The middle of March was warm and cloudless in Vilnius, but on the last day of month we saw snow on the ground and I hope it was the last snow this spring. Here in Eastern Europe we urgently need the sun in order to absorb its light and produce positive emotions. One month of war in Ukraine has already passed with thousands of deaths, millions of refugees here and all around Europe, and, frankly speaking, no understanding when this horror will end.
On March 6 flespi, as a part of Gurtam, made a decision to cease its operations in Russia. April will be the last month of commercial relationships between us and our Russian users. The datacenter for the RU region, which was widely used from other countries in the CIS region, will operate a little longer in order for such users to migrate to the EU region with the datacenters in the Netherlands. We already started migrations and hope to finish them in April/May. But in any case we will cease RU region operations for all our users no later than August 31.
From now on I will refer to the datacenter (e.g. uptime) for our EU region only.
We finished March with 99.9952% uptime. Our NOC has reported two downtimes with durations of 79 and 55 seconds. The first one was caused by a magistral network link failure and the second one was due to a DDoS attack.
Now after the first quarter is over we can measure our performance relative to the previous year. So far we doubled our growth rate in various metrics and we are slightly ahead of our initial estimations. These are great results and we hope to keep up at this level in the second quarter of the year.
This may be a hard time for us since three members of our team are currently in the relocation process from Belarus to Lithuania. But four members are already there since 2020 so once the relocation process is over, our productivity should go back to the pre-COVID and pre-war levels thanks to mostly local (offline) work. Sergei Leuchanka, our BizDev guy and Sergey Buntsevich, who is responsible for the frontend applications engineering, will continue working remotely from the USA and Poland respectively.
Now some of the things we’ve accomplished in the last month.
We integrated three new protocols: general-motors-onstar, geosensorx and linklabs.
We would like to specifically highlight three small but really valuable features that appeared in March:
It is now easy to directly access an item configuration using its ID. This is an important feature for large accounts (>30K devices) where items enumeration can take quite a while and can consume signifiant web browser resources.
Invalid parsing errors reported by our parsing system now contain raw traffic related to the failure. This should simplify debugging issues with devices even further.
Now it is possible to select devices using telemetry value in the REST API selector. A whole new world of applications is now open for developers on top of flespi with this small feature.
For those who might have missed, I want to remind of a couple of our older features covered by dedicated articles in our blog — expressions construction & testing tool and backup & restore capabilities flespi provides for you.
In March we installed a refactored version of our telematics gateway service — this was the result of nearly a year of our work. So far nobody noticed anything and this is great. We achieved expected results in terms of greatly reduced RAM and CPU usage per service with much quicker service restart/upgrade performance. It greatly widens our limits of horizontal scalability. In April we will continue to tune it further.
flespi has been staying with Ukraine from the very first day of the war and I hope my next changelog publication will be positive on this subject. Stop the war!