sf-debugging-logs

Use sf apex tail log to stream logs live, or sf apex list log followed by sf apex get log --log-id to fetch a specific log. The tail command auto-creates a trace flag unless you pass --skip-trace-flag.

Create a DebugLevel record, then create a TraceFlag via sf data create record --use-tooling-api with LogType=USER_DEBUG, the user's Id as TracedEntityId, and an ExpirationDate less than 24 hours after StartDate.

Common causes are APEX_CODE set below DEBUG, a truncated 20 MB log dropping older lines, no active trace flag, or trace flags auto-disabled after exceeding the 1000 MB generation window. Check the log header and ApexLog.LogLength first.

Query AsyncApexJob for Status, NumberOfErrors, and ExtendedStatus, which holds only the first error. Each batch chunk is a separate transaction with its own log, so fetch multiple logs and grep each for FATAL_ERROR.

A single log is capped at 20 MB with older lines removed from anywhere, logs generate at most 1000 MB per 15-minute window before trace flags are auto-disabled, and org accumulation of 1000 MB blocks new trace flags until logs are deleted.

Yes. Use the Apex Replay Debugger with checkpoints and heap dumps in VS Code, which captures variable state at specific lines without modifying code. This avoids the CPU cost and PII exposure of debug statements.