Blog
Cut up to 70% of blockchain infrastructure costs: Three cases of Dysnix solutions

Cut up to 70% of blockchain infrastructure costs: Three cases of Dysnix solutions

Daniel Yavorovych
December 2, 2022

After FTX collapsed, even the most positive participants of the global crypto market were forced to admit that the crisis had developed to its fullest. Dysnix got tens of calls from our current and potential clients with the most popular question. Most traders and projects are questioning the reliability of the crypto and exchanges they trusted a moment ago. And alongside people, money is leaving the market.

But each storm calms down after a while. And as a crypto or blockchain project, you must prepare to survive the crypto winter period and a future active spring with traffic spikes and boiling activity. The main things that will help you to stay on the market are:

  • Optimized resource usage and, as a result, lowered costs;
  • Readiness to answer your customers under any circumstances;
  • Automatization of all resource-consuming processes that your project has.

Here at Dysnix, we observed dozens of blockchain projects—from exchanges to blockchain-powered startups—that could be better prepared for the severe market challenges in terms of infrastructure. The way things are designed predicts how they will act. So, that’s why the infrastructure is the case that will save you or kill you during the crisis. 

When businesses address us with a solid desire to win the battle with circumstances and stay successful on the market, we customize the alchemical technical recipes to help them:

  1. Check the infrastructure flaws and eliminate them before the improvement stage;
  2. Check the processes flow: solve the unoptimized resource consumption;
  3. Seal everything securely: prepare for different states and events scenarios to keep infrastructure flexible and adaptable.

We’re applying complex DevOps and blockchain technologies, Dysnix custom tools, and innovations to conduct this plan. We want to share some cases of how we solve the various challenges: from high infrastructure costs, downtimes, and high latency, to hazards of security caused by infrastructure design. All our clients are still on the market and don’t feel the crisis in a painful manner; their end users feel even better. And that’s the best reward for us as technicians of these projects.

Simultaneous decrease of 2/3 of costs and speed up blockchain responses for Defi developers

The Defi developers who work with Velas blockchain and the Velas company itself are deeply interested in highly-available and cost-efficient blockchain infrastructure that’s easy to operate and access. Dysnix knows how to deal with this challenge. 

We suggested a comprehensive solution—an implementation of quick access to Velas blockchain historical data and an optimization mechanism for the response time that would reduce costs and eliminate periodical downtimes due to traffic spikes.

The solution

JSON RPC Caching Proxy, a product by Dysnix, invalidates cache with deprecated data through a smart algorithm and allows access to backend nodes only when necessary. It's a simple way to reduce costs for the node resources. 

What does our client get in the end?

  1. Significant cost decrease thanks to optimized resource use;
  2. Reduced time-to-market for blockchain projects;
  3. No downtimes, thus, significantly improved UX for Defi developers.

Cutting 70% of costs and defeating the monster of poor UX during the DEX exchange’s IFO events

One of the biggest DEX exchanges has come to us in despair. Each massive IFO on this platform turned into a nightmare because of constant high latency, user problems with buying a new token, and, as a result, a “shitstorm” of anger in social media and reputational expenses.

Solution we provided

After a brief audit of their problem, we offered a complex solution consisting of two main tools:

  • PredictKube—for auto-scaling and balancing the load to eliminate the latency spikes and prevent blockchain from overflowing.
  • JSON RPC Caching Proxy—for the overall decrease of RPS load on the nodes, fewer resources spent without affecting the user’s experience. 

This is a visible explanation of how PredictKube works:

The RPS trend and trends of prediction and nodes’ autoscaling by PredictKube

Thanks to a custom-trained AI model that predicts traffic trends, PredictKube can autoscale the project's resources in advance, so when the IFO starts burning, all needed resources are on, and there’s no delay with deploying. 

We implemented the JSON RPC Caching Proxy on December 13th:

The timescale of latency spikes on the project before and after implementation

As you can see, the latency spikes disappear and never come back. Thanks to this JSON RPC Caching Proxy, the consumption of resources becomes much more optimized and, at the same time, brings no harm to the user experience. 

What does our client get in the end?

  1. Cut off 70% of costs.
  2. Problems with downtimes—solved.
  3. Constant low latency at any load.
  4. No shitstorms on Twitter—users can buy anything anytime on this DEX exchange without errors or delays.
preencoded.png
The graph of costs of the projects after implementation of Dysnix’s solutions in February

Building up a secure blockchain infrastructure that can automatically handle ~70 inner services

We’ve got a request for the creation of blockchain infrastructure with a high degree of security and demands for load balancing of production and development environments for more than 30+ inner services wrapped up in CI/CD processes. Later, during the development process, this number rose to 70 services run on this infrastructure. And we were time-bounded, with only one month available for all processes. 

We have to offer a cost-efficient and secure blockchain infrastructure that will be able to deal with contrasting traffic trends. 

What have we created?

We created a highly-available, secure, and auto-scalable Kubernetes-based blockchain infrastructure using our tools and solutions. The project was armored with the Infrastructure as Code using Terraform, full Deployment of services as Code using Helm and Helmfile, and a fully automated CI/CD system using CloudBuild. Created infrastructure includes monitoring and alerting systems working on custom business metrics. The whole custom solution is based on a high-available scalable Ethereum JSON-RPC HTTP & Websocket service.

What does our client get in the end?

  1. Highly available, secure, and auto-scalable Kubernetes-based blockchain infrastructure from scratch in a 1-month term.
  2. An infrastructure cost optimization of 40-50%.

Meet any crisis as an adaptation challenge

The crypto winter will end, but until then, it would be wise to expect even more shocking news and be prepared for any plot twists. No matter how long it takes, things we make today to secure the business will help us even after the crisis.

Implement our solutions today to ensure your safe exit from the market’s turbulence with decreased costs and improved UX.

Daniel Yavorovych
CTO and Co-founder at Dysnix
Brainpower and problem-solver, meditating and mountain hiking.
Related articles
Subscribe to the blog
The best source of information for customer service, sales tips, guides, and industry best practices. Join us.
Thanks for subscribing to the Dysnix blog
Now you’ll be the first to know when we publish a new post
Got it
Oops! Something went wrong while submitting the form.
Copied to Clipboard
Paste it wherever you like