Several topics are usually defined in the Kafka cluster, which are created by the cluster administrator. It is not possible to create or delete topics in the cluster from ibaPDA.
If you highlight the Topics node, the list of topics is shown, that are to be written by the data store.
Buttons for the configuration of the topics:
 |
Download the topic list from the Kafka cluster (a connection to the cluster is required) |
 |
Manually add a new topic |
 |
Remove selected topic |
 |
Copy selected cell content to the clipboard |
 |
Paste data from the clipboard into cells |
 |
Move selected Topic up |
 |
Move selected Topic down |
Note |
|
|---|---|
 |
Use <Shift> and <Ctrl> to select multiple cells and copy/paste/delete them. |
If you select JSON as data format, you can set whether digital signals are sent as 1/0 or true/false.
You can reduce the amount of data to be transmitted if you enable the Send only changed values option.
The Compact JSON format option also reduces the amount of data transmitted and removes filler characters, such as multiple spaces, line breaks, etc., from the telegram.
The fields to the right show the number of active topics and active signals.
Meaning of the columns
Active
Here you enable/disable writing to a topic.
Name
Enter an unambiguous name for the topic here.
Key
The key frame property is optional You can define this property with a series of placeholders:
-
-
$identifier: Identifier, defined in the data storage configuration, see chapter Kafka cluster data store
-
$signalid: Signal ID
-
$signalname: Signal name
-
$unit: Signal unit
-
$comment1: Signal comment 1
-
$comment2: Signal comment 2
-
When using the grouped data format JSON (grouped), signal-related placeholders are replaced by empty text.
Metadata
If you chose one of the JSON formats as the data format for the topic, you can choose metadata in the drop-down menu that is saved with the topic.
Available for selection are: Unit, comment 1 + 2, timestamp, signal name, signal ID, identifier. Select the desired metadata with a check mark.
Note |
|
|---|---|
|
The metadata is included in every message. Because all metadata is constant during a measurement, except for the timestamp, this leads to a lot of redundant data written in the cluster. It is therefore important to consider which metadata is actually required. The timestamp is implicitly already included in every message. Adding the time stamp to the metadata of a message therefore also leads to redundant information. From the data analysis perspective, however, it may be useful to enable the timestamp in the metadata. |
Signal reference
In the drop-down menu, select whether the signal ID or the signal name should be used as a signal reference.
Data format
The following data formats are supported:
-
JSON (grouped): all signal data for a certain time stamp is combined in a message
-
JSON (per signal): one message is created per signal and time stamp
-
AVRO (grouped): all signal data for a certain time stamp is combined in a message
-
AVRO (per signal): one message is created per signal and time stamp
You can find examples of the description of the data formats in chapter Description of the data formats.
Signals
In order to configure the signals that you want to write in a topic, select the topic in the structure tree or click on the <…> button in the Signals column.
In the following dialog, you assign the desired signals to the topics using the storage profiles. See chapter Signal selection.