HomeCatalogingOther cataloging featuresSHAREit: Correcting failed records

7.3. SHAREit: Correcting failed records

PaILS Support Staff complete a quarterly export of records to AccessPA which go through the full file replacement process so that your library's items are kept current and can be requested through statewide Interlibrary Loan via SHAREit.

Full file exports from SPARK are generated on:

The exports include holdings received or weeded before each of those dates. The export file is uploaded to ACCESSPA within the week and the fail file is available once the vendor (Auto-Graphics) finishes processing it. This usually takes less than a week but can also take up to a month depending on the vendors schedule.

After each export, it is a good idea to verify that things went as expected in the import.  Cataloging staff should take a quick look at Access Pennsylvania to make sure holdings seem correct.  It's a good idea to spot check a few new items, weeded/discarded items, and any others you have corrected since the last upload to make sure things are correct.  If there are problems, send the examples to SPARK Support so we can compile and send on to AccessPA.

Why do records fail to upload?

When PaILS Support staff complete quarterly updates in SHAREit, we receive an email from HSLC with a link to the records that "failed" to load in the database. These records might fail for a number of reasons.

AccessPA will reject a record for upload if said record:

  1. Lacks a title (MARC tag 245); or,
  2. Lacks a publisher (MARC tag 260 $b, or MARC tag 264 $b); or,
  3. Contain fewer than six MARC fields in the entire record.

Some libraries have requested access to the failed records file so they can view and correct their records. Since the records are loaded as the combined catalog, it can be difficult to find and isolate records from a particular library. 

Use this workflow to find your library's "failed" records:

1.  Download a copy of the fail file, pails.fail.mrc

2.  Download a program to open the list of MARC records.  We recommend the application "MarcEdit", which is available for download here: https://marcedit.reeset.net/downloads

3.  Open MarcEdit, and click the "MarcEditor" icon.


4.  At the next window, go to "File --> Open".  Browse to the location where you saved the fail file.  Select "MARC Files (*.mrc)" from the dropdown box, and open the fail file.


5.  Press Control-F to bring up the find function.  In the space next to "Find what:", enter $b plus your library's SPARK short code (ie, $bSOC for Somerset).  Then click the "Find All" button.


6.  Select a line from the results, and click "Jump to Page" to view the failed MARC record. Note that "$bSOC" is found TWICE or more per 852 line, so the correct number of found records in this search is often half of the number of times the "$bSOC" was found.

If you would like to export a list of all your records including the TCN, see directions further down this page.

Please remember:

Once you have located your failed records, and you have identified why the record failed, you still must correct them in Evergreen!

Find the records in Evergreen, make corrections, and save the changes.  This way, the improved records will be included with the next quarterly upload to SHAREit.

When correcting your records, please adhere to the guidelines in Cataloging in SPARK.


Exporting your library Failed Records

Using some of the advanced features of MarcEdit, it is possible to export a list of everything you need to correct, including the TCN of each problem record.  If you wish to do this, here are directions:

1.  Follow steps 1 through 4 from the instructions above.

2.  Once the Fail File is open, go to the "Tools" menu, and select Swap Subfield Data


3.  Make the following changes:

    1. Under "Original Data", in the area beneath "Field", enter 001

    2. Under "Modified Data", in the area beneath "Field", enter 852

    3. Also under "Modified Data", in the area beneath "Subfields", enter t

    4. Under "Search Options", make sure the boxes are checked next to Copy Source and Add to existing field

    5. Finally, click the Process button


4.  A box will appear, telling you how many changes were made.  Click OK

5.  Press Control-F to bring up the "Find" function:

      1. In the area next to "Find what", enter $b plus your library's SPARK short code (ie, $bBANG for Bangor)

      2. Then, click the Find All button


6.  In the search results, the TCNs will be included under "Found Text" as the final entry of each line.  To save your list of records as a text file, click the icon indicated below (the one that looks like a green arrow going into a folder).  Enter a name for the text file, and  click "Save"


7.  You can now open the text file, and copy the TCN from each line.  Go to the Evergreen web client, and from the "Cataloging" menu, select Retrieve Bib Record by TCN

Knowledge Tags


This page was: Helpful | Not Helpful