|
|
eels_log_archive -a archive_name [-b time_spec] [-d database] [-e time_spec] [-p] [-q SQL_Query] [-s]
The records are stored in the archives in compressed ASCII format, using the UNIX compress(1) command. The first line of the archive provides information on the order of fields in the database. The format string is a vertical bar ``|'' separated list of database fields, that specify the order of fields in the archive. This information is used when restoring the archive to make sure that the data ends up in the correct field of the database. The remainder of the archive file contains a record per line with each field separated by a vertical bar ``|''.
eels_log_archive is a perl(1) script that uses eels_db_query(1Meels) to do most of its work. Either you must specify an SQL query when you call eels_log_archive or a default of ``select * from events'' is used which is then passed to eels_db_query to select records from the database. The eels_db_query command outputs records using a vertical bar ``|'' to delimit each field, so the output of eels_db_query can be directly written to the archive file. The data is not actually written directly to the file. It is first compressed by piping the output through the compress command.
The eels_log_archive command prints status information when archiving which includes the number of records to archive, and a percentage of the total records that have been archived. This information updates continually to give the operator an indication of how the archive is proceeding.
The above example archives all records that originated from ``syslog'' to an archive file named /etc/eels/archives/eels_archive_03161998.