1 April, 2019

March 2019 change log

Major flespi improvements in March 2019

March is the greatest month in the year because cold and snowy winter suddenly surrenders to a warm and sunny spring. Flespi team loves the sun, as most Belarusians do, and we have proudly delivered quite a few features at a surprisingly high uptime — pure 100% of platform availability this month. This is our third 100% uptime month since we started to publically monitor and announce all our issues — either network or platform related. In April we expect some movements in our datacenter network structure, so the availability might not be that high — I will cover the details in next changelog. 

We have two main announcements for our users: 

  • First of all — we changed terms of use for our Free users. Our MQTT users should not be affected, but telematics hub users are now quite limited and basically current restrictions can only be used for development and testing. Any kind of commercial usage with 10+ devices is almost impossible. Our initial limits for Free accounts, like connecting up to 3,000 devices, were offered mostly to test the platform possibilities and to attract more developers. Now we see a steady growth of applications using flespi for various purposes and want to concentrate on companies that rely on our product for their commercial projects. The announcement was sent to all Free users in the middle of March and all users are upgraded to new terms starting April 1.
  • The second announcement is regarding our third layer — the Analytics system. Its 0.4 alpha-beta version is already in production and available via REST API. The system is only making the first steps into the world and will gain its main power when it starts operating in the automatic mode providing its users with precalculated and ready-to-use data. Right now it works only in manual mode, meaning each time you want to calculate something you need to do a REST API call. The engine can split device messages into intervals based on certain criteria, like parameters, geofences, date/time and then precalculate these intervals using custom counters. The only documentation we have now is the REST API docs, but soon we will publish a series of blog and knowledgebase articles on how to use it. You are welcome to shoot a question in HelpBox if you get stuck testing the newborn analytics — we are available there almost 24/7 ;-)

And some more stuff we've worked on:

  • For devices and storage containers it is now possible to access daily messages snapshots for the last 10 days. This service is free for all users. Snapshots are stored on separate storage servers and I believe this is an important feature for our users to feel confident about the flespi database.
  • Our Commercial users doing a hierarchical grouping of their items into sub-accounts can now enjoy it with greater experience as the flespi.io interface provides more and more tools for this — like the selection of a sub-account during item creation or tokens login links.
  • Speaking about tokens — we extended the possibilities for tokens ACL configuration and now it is possible to issue tokens with a limited set of 3rd level operations — e.g. issue a token allowing full remote device configuration, but not allowing to modify the device itself. 
  • In March we did a big revision of some protocols, namely wialon_retranslator, wialon_ips, and teltonika and performed the tuning of their parameters naming scheme. A lot of efforts was invested into Teltonika protocol — now it’s highly important to pick the right device type for tracker since parsing heavily relies on the type of device — not only for settings management but also for mapping of parameters ID into correct names and measurements. Just imagine — we have 70+ channels that are using Teltonika protocol and the protocol upgrade went seamlessly.
  • We also integrated two new protocolsmaxtrack and egts.
  • One interesting article I recommend to our Telematics Hub users for reading on a Sunday afternoon is a short overview of flespi MQTT ecosystem and how it can be used in Telematics. Together with sticky shared subscriptions, these are used to control load balancing and messages distribution.
  • Going further with MQTT — we moved MQTT sessions, subscriptions, and tokens into MQTT state — so now flespi.io is 100% reliant on data from MQTT API and almost does very few REST API calls.
  • To make life with flespi MQTT state easier and for those who like to explore the world, I recommend to open the built-in MQTT Board and subscribe to flespi/state/# topic. After subscription change subscription mode from History into Tree and enjoy the hierarchical way of your account data visualization.
  • And last but not least, we are almost ready to present our MQTT-based data visualization tool called MQTTTiles. It allows creating custom dashboards based on data coming from MQTT. Remember our article about how to manage light bulb from a smartphone using GPS tracker? Now you can reproduce it with MQTTTiles.
    Anyway, we are still updating Tiles with the new features and once it has all the tools for exporting constructed dashboards with the link, we will cover it in details with a dedicated blog article.

In April we will concentrate on the projects of our users to help them to get the most out of flespi. 

We’ll also do the data center network upgrade (fingers crossed). 

As usual, we’ll deliver some new features and do protocols integrations. 

And will enjoy the lovely spring in Minsk.