_historian Messages for our historian feature
Topic structure flowchart LR
topicStart["umh.v1..."] --> _historian
_historian --> |Optional| tagName
_historian --> |Optional| tagGroup
tagGroup --> tagName
classDef mqtt fill:#00dd00,stroke:#333,stroke-width:4px;
class umh,v1,enterprise,_historian mqtt;
classDef optional fill:#77aa77,stroke:#333,stroke-width:4px;
class site,area,productionLine,workCell,originID,tagGroup,tagName optional;
tagGroup -.-> |1-N| tagGroup
click topicStart href "../"
flowchart LR
topicStart["umh.v1..."] --> _historian
_historian --> |Optional| tagName
_historian --> |Optional| tagGroup
tagGroup --> tagName
classDef mqtt fill:#00dd00,stroke:#333,stroke-width:4px;
class umh,v1,enterprise,_historian mqtt;
classDef optional fill:#77aa77,stroke:#333,stroke-width:4px;
class site,area,productionLine,workCell,originID,tagGroup,tagName optional;
tagGroup -.-> |1-N| tagGroup
click topicStart href "../"
[i18n] javascript_required
Message structure Our _historian messages are JSON containing a unix timestamp as milliseconds (timestamp_ms) and one or more key value pairs.
Each key value pair will be inserted at the given timestamp into the database.
Show JSON Schema
Hide JSON Schema The messages in our data model must be JSON using the following format:
{
"$schema" : "http://json-schema.org/draft-07/schema#" ,
"type" : "object" ,
"properties" : {
"timestamp_ms" : {
"type" : "integer"
}
},
"required" : ["timestamp_ms" ],
"additionalProperties" : {
"oneOf" : [
{ "type" : "string" },
{ "type" : "number" },
{ "type" : "boolean" }
]
},
"minProperties" : 2
}
Examples:
{
"timestamp_ms" : 1702286893 ,
"temperature_c" : 154.1
}
{
"timestamp_ms" : 1702286893 ,
"temperature_c" : 154.1 ,
"pressure_bar" : 5 ,
"notes" : "sensor 1 is faulty"
}
If you use a boolean value, it will be interpreted as a number.
Tag grouping Sometimes it makes sense to further group data together.
In the following example we have a CNC cutter, emitting data about it’s head position.
If we want to group this for easier access in Grafana, we could use two types of grouping.
Using Tags / Tag Groups in the Topic :
This will result in 3 new database entries, grouped by head & pos.
Topic: umh/v1/cuttingincorperated/cologne/cnc-cutter/_historian/head/pos
{
"timestamp_ms" : 1670001234567 ,
"x" : 12.5 ,
"y" : 7.3 ,
"z" : 3.2
}
This method allows very easy monitoring of the data in tools like our Management Console or MQTT Explorer, as each new / will be displayed as a Tree.
What’s next? Find out how the data is stored and can be retrieved from our database .