Content-type: text/html Manpage of TARCHIVE

TARCHIVE

Section: User Commands (1)
Index Return to Main Contents

 

NAME

tarchive - create tar file(s) from sequence(s) of files, then move the tar file(s) to archival storage at various supercomputer centers

 

SYNOPSIS

tarchive [-d] [-m] [-n] [-g] [-dat0] path list

 

DESCRIPTION

tarchive tars files identified by the blank-separated list list, then transfers the resulting tar file(s) to the destination directory path on the archival storage system. Each element of list (list[1], list[2], etc.) identifies a sequence of files named list[1].????, list[2].????, etc.

All options do not work on every center.

 

COMMAND LINE OPTIONS

-d
Delete the local files (including list[1].tar, list[2].tar, etc.) after transfer. If this option is not set, the tar file is copied to archival storage, but the local raw data files and the local tar file(s) are retained.

-m
Migrate data off-line. Some centers allow for user-controlled migration of archival data off-line to long-term storage on slow media. This is useful for users with numerous large data files to move, which, if moved all at once, would threaten to fill the first tier of the archival storage system. By specifying this option, the user immediately makes room for subsequent archival data transfers.

-n
No wait. If set, archive data in background mode. (Not all centers allow this option).

-g
gzip. If set, tarchive tars files named list[1].????.gz, etc.

-dat0
Retain .dat0 files. If set, tarchive does not delete the raw data files, regardless whether -d is set.

 

EXAMPLES

To copy the files run43.txy.0000, run43.txy.0001, run43.txy.0002, etc. to the archival storage directory KH/3D/R25003d/run43, do the following:
tarchive KH/3D/R25003d/run43 run43.txy

To perform the same as above, but for run43.uxy.0000, ... as well, do the following:

tarchive KH/3D/R25003d/run43 run43.txy run43.uxy

 

AUTHOR

Written and Copyright (C) 2000 by Joseph Werne.

 

SEE ALSO

archive(1), darchive(1), dtarchive(1)

 

NOTES

An attempt is made to create the destination directory on the archival storage system, regardless of whether the directory already exists. This can produce a superfluous error informing the user that the destination directory cannot be created because it already exists. It would be better to check to see if the directory already exists before attempting to create it by default.

The machine name is checked via nslookup operating on the output from hostname. Centers currently supported include PSC, SDSC, ERDC, NAVO, ASC, and AHPCRC.


 

Index

NAME
SYNOPSIS
DESCRIPTION
COMMAND LINE OPTIONS
EXAMPLES
AUTHOR
SEE ALSO
NOTES

This document was created by man2html, using the manual pages.
Time: 21:21:46 GMT, January 07, 2001