Traditionally, when one year changes the other, we reminisce the goods and the bads, learn the lessons, and plan ahead. Flespi team is by no means different. While last year we focused on our faults, this year we are much more positive — I can’t even remember any really epic failures we had. Just some minor mistakes that any production-oriented company will have. Nothing extraordinary.
When I analyzed flespi achievements in 2018 I ended up with a list of 30+ items. Sorting it by importance I ran into one item that stood out and which I consider our major highlight of the year. All the features we implemented, all new products we delivered to our users, and all technologies we developed in-house will be absolutely useless without this notable circumstance.
It’s our final yearly uptime. According to our internal uptime calculation system operating in two independent locations and checking platform behavior and performance each minute with 10+ different tests we reached 99.995% yearly uptime in 2018. Just imagine, this is the “four nines and half” mark which means that our platform was (partly) unavailable only for about 27 minutes throughout the year. This includes everything — network failures, overloads, some unsuccessful upgrades — any kind of problem when part of the platform was not operating properly. Only 27 minutes!
Let us explain what makes us take special pride in this uptime number. It’s the fact that we installed 700+ upgrades in the course of 2018. Two or three upgrades every single working day of the year! This is what we call a dynamic environment ;)
We introduced conceptual as well as aesthetic improvements to various flespi platform components.
We released and opened our MQTT Broker to the public.
We grew to 100K connected telematics devices with 9TB 8+ months history backed by our storage system which implementation was also totally replaced this year.
We survived the mass removal of the test devices (huge load on our servers) and did many other less noticeable things.
We put a lot of resources in adapting the platform to the growth of our user base from 160 to 1000 accounts. As flespi is a backend available via REST API, many of our users occasionally tried this or that API call not thinking about the load. Calls like “get all (millions) messages from all devices” or “perform thousands of modification operations”. We survived those as well. I never saw similar a product getting features at the same rate as ours. Even with a much bigger team than ours. Never.
Such impressive stability is only possible due to the following three factors:
Network & servers. It is out of our control, but our hardware provider is doing magical things and is extremely reliable. We are with zylon internet for 10+ years already for our Wialon Hosting product and believe me — they are the best provider I’ve ever met.
Product-oriented leadership. Our goal is not to earn money by growing sales. Our goal is to deliver the best product. The one that will always be ahead of any competitors on the market. Years ahead. Then we hand-pick and grow the customers who can truly benefit from such product.
The team. Any technology is nothing without people and our people are the best.
In 2019 we are going to deliver some new exciting features among which not only the new technology for software engineers but also the Stage 3 product operating on top of the gateway and registry layers we already offer.
Our primary goal for the year is to deploy this new Analytics layer and keep the highest uptime possible.
In the background, we’ll keep upgrading the internal implementation of the platform by using the new technologies developed in-house. These upgrade projects can run for months — continuous improvement is part of our philosophy so we won’t count them as goals for the year. And of course, we will implement 100+ other new features, integrate more devices and protocols, and introduce a number of usability improvements.
According to the central figure of the previous year — 99.995% platform uptime — flespi is not a startup anymore. flespi is a stable, dynamically growing backend delivering highest quality service, first class customer support and unbeatable functionality that allows our Free and Commercial users to keep calm and just rely on us to significantly cut development and operational costs in their MQTT and telematics projects.
As we mature, we’ll do our best to make uptime for 2019 at least 99.995% to further reinforce your confidence and trust in us.
Thanks for motivating us to be better every day!