General Information
With the Event Schemas, you define how the data coming from the SF Edge devices should be structured. Meaning that depending on your Event Schemas the representation of the data points sent by the SF Edge devices will be different in the Dataset Builder.
An event schema consists of a schema name, key, and its attributes. Every attribute consists of a name, key, data type, and a unit (and if the data type is double a number of decimals). You can picture the schema as a table and the attributes as columns of this table.
The following picture shows how the definition of the schema affects the structure in the dataset builder.
...
The data type of the attributes defines how the data should be interpreted and if the data is valid. The unit helps you describe your data and can make widgets more explanatory.
Note that you do not need to represent all data points of a machine in one schema you can use multiple schemas. For example one schema for sensor data and one for alarms. On the other hand, not everything a machine sends must have all attributes. Every column is allowed to have NULL entries. In general, it depends on your data how many schemas are practical for you.
| Info |
|---|
If you have admin permissions you are able to mark attributes in an event schema for deletion. During the next publishing of the event schema, all the marked attributes will get removed. As an admin, you can also remove a complete event schema via the event schema list. |
| Note |
|---|
After saving your schema you need to publish it if you want an edge client to be able to send data for this schema. |
The status displayed next to every attribute tells you the state the attribute is currently in. There are three states that an attribute can have:
Inactive: This attribute hasn't been published yet
Active: This attribute is published
To delete: This attribute is published and marked for deletion. During the next publish of the event schema, this attribute will get removed.
It is not allowed to create attributes with the metadata keys (Id, Inserted, Size, Thing, Timestamp, Type) explicitly. Every schema will have those attributes automatically.
Matching Schema to SF Edge Configurations
If a SF Edge is supposed to send data to the platform (via MQTT) some configurations need to match the defined event schemas, basically, we need to define which data point goes into which attribute. For detail have a look at Edge Data Routing.
Invalid Messages
If you configured an edge that is supposed to send data to the platform but you can not find any entries in the respective schemas in the dataset builder have a look at the invalid messages. The following video shows you how to access them.
...