9.2. General Output Database Commands

A number of commands that control the format of the output database are available in several types of output (results, restart, history, and/or heartbeat). These commands are described in detail below.

DATABASE NAME = <string>results_file_name
DATABASE TYPE = <string>database_type(exodusII)
OVERWRITE = <string>OFF|ON|TRUE|FALSE|YES|NO
  (ON|TRUE|YES)
TITLE <string>user_title
PROPERTY = <string>propName = <string>propValue

The first two command lines (DATABASE NAME, DATABASE TYPE) give pertinent information about the database file. The command line

DATABASE NAME = <string>results_file_name

gives the name of the results file with the string results_file_name. If the file is to appear in the current directory and is named job.e, this command line would appear as:

DATABASE NAME = job.e

If the file is to be created in some other directory, the command line must include the path to that directory. If the directory does not exist, all necessary directories will be created automatically.

Two special, shorthand meta-characters may be used in the results file name specification in the input deck: %P represents the number of processors in use for the analysis, while %B represents the base name of the analysis input file. If the %P character is used, for example, a specified database name of results-%P/job.e for a job running on \(N\) processors would be expanded to results-N/job.e, and the actual results files would be of the form results-N/job.e.N.i, where \(i = \{0000, 0001, \ldots, 1023\}\). If the %B character is used, for example, a results database name specified as %B.e in an input file named my_analysis_run.i and the , would yield a results output file named my_analysis_run.e. If no results database file name is given, the default file name is %B.e, the base name of the input file with the extension .e.

If the results file does not use the Exodus II format [[1]], the format for the results file must be specified using the command line:

DATABASE TYPE = <string>database_type(exodusII)

The Exodus II database is supported in Presto and Adagio. Exodus II is currently the only supported format, but other options may be added in the future.

The OVERWRITE command line can be used to prevent the overwriting of existing results files.

OVERWRITE = <string>OFF|ON|TRUE|FALSE|YES|NO (ON|TRUE|YES)

The OVERWRITE command line allows only a single value. The options FALSE, NO, or OFF will cause the code to terminate before existing results files can be overwritten, while TRUE, YES, or ON will cause existing results files to be overwritten by new files (the default status). Suppose, for example, that an existing results is named file named job21.e, and suppose that an input file with a RESULTS OUTPUT command block containing

OVERWRITE = ON
DATABASE NAME = job21.e

is run. In this case the existing file job21.e will be overwritten.

Whether or not files are overwritten is also impacted by the use of the automatic read and write option for restart files described in Section 9.8.1.2. If using the automatic read and write option for restart files, the results files, like the restart files, are automatically managed. The automatic read and write option in restart adds extensions to file names and prevents the overwriting of any existing restart or results files. For the case of a user-controlled read and write of restart files (Section 9.8.1.3) or of no restart, however, the OVERWRITE command line is useful for preventing the overwriting of results files.

A title may be added to the results file with the TITLE command line. The specified user_title will be written to the “title” field of the results file. Some of the programs that process the results file (such as various SEACAS programs [[2]]) can read and display this information.

The PROPERTY command can change some of the fundamental properties of the output database. For example, maximum variable name length, which, in Exodus II format, has a limit of 100 characters. Note, changing this properties is not recommended unless by direction of the development team.

---

[1]

L.A. Schoof and V.R. Yarberry. EXODUS II: a finite element data model. Technical Report SAND92-2137, Sandia National Laboratories, Albuquerque, NM, September 1994. URL: http://infoserve.sandia.gov/sand_doc/1992/922137.pdf.

[2]

G.D. Sjaardema. APREPRO: an algebraic preprocessor for parameterizing finite element analyses. Technical Report SAND92-2291, Sandia National Laboratories, Albuquerque, NM, 1999. URL: http://infoserve.sandia.gov/sand_doc/1992/922291.pdf.