Roadmap Update 11.19–12.03
Welcome to the fifth roadmap update! It’s a busy and demanding period, but also very exciting. We have a few important events ahead of us this month, and we can’t wait to share everything with you. For now, let’s look into what we have been up to in these recent weeks.
Pre-Launch Testing Results
In October, we pre-launched Syntropy Network and immediately started carrying out internal tests and aggregating data. The results exceed our expectations, with many more potential use cases opening up.
The pre-launch data proved that there are many optimizations that Syntropy can facilitate through public nodes. We tested around 200,000 pathways on the public Internet and compared the performance against our network. We managed to validate the technology by showing that more than half of web traffic can be improved using Syntropy.
In some cases, we managed to reduce latency by as much as 700ms. We are analyzing all of the outliers to understand such irregularities. We need to know whether this is due to a slow performing egress VM, congestion, software, or actual improvement. But the main point is that even if you cut out all of the outliers, most of the data shows ping improvements between 20ms and 200ms.
We have taken a more in-depth look into some of these improvements. For instance, we registered performance gains in a connection from Vancouver, USA to Amsterdam, NL. In this case, the Syntropy node that optimizes the route is located in the same location where the destination is. This could indicate a last-mile problem, where traffic is satisfactory on the public Internet, but performance slows down in the local network. The cause may be that the ISP is offloading the traffic onto a cheap network. This happens all the time. A lot of the Internet is still run on a least-cost basis.
Although optimizing long-haul routes is a high-potential use case of Syntropy, it’s interesting that the long-distance aspect is not always the main issue. We examined another interesting optimization opportunity from San Francisco, USA to Washington, USA. This is exciting because, as our CTO Jonas wrote on Twitter, if Syntropy reduces latency from SF to DC by 50%+, that represents a very tangible, high-demand use case for our tech. We found many more routes like this, and that is intriguing because we’re proving that Syntropy can route around congestion, not just optimize for geographical considerations.
We also ran tests with AWS Cloud system and discovered hundreds of potential optimizations, from short to long-haul distances. Some of the clients of AWS clouds spend millions every month, and they are very sensitive about performance. These optimizations would represent significant savings to them.
That is only a sample size of how much Syntropy could optimize AWS Cloud. Still, the exciting part is that these results should be replicable across other cloud platforms too, like Oracle Cloud, Google Cloud, Microsoft Azure and IBM Cloud. In fact, we’ll be sharing results like these early next week.
It’s important to note that we achieved all of these results using our pre-launch network solely. It will evolve and improve over time, exposing many more alternative routes and improvements. Once crowdsourced public relay nodes join the system, we will have an even larger and more dispersed global network, running on top of the public Internet. We are excited to witness its evolution from the very humble beginnings, and we hope that you are too!
Our developers have been on the frontlines of operation. From the very brief look into their list of achievements, it might seem slightly shorter than previous updates, but in reality, it’s the opposite.
Apart from the list of their achievements below, developers have been busy migrating to GitLab, cherry-picking bug fixes, setting up staging environments and polishing the release process between them.
Also, they have been working on extending and improving our documentation. That is a critical part to successfully onboard and activate users of Syntropy Stack.
What does it all mean? It means that we’re approaching launch faster than expected!
Below is a brief list of developers’ achievements over the last two weeks. Progress is now slightly ahead of our roadmap timeline.
- Set up specific ports for Agent so that new interface can be created only on those ports;
- Refactoring of the route overlaps detection;
- Set up PyPI deployment pipelines.
- Migrated Netbox to Kubernetes;
- Migrated Grafana and all of the data to Kubernetes;
- Working on alerting mechanisms and dashboards.
- UDP Load balancer and UDP NAT integration.
DARP (Distributed Autonomous Routing Protocol)
- Made a simplified command for an alpha tester to run DARP. Now it’s a single command as opposed to a sequence of docker commands;
- Nodes now connect automatically to the closest genesis node;
- Writing patents on several aspects of DARP technology.
- Implementation of Kubernetes and Host types of Services;
- Auth service decoupling and refactoring;
- Made improvements to user ID checking while performing various operations;
- Made improvements to Network filtering;
- Achieved an increase in Elastic integration performance.
Business Intelligence Module
- Performed lots of analysis to extract the optimized routes and visualize the results. You’ve probably already seen them!
- Performed heavy stress tests to define the capability limits of our current setup;
- Manually tested UI to nitpick the tiniest bugs;
- Automated testing initialization;
- Test code optimizations
- Connection service implementation and management via code;
- Implemented a microservice to export a .yaml file of a Network setup;
- Various updates to Ansible module;
- Set up PyPI deployment pipelines;
- Held live user-testing to polish the workflows and documentation.
- Improvements to Connection Status representation — implemented a ‘sleeping connection’ functionality to signify the period during which we haven’t received any updates about given Connection;
- Implemented filtering of endpoints while adding them to a Network Graph;
- Multiple design fixes and usability improvements.
Graphs & Analytics
- Graph and graph elements scaling improvements;
- Connection status representation within network graph;
- Added Connection status informational modal;
- Lots of usability improvements.
That is all for now. It’s Christmas season, so we are actively working towards milestones set in our roadmap. By the time you are reading the next update, we will have accomplished some of those.
We advise that you monitor these updates regularly as they will consistently track our progress towards full launch. Be sure to follow us on Medium, Twitter, Reddit or Facebook for a daily stream of updates. If you have questions, we are always on Telegram.