One of the major tech stories in the past 12 months has been the value of bitcoin. The exchanges have seen highs and lows, with huge spikes in traffic in some cases causing huge outages. For example, in November 2017, two of the biggest cryptocurrency exchanges, Coinbase and Gemini, went down, making it impossible for many users to buy or sell the digital currency. The result of this outage was a steep drop of 20% in the value of bitcoin.
At the time, Coinbase reported that it saw an "all-time-high" in web traffic that contributed to the system outage. However, this wasn't the first time it had experienced problems. Earlier in 2017, Coinbase crashed after an Asian buying frenzy, with Japanese investors rushing in to snap up cryptocurrency after it was made legal tender. Coinbase adopted an honesty-first approach, with CEO Brian Armstrong saying that the company is expecting more outages in the future during periods of high-volume traffic.
(Image: Thought Catalog, Unsplash)
Is blockchain making things difficult?
Poor performance isn't something that modern businesses and consumers will just put up with, so these issues need to be brought under control. So, what's making it difficult for the bitcoin exchanges? Bitcoin was the originator of blockchain -- some have called it blockchain's "proof of concept" -- and so it has seen a growing amount of attention as the value of one bitcoin rises above $1,000 and far beyond. In basic terms, bitcoin is a constantly growing list of records called blocks, connected and secured with cryptography. These records provide a verifiable means of recording transactions, but because the transactions are fully anonymous, privacy is inherent, meaning the cryptocurrency could be being used to buy and sell absolutely anything.
As blockchain becomes more widespread, the onus on how the technology behaves and performs will fall on teams integrating it into their backend stacks. However, blockchain use in web and cloud applications will happen in very complex IT environments, which will almost certainly include elastically scaling compute resources that exist only temporarily. Not only that, but these blockchains will be processing thousands of transactions every minute, the sheer scale of which will create monitoring challenges even for the most advanced IT organizations. This explosion in transactions and complexity will require an entirely new monitoring approach.
Monitor and track to understand the chain
Monitoring blockchains requires visibility into the entire technology stack and every digital transaction that is processed through it. You can't skip requests or sample/throttle data when monitoring blockchains; blindly trusting that the application, services, process, network or infrastructure layers are always providing 100% availability and optimal performance.
However, IT teams will also need to make sense of the insights generated through their monitoring processes. This is where deterministic artificial intelligence (AI) capabilities will be essential. By using tailored machine-learning algorithms, organisations can auto-discover and auto-baseline 100% of traffic, from the end user to the application to the blockchain and to their IT infrastructure. This AI-generated map will provide a complete understanding of all the entities, relationships and dependencies involved in the operation of a blockchain-based application, helping IT teams to determine the impact of performance-related events and what caused them.
However, understanding is only part of the puzzle. The next step is to use machine-learning algorithms to baseline, compare time frames and create multi-dimensional views of data to then determine and rank issues that inevitably occur in the complex environments where blockchain resides. As a result, businesses can build self-healing into their blockchain applications. For example, if a blockchain application is having a memory-based issue, a deterministic AI (machine learning) could detect this state and trigger a fix to adjust the memory setting automatically, maintaining the application's ability to process blocks.
An immense amount of power is needed to mine cryptocurrencies. (Image: Rafael Pol, Unsplash)
Blockchain has wider applications
Monitoring blockchain performance will be key if businesses are to avoid the type of outages we saw with the bitcoin exchanges. But it's not just the value of bitcoin that can be affected by outages; blockchain has other applications outside of the finance world that could be impacted.
For example, in the supply chain, blockchain applications track goods as they move and change hands in the supply chain, better organising tracking data and putting it to use. Walmart is currently working with IBM to use blockchain technology to track goods across every step of the supply chain -- getting fresh food to customers quicker and providing complete traceability. But an outage here could mean losing track of stock, or goods not being in the right place at the right time, slowing delivery to consumers and creating a negative impact on their experience.
As the use of blockchain technology becomes more widespread, it's going to become too important not to monitor it. Whilst it might feel like another brick in the wall of IT complexity, with proper management, blockchain can fit seamlessly into the digital ecosystem and provide value, instead of headaches for IT teams.
David Jones, APM Evangelist at Dynatrace