IBM TS2340 User Guide - Page 175

Solaris Device Driver (IBMtape) The following options can be used to configure the tape monitor daemon running it on the command line. Most options can also be specified in the /etc/tmd.conf configuration file. However, the command line options override any configuration file options. The following options can be used to configure the tape monitor daemon running it on the command line. Most options can also be specified in the /etc/tmd.conf configuration file. However, the command line options override any configuration file options: -s Stop any currently running instance of the tape monitor daemon. -r Restart the tape monitor daemon and reload all configuration settings. -d Turn on drive error diagnostic retrieval and storage. This option is enabled by default. -D Turn off drive error diagnostic retrieval and storage. -p Specify an alternate directory for the storage of drive diagnostic information. Default directory is /var/opt/IBMtape -l Specify a file for writing daemon related log messages. By default, the tmd only writes status information to the syslog file of /var/adm/messages. -y Turns off writing log messages to syslog. -z Turn off compression. By default, the tmd will use a form of file compression to reduce the size of stored diagnostic information. The file name of dump presents some useful information. An example of the dump file is: IBMtape.000001300148.2004-04-09-14:54:14.dump.gz Here, 000001300148 represents the serial number of the tape device, 2004-04-09-14:54:14 is the time stamp for the dump retrieval. A message is also logged in the syslog file of /var/adm/messages after a drive dump is retrieved by tmd. For example, Apr 9 14:54:21 Java tmd[3279]: Drive dump saved to /var/opt/IBMtape IBMtape.000001300148.2004-04-09-14:54:14.dump. Tracing Facility IBMtape incorporates a tracing facility that is useful for performing problem determination. The tracing facility logs diagnostic information to /var/adm/messages based on the control variable IBM_trace. Refer to "Setting the IBM_trace Level for Static Tracing" on page 159 for instructions on how to set the trace value. IBM_trace values range from 0-13 and result in posted messages as shown in Table 29. Postings are cumulative, so trace level 3 also posts items for levels 2, 1, and 0. A trace value of 2 or 3 is suitable for most normal production environments, with little or no degradation of throughput. IBM_trace values of 4 and higher increasingly degrade performance and should generally be used only when directed by IBM support personnel. Table 29. Tracing Facility Trace Level 0 1 Items Traced Severe error conditions only. For installations with extremely small /var file systems, this setting can prevent filling the file system unexpectedly. However, this may be at the cost of not recording messages related to serious device or system environment errors. Device sense data. Sense data can help in diagnosing the source of unexpected error conditions. Chapter 6. Solaris Tape and Medium Changer Device Driver 157