How to migrate TTIG to v3 cluster

Hello

I am a bit afraid because an pilot experience to monitor the bud temperature (spring froze) is starting very soon. I am using the v2 cluster and I noticed is going to shutdown soon.

Is there a date? A delay to migrate?

Actually, I am using one application for two vineyard field, but I would like now to create one application by field, then I will need to create two application. I wonder if it is time to migrate and doing all in v3 cluster.

But my action should not stop my measure (for one day, it’s OK), so it’s bit critial.
Second issue, all of my devises and the gateway are in the fields.

I login to v3 cluster and try to create a new gateway to register the gateway which is in the field.
That action will not give troubles to gateway register in v2 cluster?

The step to register a gateway in v3 cluster is a bit different (it look like)

Screenshot 2021-02-12 at 11.32.28

Following the above image, in v3 cluster, I am not sure what I have to enter in Gateway ID and Gateway EUI fields

a Gateway ID is an unique, human-readable identifier for your gateway
A Gateway EUI is a 64 bit extended unique identifier for your gateway

But in my above image, the Gateway ID contain the EUI 64 bit code

In v3 cluster, I supposed, I should give a gateway ID, as TTIG-1
and in the Gateway EUI field, I should copy the id eui-58…?

If now, I want to continue the migration.

  • I create two new applications, one for year field
  • I create all devises. 7 in the application 1, 5 in the application
  • I go to the field, and update the devise software with the key
  • Then, I can delete the devise and application in v2 cluster with may a couple of hour with an interruption of service.

For now my concern is to migrate ma TTIG and TTOG, and my personnal gateway gateways

Many thanks for your clarification. I am afraid making a issue now as the projet is running

Please see

or the TLDR:

  • No it doesn’t support V3 yet.
  • Yes it’s coming
  • We don’t know when
  • Migrate gateways last
1 Like

May be let me ask you a addtional question about strategy.
My project is to measure bud temparture in two wineyard and later more.
For now in TTN console, I have created one application with an HTTP intregartion pointing to a PHP files. That’s work fine.
Would it be a good idea to create two applications (on application by winyard) pointing to the same PHP files? My “worry” is if later a wine owner stop the project, I could delete the application without affecting the othe wine owner. Also, I could detect from which application the measures are collected.

Ok thank, then I will not migrate yet, but after the experiment, in may or june 2021

Hi @pierrot10,

Just seen you additional post about multi tenancy, V3 has support for organisations which allows you to group tenants and their devices, they can of course use the same HTTP integration.

Other stuff you should plan for in V2 to V3 migration include:

  • For OTAA devices they need to rejoin, some devices do this automatically, others need a restart, you may be wandering between the vines.
  • The decoder script format changes from from function decoder(bytes, port) to function decodeUplink(input) where bytes and port and a whole bunch of other data are child objects of the input object.

Until then the best way to segregate tenants is probably having separate applications in V2 and adding collaborators from each tenant to the application.

However in this use case it sounds like your end user are non techies and more about growing grapes so I would argue they should not need TTN access at all and instead should just have some friendly web portal where they can see their data and be abstracted from the tech. Just my thoughts.

Best Regards,

Rob

However in this use case it sounds like your end user are non techies and more about growing grapes so I would argue they should not need TTN access at all and instead should just have some friendly web portal where they can see their data and be abstracted from the tech. Just my thoughts.

Hello, yes that’s correct :grinning:

That question come into my mind, because I am using a Jeedom box with MQTT to generate an alarm when a temprature is below a limit.

All of the measures are sent to a MySQL via TTN and I provide a web application to see the graph of the temperatures.

But to generate the alarm, I chose jeedom. With MQTT, I can read the message sent to TTN. That’s work a bit in parallel. But the advantage, I can send a Telegram message and the wine owner can validate the alarm. Additionally, when a alarm “ring”, the domotic will allow me to ring a siren, or turn on a gyrophare ot the light of the wine domain.

That question, about one or two application by wineyard, come into my mind, because (from Jeedom) I can not detect from which wineyard the message are collected and receive via MQTT. I could update the station, adding a new parameter to identify the field, but I want to keep the messages as short as possible.

As in TTN payload, I can read ‘dev_id’, I could create to application, one for each field

{“dev_id”:“wineyard-1”}
{“dev_id”:“wineyrad-2”}

I beleive that a good idea, specialy if the one owner leave the project.

I am checking my existing application in v2 cluster, and I observed, the Process ID of the HTTP integration , has the name as my application. I beleive the dev_id value is the name of your application and NOT the name of the Process ID of the HTTP integration?

PS: For now the user will not have access to TTN console. They will not understand, as I do not understand wine, excepted in a glass :grinning: :grinning: :grinning: :grinning:

Cheers and have a nice day.

2 posts were merged into an existing topic: Connecting TTIG to The Things Network V3