Do you know what an ideal backend means? Rock solid reliability, adequate performance, full transparency, and moderate growth in new features.
flespi in April was an absolutely ideal backend. Until 10 pm on the evening of April 30, with 2 hours left in the FET (UTC+3) time zone until the end of the month. We were holding a 100% monthly uptime in both regions and then one of our uplink providers — KPN experienced some issues in their network and BGP automatically switched to alternative uplink within a minute. Our monitoring system immediately detected and reported to the EU NOC 53 seconds of downtime for the region and we ended up in April with 99.998% monthly uptime for the EU region. But we did once again 100% monthly uptime for the RU region where almost 15% of our users are located. And this is great!
Before I continue and list the new features delivered in April, I would like to say thanks to all our users who believe in the flespi team, its manpower, and our technologies. April 2020 was so far the most successful month in the flespi history in terms of growth of everything — users, channels, registered devices, data stored, commercial leads, new contracts, and even traffic on flespi.com. Almost every countable metric grew up by 30% within one month. And this is not just one or a few commercial users that started using flespi in their software platforms. No, this is a lot of new users and projects from a lot of application areas — both IoT and vehicle telematics segments. We now see a lot of smart people doing cool projects with flespi, discussing them with our engineers in the flespi chat, and launching them on the market. Thank you a lot! Be sure — we will not let you down.
- A few new members joined the COVID-19 IoT Alliance and some flespi-based free projects are already running in order to help people and governments in some countries to overcome the virus.
- One of the most interesting and powerful features we introduced in April is the complete ecosystem for accessing original (raw) traffic received by the channel. All incoming and outgoing channel traffic is stored separately for each identified device. This is available to all our users for free and by default, we store up to 2MB of traffic per device for up to 10 days. This raises the flespi gateway transparency to the highest possible level and makes the communication with hardware support teams easy and straightforward as they always want to have a traffic sample for any kind of problem.
- Nowadays when most of us are working remotely, availability is one of the most important factors for any service. That’s why we saw a lot of DDoS attacks last month and I suppose their number and intensity will grow each month. To protect our users we introduced a personal DNS name for each channel bound to a special IP address and whenever a DDoS attack is detected, we will manipulate with these names in order to protect our users. Previous IP-based server specification in devices configuration is yet 100% valid but won’t give you the protection against DDoS so we recommend everybody to switch to DNS-based server specification.
- Again one more trend we noticed in the last few months is the use of BLE tags in all kinds of projects. In flespi, we could not stay away from this trend and did some research work on this matter. One of these research activities resulted in quite an interesting direction — indoor navigation with BLE tags. After the publication of this article, we got even more requests in that direction and now we are trying to implement additional indoor navigation use-case based on micro-computers. Stay tuned to receive it immediately after publication.
- One small notice for the flespi analytics system — we removed the experimental label from it and now you may use it in production services as the changes we introduce there are now regulated by our SLA.
- To increase usability and speed up the process, we gave our users a possibility to connect WiaTag and DistanceTag mobile applications to a flespi channel by simply scanning the QR code of the flespi device.
- A few more usability features for the flespi panel usage on large volumes include profiles configuration and simple one-click filtration by subaccount.
- We also implemented one more software tracker into flespi — Traccar client. It is open-source, supports both iOS and Android platforms, and can be white-labeled. Thanks to Anton Tananaev for this wonderful app.
We are working on the firmware upgrade feature and are now converting all integrated protocols from the older pvm engine to its new generation, which is internally labeled pvmII. The whole conversion process will take months but we are gradually performing it in the background and someday will finish it successfully.
We also work now on the user-configurable device plugins. It will give our users the possibility to define their own plugins and store additional structured information attached to devices. This information can be visualized in the panel, changed via REST API with integrated data format validation via user-supplied JSON schema, read with MQTT API, and attached to the streams.
We are moving fast, so keep up.