We enhanced flespi functionality with device groups feature. Currently it is only available in REST API, but in October we will add groups to the flespi.io GUI.

Streams, calculators and plugins can now link to a device group instead of a separate device and once you add device to the group it will be automatically linked to all them. Same with ACL - if token provides access to device group, it will automatically provide access to all devices in this group.

Groups are mostly designed to simplify operations on large volumes of devices and automate various assignments.

13 days later

In order to more effectively mitigate DDoS attacks now and in the future we are changing the default gateway IP for all connections originating from flespi network.

If you configured the firewall in your infrastructure to accept traffic from our platform (connections from streams, plugins, proxy and various API or MQTT based channels), please update your configuration as soon as possible:

Old IP: 193.193.165.35 [is active gateway now]
New IP: 185.213.2.10 [will be active gateway soon]

The change will come into effect on Monday, 17 October 2022 or earlier in case we will experience DDoS attack.

This week we are also changing IP addresses for REST(flespi.io) and MQTT(mqtt.flespi.io) services, but this is based on DNS records and the change will go smoothly.

And finally we will setup new range of IP addresses for channels, although current addresses 193.193.165.X will be retained and fully operative as an entry point for existing channels. If your devices are using channel URI with domain name you are totally safe for any change. For those who hard configured in devices IP 193.193.165.X you may be sure we will continue to serve it now and in the future although its routing and DDoS protection level may differ.

4 days later
  • Edited

We applied announced change of default flespi IP for outbound connections from 193.193.165.35 to 185.213.2.10. We recommend all our users who use streams, plugins and special channel protocols (usually OEMs related) to check the status of their systems.

We have fixed a bug in our SSL library that could cause instability on long-term secure TCP connections.

An acknowledgment latency has been added to the log of sent messages. It shows how long it took to send a batch of messages.

5 days later

shal Our primary IP address for channels from now is 185.213.2.30.

IP address 193.193.165.37 is still operational and will continue to operate in the future, however all our uptime tests that we constantly run from multiple locations will not count downtime if this address is not available due to any reason.

Nevertheless we still keep an eye on it inside internal monitoring system and will immediate react in case of any troubles with it but this won't be automatically announced in the NOC comparing with default channel IP addresses 185.213.2.30.

How this may affect you:

  • If you are using IP address in devices configuration you are free to use any of these: 193.193.165.37 or 185.213.2.30 but I suggest to configure newly provisioned devices to new IP address.
  • If you are using channel DNS record in devices configuration (e.g. xxxxx.flespi.gw) than your devices will automatically use new IP address. But in case you whitelisted 193.193.165.37 somewhere in the operator network you have to whitelist also new IP address. If this can not be done quickly please contact us in HelpBox in order to temporary fix DNS records of your specific channels to old IP address.
  • Edited

We're announcing an update that will change the way device telemetry is published into MQTT. At the moment, telemetry is published regardless of it freshness.
We will change this behavior and will publish only the latest telemetry data.
If you're filtering telemetry at client side, you can subscribe to the entire message topic: flespi/message/gw/devices/{id}

The update will be installed on November 23rd.

It is now possible to refer to the last known parameter value from the plugin validator and the msg-expression plugin. For example, you can use speed - #speed to determine the difference in speed between two consecutive device messages.

15 days later

New Modem of type POD IoT Suite is available in flespi.

perform request via interactive API

response will be like

12 days later
5 days later

We plan to get rid of duplicate messages if the stream is subscribed to both the device and the channel. In this case, only the device message will be sent. The update will take place on December 8th.

a month later

If you want to restrict access to devices only from groups, you must use the newly added in-groups selector instead of all in the token ACL.

a month later
  • Edited

We found out that device groups are becoming quite popular tool to simplify and automate bulk management of device connections with various plugins, streams and calculators. That's why we decided to increase the number of included device groups by an order of magnitude per each Commercial Plan:

This change extends flespi functionality and does not add any extra fee. We already implemented it and welcome our commercial users to enjoy more device groups available in their plan.

    We enabled the possibility to store JSON arrays in device telemetry. Previously message parameters that contained JSON array as a value were skipped from telemetry publications and now this feature is unlocked.

    Mostly the change will be visible for users of BLE-enabled devices.

    a month later

    shal
    We discovered that we left the method by mistake during this update. The method is GET gw/channel-protocols/{selector}/commands/{selector}. We will mark it as deprecated and give it one month to transition to the GET gw/channel-protocols/{selector}/device-types/{selector}?fields=commands method.

      5 days later

      We are on the half way to start flespi operations from the second datacenter.

      Second datacenter will contain a set of all flespi services but with dedicated set of IP addresses to split traffic between datacenters. Thus you may need to adjust IP whitelisting configuration on your side. Please update any such configuration before next week (up to April 3).

      For those who do not use any kind of IP whitelisting there won't be any noticable changes.

      For those who protected their integration with IP filter (this is mostly related to streams and some channels) please be informed that flespi may use alternative source IP for its requests towards your integration. Thus any kind of firewall or IP filter where you previously configured 185.213.2.10 as whitelisted IP address please also add 185.213.2.110 which is an alternative source IP for flespi services or just specify our whole subnet 185.213.2.0/24 which is operated solely by flespi.

      For those who are filtering traffic on MNO towards flespi channel's IP 185.213.2.30 please be aware that starting from next week channel host may also resolve to an alternative address 185.213.2.130.

      2 months later

      We have resolved the issue where two channels at the same hierarchy level were simultaneously processing TCP connections from a single device. Now, when a connection is established to any channel, it will break the existing connection.

      16 days later

      We have added a new field to the device called "groups," which contains a list of groups that the device is a member of. The field can be used to select devices based on group name or ID. The field format is as follows:

      [
        {"id": 1, "name": "group1"},
        {"id": 2, "name": "group2"},
      ]

      Examples:

      • To get a list of groups: GET /gw/devices/all?fields=groups
      • To send a command to all members of a group (by name): POST /gw/devices/groups.name=group1/commands
      • To send a command to all members of a group (by ID): POST /gw/devices/groups.id=1/commands
      • To send a command to all devices that are not in any group: POST /gw/devices/groups=[]/commands