I've just implemented the Nebula Logger in my Salesforce org and I'm curious if I really need to bring over Log__c, LogEntry__c, and LogEntryEvent__e for full visibility? I'm just logging the exception in APEX. It appears that the LogEntryEvent__e will contain all the details I need and that the other Logging records form a redundancy or duplicate reporting. Doing a cost benefit analysis what would the company lose and gain by this decision and would it just be best to send all the object records through? Side note is the LogEntryEvent__e really being sent over to Datadog in realtime or is it being queried on a scheduled bases? It appears to be the later one in my observation, but curious to get confirmation.
Great question, Brandon. From a cost/benefit perspective, I’d avoid sending every Nebula Logger record to Datadog without first confirming which level of detail your monitoring and troubleshooting actually require.
If you’re primarily interested in exception visibility, LogEntryEvent__e may provide the event-level information you need, while Log__c and LogEntry__c can be valuable for retaining and correlating the broader logging context within Salesforce. Sending everything could improve traceability but may also increase data volume, storage, and Datadog ingestion costs.
For the realtime vs. scheduled question, I’d verify the specific Nebula Logger/Datadog integration implementation and configuration rather than assuming from the object type alone. Platform Events can be delivered asynchronously, but the integration architecture determines whether Datadog receives them immediately or through a polling/scheduled process.
Interested to hear what others have observed with this integration as well!
