Skip to content
Nebula Logger

Logging levels

Nebula Logger supports seven logging levels, in descending severity:

ERROR > WARN > INFO > DEBUG > FINE > FINER > FINEST

Whether an entry persists depends on the effective logging level for the transaction. Entries below the effective level are dropped at the framework layer - no CPU spent formatting them, no record inserted.

Precedence

Section titled “Precedence”

The framework picks the effective level in this order (highest priority first):

  1. Per-scenario override - if the transaction has a scenario and a matching LoggerScenarioRule__mdt exists with a level override enabled, that level wins for the whole transaction.
  2. User-level LoggerSettings__c.LoggingLevel__c - if the current user has a user-level record, its level wins.
  3. Profile-level LoggerSettings__c.LoggingLevel__c - if the user’s profile has a record, its level wins.
  4. Org default LoggerSettings__c.LoggingLevel__c - the fallback.

If no level is defined anywhere, entries do not persist. This is intentional - it forces every org to make an explicit choice.

What each level means

Section titled “What each level means”

There is no framework enforcement of what belongs at each level - the levels are conventions. The one most teams settle on:

Level Convention
ERROR Something failed. Requires investigation.
WARN Something is suspicious but not a failure. Might require investigation.
INFO Business milestone. Useful in a normal transaction.
DEBUG Diagnostic detail useful when investigating.
FINE Verbose diagnostic detail - method entry/exit, decision points.
FINER More verbose - individual iterations, intermediate values.
FINEST Most verbose - everything.

Rule of thumb: run production at INFO or WARN. Leave DEBUG / FINE / FINER / FINEST calls in the code - they cost nothing at runtime because the framework drops them before they cost anything. Turn them on via LoggerSettings__c when investigating.

Per-scenario overrides

Section titled “Per-scenario overrides”

To temporarily raise the level for one business process without lifting it org-wide, create a LoggerScenarioRule__mdt record whose Scenario__c matches the value passed to Logger.setScenario(...).

Set:

  • IsEnabled__c = true.
  • The relevant “override enabled” flag for logging level on the CMDT.
  • The target level.

The scenario rule beats the hierarchy for transactions with a matching scenario.

Where next

Section titled “Where next”