System Trace

The trace and HTTP access log files of the tenant, read in the browser instead of downloading a .gz, unpacking it and opening it in an editor. It replaces the System Log Viewer of version 1.

When to use it

• You develop an adapter or debug a flow and look at the end of today’s trace log many times an hour.
• You want only the lines that mention a flow name, an artifact or a correlation ID, not a day of framework noise.
• You want to know whether a call reached the tenant at all, and with which status: “the sender never called” or “the call was rejected with 401”.

How it works

1. The page lists every log file of the tenant, newest first, with its type (trace or HTTP) and day. Filter by name or type.
2. Open a file. It is downloaded once and stays available for 30 minutes, so going back to the list and opening it again is instant.
3. Read it in one of three views, which share one search box. Among them: errors and warnings on their own, and for HTTP files the calls with their status.
4. Open one record to see it whole (logger, thread, bundle, node, correlation ID) and copy it into a ticket. “Save as .log” keeps the unpacked file.

Good to know

• Nothing is stored on the server; the file is downloaded and read in your browser. Very large files take a moment.
• The “size” Cloud Integration reports for a log file is a line count, not bytes.
• The level column of the trace file says ERROR on lines that are plain information. The errors view looks at the record, not only at that column.