Debugging the IBM Storage Scale native REST API issues

The IBM Storage Scale native REST API feature introduces a new trace class, grpc, with a maximum level of 6, is available for IO-daemon traces. You can set the new trace class through the normal mmtracectl --set option.

The scalectl command supports a global --debug flag to troubleshoot issues. If no string is provided after --debug, it prints the output to stderr. You can also specify a file to direct output.

Examples:

scalectl list nsd --debug=list_nsd.log

scalectl list nsd --debug (stderr)

scalectl list nsd --debug=stdout

Admin-daemon writes logs to /var/mmfs/scaleadm/log/scaleadmd.log. The log level is determined by the log_level value in /var/mmfs/etc/scaleadmd_config.toml. The possible values are "error”, “info”, and “debug”. If the setting is changed, then it requires an administration daemon restart and applies only to the local node. Admin-daemon logs are rotated by using logrotate. By default, logrotate runs once per day, but the system admin can modify the frequency. The settings for the admin daemon are defined in /etc/logrotate.d/scaleadmd-log.

The scalectl command generates a unique request-ID (uuidv4) per request. This ID is sent to the administration daemon as part of the request and the administration daemon uses this request-ID during the request. This ID is useful for tracing a command from scalectl to administration daemon and back across different logs.