How to correctly map protocol table columns (Kafka collector) to Alarm Console properties ?

47 viewsAlarm Console Connector Visible alarms XML
1
0 Comments

Hi Dojo community !
I’m working on a custom Kafka collector and I’m trying to understand how it maps protocol table columns to the Alarm Console fields.

In my protocol XML, I’ve set all my read parameter :

<code>Monitored = "true"
</code>

This allows me to alarm on the table columns when assigning an alarm template.

However, I’m seeing an issue:

  • When I create an alarm based on a single parameter match (e.g., matching a keyword in one column),
  • The Parameter Description and several other Alarm Console fields remain empty,
  • Which suggests it is not mapping the protocol columns to the Alarm Properties correctly.

My protocol table has the following columns:

<code> AlarmID  
 AlarmName  
 ApplicationName  
 SubcomponentName  
 Impact  
 Urgency  
 Description  
 SupportingData  
 OperatorActions  
 MessageTimestamp (payload)  
 ReceivedTime (DM)
</code>

My question is:
Is there a documented way (or best practice) to understand which protocol table columns get mapped to each field in the Alarm Console?

For SNMP traps, we rely on OID bindings and can easily map values to Alarm properties.
Is there an equivalent mapping mechanism for non‑SNMP protocols (like Kafka)?
Or a method to ensure columns are being mapped so the Alarm Console fields populate correctly?

My goal is to provide proper baseline visualisation for our operations team, and right now the alarms appear partially empty.

Any guidance, examples, or documentation would be greatly appreciated!

Luís Freitas [SLC] [DevOps Advocate] Answered question

1 Answer

0
960 1 Comment

Hi A B M Siddique,

You’re correct that for non‑SNMP protocols such as Kafka, there isn’t an equivalent to OID bindings. In these cases, the mapping of protocol table columns to the Alarm Console fields depends entirely on how each parameter is defined in the protocol XML.
To populate the Alarm Console fields the way you expect, you can use the Alarm-options on the parameters you want exposed. In particular:

  • properties – specifies the format and values that should appear in the alarm’s Properties tab.
  • propertyNames – defines the labels that accompany those values.

These options give you explicit control over how table columns are represented in generated alarms.
Documentation reference: options attribute | DataMiner Docs
I hope this helps!

Kind regards,
Luís Freitas

A B M Siddique Posted new comment
A B M Siddique commented

Thank you very much Luis for your suggestion on this , I will check it out and try to implement it !