University of Southern California
Import process:

Import a Collection Into Documentum

 

This section details the steps necessary to import a collection of both metadata and content. 

The import process will import the object (metadata + content), create any renditions required, and place it on the Documentum lifecycle.

Once it’s approved, the object will be published (via Documentum Site Caching Services).

See the Overview CIS Life Cycle Diagram for detail.

 

  1. Pre-Import:

 

    1. Get the required information from Wayne Shoaf:

·      Specify the renditions needed.  Sample settings are:  1024, 256 and 64

·      Specify which rendition to be used as default

·      Specify the next autonumber to be used (# of digital objects in the collection+ 1)

·      Specify the caption text.  Sample caption:

"Digitally reproduced by USC Digital Archive C2005, First Families Postcards: [filename]"

where [filename] is the name of the image file.

 

    1. Get the required information from Nava:

·      Determine location of staging area for collection

·      Determine size of collection

 

    1. Check disks to ensure enough space can be allocated to Documentum:

·      Currently, CIS has been allocated 25 file systems with 60GB each.

·      As of May 2005, we are at file system #16.

·      Use the following commands to check size:

df –h /var/local/documentum/data/cisprod/content_storage_[xx]

where [xx] is the file system #

du –sh <…>/[collection]

 

  1. Prepare Documentum

 

    1. create location & file storage (use DA)

·      create location  (object: dm_location).  DA > admin > storage > locations

·      create file storage (object:  dm_filestore). DA > admin > storage > file_storage

 

    1. create object cis_collection (use DA)
    2. link virtual staging area to actual staging area (see 1b)

·      change directory to virtual staging area

cd /auto/docu1-02/cismig           

where

         virtual staging area is /auto/docu1-02/cismig                          

and directory structure of cismig is as followed:

            - cismig

            |--  [collection]

            |------  new

            |----------  [batchId]      (batchId = date of staging)

            |--------------  [contents …]

|------  replace

·      link to staging area:

ln – s  [actual staging area (see 1b)]  [virtual staging area /collection/{new/batchId/}]

 

  note 1: after import, manually rename directory from <collection> to <collection>_done

to prevent the import job from picking it up the next time around

  note 2: the import job also renames the files in the collection from <filename> to <fileName>_done

 

    1. update local copy of c:\collection.txt to contain the new collection name (see collection.txt)
    2. run CreateCollection script from a windows client (CisprodCreateCollection.bat)
    3. run UpdateLifeCycle script from a windows client (CisprodUpdateLifecycle.bat)

  note: do not run if other users are updating lifecycle (import jobs, add/modify objects)

since UpdateLifeCycle will uninstall all lifecycles in Documentum)

    1. verify that aliases are created (use Samson)

iDQL: select * from dm_alias_set;

Application Builder: open USC_CISR2 > view lifecycle > status must be set as “INSTALLED”

    1. create derivatives (use Samson) (see createDerivatives.dql)

iDQL: create …

    1. create autonumber (use Samson) (see createAutonumber.dql)

·      iDQL: create …

    1. Ensure that caption text is UTF8 compatible, for example: copyright symbol = Alt-0169 (use sql*plus)

·      sqlplus: select * from derivatives_s;

    1. grant roles (use DA):

·      <collection>_digitizer to cismig (CIS migration user)

·      <collection>_supercataloger to wayne

    1. Ensure that all collections are moved/renamed to <collection>_done or <collection>_wip

 

  1. Run Import Job
    1. Ensure that the Media Server is running
    2. run java program to import xml files

·      cd /var/local/documentum/dmadmin

·      . ./runimport > ~/logs/[logFileName.log] &

    1. check log file for errors

·      tail –f ~/logs/[logFileName.log]

    1. rename directory:  [collection] to [collection]_done

 

  1.  Post Import Sanity Check
    1. check to make sure all renditions are created (use Samson); replace the CollectionName with the actual name

 

select i_full_format, page_modifier, count(*) from dmr_content where any parent_id in (select r_object_id from <CollectionName> where r_is_virtual_doc = 0) group by i_full_format, page_modifier;

 

    1. run java program to create renditions (use Intellij)

·      Java Project: ImportJob:

-       RequestRend64

-       RequestThumbnail  (100)

-       RequestRend256

-       RequestRend1024

-       RequestSid

·      Use the following parameters

-        <docbroker> <user> <password> <collection> tiff <# of rendition>

-       where

·      docbroker = cisprod

·      user = dmadmin

·      # of renditions = total number of renditions

tiff

64

thumbnail

256

1024

sid

 

 

    1. If 1024 image needs caption

·      Login to docu1 as dmadmin

·      Run

-       . ./runtextrend > logs/<filename>

 

    1. check log file for errors