Re-Inventing Cloud Computing
Service uptime is the holy grail amongst all of the public Cloud Computing provider giants such as Amazon Web Services, Google Cloud Platform and Microsoft Azure. Take the events from last week as an example. The AWS S3 (Simple Storage Service) suffered yet another setback this calendar year. It’s a minor degradation of service as compared to the major outage across the same U.S. East Coast region back in February 2017.
The latter was an infamous AWS S3 outage publicly described in the post-mortem here; due to an inadvertently overreaching service operation that cascaded across a network of subsystems which made it harder to recover quickly — and it all started from a human error! The economic impact to businesses were incalculable, likely in the millions of dollars, if not more! Much has been written about these major outages. They were heavily debated while others have been highly critical of how things are run within Cloud Computing provider giants.
Can blockchain’s Cloud Computing do any better? I couldn’t find a mapping of blockchain equivalents for Cloud Service stacks. So, I drew one up.
It’s awfully unfair to map the traditional Cloud Computing stack to the Blockchain technologies above. Take Etherum for example.
Ethereum not just a distributed database. Ethereum can also represent real asset value, and has an application service layer (via smart contracts) — this has no parity when in Cloud Computing. The database portion of Ethereum is a side effect from its built-in state transition functionality which acts more like a ledger. Its built in smart contract concept enables Ethereum to behave much more than just a system of records which traditional database tends to offer.
What else can Ethereum do? The white paper suggests that it can perform actions on tokens that represent real assets such as transferring them between parties securely. While doing so, these transactions are performed through a fee-based service by paying miners to help with confirming the transfers by multiple decentralized parties, in a bank settlement-like manner but without a concept of a central bank. Payments are sent in a currency called Ether.
Ethereum can operate Decentralized Autonomous Organizations (DAOs) with voting rules built in. It can also manage workflows such as multi-signature escrows. Insurance companies could build smart contracts that help hedge against crop yields. The white paper also suggests possibilities in peer-to-peer gambling!
As for IPFS (or Inter-Planetary File System), it is way more than content distribution because it’s touted as the HTTP for Blockchain. Filecoin and Sia both excels at Storage and Content Distribution, not only hosted in a de-centralized manner, but also secured by the blockchain’s cryptographic protocol.
And gosh, there’s also this thing called EOS, where would it even fit within Cloud Computing? Dan Larimer, the man behind EOS claims that it is an Operating System for blockchain development… but it reminds me of Heroku which offer developers to focus on building apps instead of worrying about deploying them at scale.
There are blockchain technologies that actually maps tightly to traditional Cloud Computing stacks. Take Golem, which maps to the Compute stack, offering computing resources from a decentralized network of users. Also, Namecoin maps closely to DNS, as a decentralized way of registering easy-to-remember names such as bitcoin.bit domain (representing blockchain address, democratizing and securing the Dot-bit domains names). IPDB is modeled after IPFS where it promises to be a secure and performant store of records amongst a decentralized set of nodes.
I thus challenge thee with this quote.
“It’s lunacy to imagine a decentralized set of Cloud Services in its current state of infancy to claim being able to demonstrably outperform world-class Cloud Computing providers with better service uptime? That … is … simply … impossible!”
It wouldn’t be just impossible, it would surpass the possible realm and it will be much greater than that. Over the course of time spent developing these decentralized blockchain solutions, there will be much more practical to use of these Cloud-like components to securely deliver highly available and resilient services imaginable.
It was naive for me to quote earlier that existing Cloud Computing providers will wither away. These major providers will likely contribute to the foundation of blockchains and thus strengthen the decentralized nature of blockchains. Once they start participating to throw in compute power, storage capacity and throughput bandwidth into the blockchain network, they will strengthen the blockchain through the network effect every time new nodes are added. When actual users begin interacting on applications built on blockchain-backed Cloud Computing platforms, it theoretically becomes stronger, more responsive, more secure against attacks, more valuable with its tokenized economy and feeds the developer community that will continuously add value to the ecosystem.
The world’s largest public Cloud Computing provider, Amazon Web Services calls its annual conference re:Invent. It’s simply ironic because the decentralized world of blockchain technologies are beginning to solve many of today’s Cloud providers offerings that may one day actually re-invent Cloud Computing from the outside.
