Skip to main content
ExLibris
  • Subscribe by RSS
    • Back
    Aleph
    Ex Libris Knowledge Center

    MARC record load - p_file_01 / p_file_02 / p_manage_18 parameters

    1. Last updated
    2. Save as PDF
    • Article Type: General
    • Product: Aleph
    • Product Version: 16.02

    Description:
    We're purchasing bibliographic records for the Evans Digital Collection online and Shaw-Shoemaker files of Early American Imprints.
    We will be loading over 50,000 records.

    How should I specify the format of the file to be loaded?

    What parameters should I use in the load BIB records process?

    How can I set the system to add a portion of the file at a time (based on previous experience, these records take a long time to process because the authors and titles are often a paragraph long)?

    We would not want to freeze our catalog or cause slowdown while doing these loads.

    We may be doing this on version 16.02, but more likely version 18 if we can get that going soon.

    Resolution:
    We suggest that you not try to break up the file. Following the sequence described below should enable you to process these records over a weekend.

    (We presume that these are all new records and none of them need to overlay existing records.)

    You need to run p_file_01, p_file_02, and p_manage_18.

    * p_file_01 parameters:

    MARC Record Format: 04 <MARC records delimited by '1D' separator> -- this is by far the most common format

    * p_file_02 parameters:

    MARC Record Format: 01 <MARC Record>

    * p_manage_18:

    Procedure to Run: New
    Indexing: DOC_ONLY <None II>
    Fix Routine: <This will vary depending on the input records. "None" may be OK>
    Character Conversion: MARC8_TO_UTF
    No. of Catalogers: M <Multi-user>

    Specifying DOC_ONLY <None II> will let the records load faster. With "Multi-user", you will be able to do this when other cataloging is going on.

    You could start this on Friday pm. It should finish in a couple hours.

    Then, using the "Output File for Logging System Numbers" produced by p_manage_18, you can run p_manage_40, specifying that certain ranges of records be indexed. You could start with 1,000; see (in the abc01/scratch/run_e_01 log) how they run; and then, if they are going OK, specify a range which includes the rest to run over the weekend. (This indexing should take no more than 2 days for 50,000 records.)

    If you want to automatically create item records, you would run p_manage_50 at some point later.

    (Keywords: p-file-01 file_01 p-file-02 file_02 p-manage-18 manage_18 p-manage-50 manage_50 p-manage-40 manage_40)

    • Article last edited: 10/8/2013