Adding a New Database File

A new database file is created using the sncreate program. Assume that the file we wish to create is called foobar.dbs and is referred to by the two letter abbreviation fb. (To add a new bank to this new database file, refer to Sec. 8.1.)

1. Stop the server if it is running by issuing the command
   snodb.perl -stop_server
2. Edit /$SNODB_VER/scripts/snodb_lib/customize_hepdb_names.pl to include a new line for the fb database file
3. Run snodb.perl -names to create a modified version of the hepdb.names file which includes reference to foobar.dbs
4. Run snodb.perl -server to restart the server.

Note that it may take a significant length of time before a new database file gets installed on all slave nodes. This is because there is no mechanism for automatically propagating file creation directives using standard HEPDB journal files. Hence, the process depends on individual database czars running sncreate at each slave node. Following that, automatic updates creating an internal directory structure and bank aliases would follow. Before updates containing references to banks contained in the new database file can be propagated, all these steps must be executed at all the slave nodes. The entire process may take as long as a few weeks.