Where Does Blockchain Scalability Matter?

Posted Under: In-depth
3 years ago

Specific Use Cases, from Digital Art to HR

For blockchain tech to go mainstream, it needs to scale by Trent McConaghy:

In a recent Reddit thread, a user asked:

What’s the intended use case of BigchainDB? [the scalable blockchain database]

My short answer was: It’s for people who want a scalable, queryable database with blockchain characteristics (decentralized, immutable, native assets).

There’s a medium length answer too: put yourself in the shoes of a developer who’s built a career building applications for the web. You’ve become very good at using databases like MongoDB to manage mountains of data. You insert, retrieve, and update data using a query language. It handles massive scale — millions of records in a heartbeat. “It just works”.


Credit: Good Free Photos, CC0/Public Domain License

Now… imagine you’ve just heard about graph databases, and see the benefit because you model some artifacts as graphs — great, learn the interface and use it side by side with your relational databases, document stores, etc.

Now… imagine you’ve just heard about blockchain databases, and want one of the benefits: you want shared control (decentralization), audit trails (immutability), or assets that just “live” without a substrate that owns it (native assets). You want one of the blockchain benefits for yourself, and decide to deploy a blockchain database. Great — learn the BigchainDB interface and use it side by side with your other databases.

An even longer answer is in our whitepaper.

Specific Use Cases

My answer above led to a follow-up question:

Hi Trent. Can you be more specific or concrete in the use case? Anything meaty or tangible you can link to?

It was my pleasure to respond to this one! Here are a few.

1. The use case that started it for us was ascribe, to secure attribution and easily license digital art works to others. We had been iterating with photo marketplaces who wanted to have attribution claims of 100K images/day onto a blockchain, and licensing on top of that, with clean provenance (which means Merkle trees are out). That was infeasible with existing tech.

We also wanted to be able to query based on creator, title, and other metadata; without querying being slow. Infeasible. (You basically always need querying; if you can’t readily access the data you write, then why write in the first place?)

2. Then we built whereonthe.net, leveraging an index of 15 billion images on the web. We wanted to record the sightings of those images; and to be able to efficiently query what was written. This was clearly infeasible with existing tech: at Bitcoin’s throughput it would take a century to populate:/

To meet these needs, and seeing that others had challenges too, we then built BigchainDB.

3. The first lead user was Everledger, for diamond supply chain. Time-stamping when you have sightings of millions of diamonds. This was infeasible with existing tech.

4. HR credentials, with Recruit Technologies, which has millions of users. Link

5. Medical health records in the UK, via Tangent90. Link

6. Loyalty and rewards system, via Capgemini. Link

Each of these real-world applications has big data needs. When it’s doing attribution, licensing, value transfer, and so on. It needs scale for both writes, and reads.

We’re seeing many more applications, for three key benefits: decentralized / shared control, immutable / audit trail, native assets. Details. We’re seeing those benefits across six main verticals: here.

Finally, if you think about any web or industrial application that’s truly gone mainstream, then they need to have the scale to serve the mainstream. For example:

— The medium-size photo marketplace we described earlier: 100K images/day.
— Instagram has 5K images per second (=432M images per day)
— Electric meter billing in a single country like the UK is 50K records/s.
— Poor ad exchanges run at 100K transactions/s; good ones at 500K.

For blockchain tech to go mainstream, it needs to scale. At BigchainDB, scalability is job one.

# # # #

Trent McConagh has been raised in a pig farm in Canada, “hacking away on cold winter nights. 3D CAD tool, wordprocessor, dozens of games.” He holds a PhD in EE from KU Leuven, Belgium. Awarded #1 thesis worldwide in the field.

- Click to Read and Post Comments Total Comments: 0
  • No comments posted

Post Comment