and, latterly, Microsoft, which has spent the last two years firing various Ethereum-based initiatives at blockchains through Project Bletchley and the Coco Framework.
Other big brands have already piled into versions of the BaaS concept, including SAP, HP, Accenture, Deloitte (Rubix), and PwC, plus a clutch of financial-oriented names such as JP Morgan Chase with its payments network. Analysts often talk up the startups, but what is striking about the so-called blockchain revolution is how many familiar names have positioned themselves right in the middle of things early on.
It looks like the beginnings of a competitive market. Does this mean mainstream enterprise blockchains might be on the horizon?
Like its rivals, Oracle likes to use the BaaS tag to shield a lot of uncomfortable complexity behind a familiar idea. Blockchains aren't easy to explain, but the concept of service sounds simple enough, so hell, let's use it. Sceptics are not hard to find.
"The term BaaS is meaningless," argues Forrester principal CIO analyst Martha Bennett. "What is being talked about is mainly an infrastructural service on which you can run your blockchain."
If you're a cloud customer, a BaaS will allow an enterprise to host their blockchain distributed ledgers inside an environment already being used to host other applications, processes and management functions. The enterprise builds out rather than starting from scratch. What it's not is a ready-made blockchain – it's still up to the BaaS customer to get busy.
"If I want a bit of CRM, I get my credit card out and go to Salesforce and I'm in business. Every single blockchain network still has to be built. The degree to which it can leverage components that are already there depends on the use case."
In the case of blockchain consortiums, a customer looking at the BaaS concept right now would tend to be the "genesis block" on which this is founded. That alone implies that it will appeal to very large organisations, be those enterprises of key importance to a supply chain network, or perhaps even governments.
BaaS to the future
What IBM, Oracle, SAP and, to some extent, Microsoft are promoting with BaaS are parallel universes in which public, private, hybrid or consortium blockchains can exist. Within each, even industry-standard Hyperledger, pretty much everyone admits that buying into blockchains as a cloud extension is only the start of the technical learning curve. These range from basic problems such as performance and latency (blockchains are slow by comparison with traditional databases) to the scalability that BaaS claims it smooths.
Security – including blockchain key management – also needs some explaining, albeit the same concerns plague any enterprise technology these days. And who understands these problems? Developers are still in chronically short supply, as few as one in a thousand of the pro coding population, according to an estimate by analyst Juniper.
Bennett thinks that before organisations worry about the basic parameters of a blockchain deployments on BaaS, enterprises must confront deeper operational objectives. BaaS is at best a modest building block.
"A lot of enterprises still approach blockchains the wrong way. Let's start with a tech and find something to do with it," she says.
"You need to start with your use case. It's your use case that determines your governance model, and your governance model that determines your architecture. And until you know your architecture, you can't even select a supplier."
For private blockchains to function and scale, they must define an agreed set of rules to cover things like exception management, error handling and fraudulent transactions, without which they will never extend far beyond the enterprise itself.
"They need to work on governance models, that is what are the network participants' rights and obligations, what is the legal framework? Everybody talks about blockchain networks introducing transparency. But with every project you look at you hit the confidentiality head on."
Bennett's broader point is that enterprise blockchains, whether configured on top of BaaS or not, are primarily an organisational challenge. This is not an investment in a new service or technology but an entirely different, radically distributed way of doing business that is likely to have unintended consequences.
Public blockchains serving booming cryptocurrencies are still held up as the use case for so much of what falls under blockchain hype. Granted, the potential for large-scale disruption here is huge. By contrast, building private blockchains to serve business niches will never attract the same attention, leaving their complexities underestimated.
Today, Hyperledger is probably about as good as it gets, which is presumably why Oracle opted to take the path of least resistance. Just don't expect the future of BaaS, whatever it evolves into, to be simple, or to answer the deeper questions. Blockchains might even turn out to be less disruptive of mainstream business than some people claim, a glorified version of the business process re-engineering mirage that has been marketed like crazy at companies for 20 years.
Says Bennett: "A lot of the time when you look at use cases, you can solve them better, faster, cheaper with existing technologies."
Right now, this might not be what organisations want to hear. ®