Skip to main content
Ex Libris Knowledge Center

Number of z00p records fewer than should be

  • Article Type: General
  • Product: Aleph
  • Product Version: 20
  • Relevant for Installation Type: Dedicated-Direct; Direct; Local;


After running publish-04 to initially extract records to the $usr_library, the number of Z00P records ("ALEPH Published Documents") is significantly lower than it should be. (And, of course, these records lacking z00p's do not appear in Primo.)

Note: If certain records which *do* have z00p records do not appear in Primo, see links for the two articles in Additional Information below.

Analysis:
Case #1. publish-04 failed for some reason. A common cause of failure is lack of space.... For instance, the $alephe_scratch/xxx01_p_publish_04.00nnn may have: "Exiting due to job suspension."
And the ./xxx01/scratch/p_publish_04_b_1.log, could show:
Error: write_xml_to_clob: ORA-01691: unable to extend lob segment XXX50.Z00P_SEGNAME by 512 in t

See Additional Information #B below for Complete analysis of publish-04 failure.

Case #2. If some time has elapsed from the last run of publish-04, various problems could have intervened.
To see just how many fewer z00p records there are than there should be....

The number of Z00P records for the PRIMO-FULL publishing set (or its equivalent, with a different name) will be equal to the number of records in the logical base (such as "xxx01PUB") specified in column 2 of the ./xxx01/tab/tab_publish entry for the publishing set.

To calculate this number of records for the col. 2 base for the "PRIMO-FULL" publishing set:

a. do the following SQL to get a count of non-deleted bib records {those which have a non-null 245 (title) field in the z13}:

SQL> select count(*) from xxx01.z13 where z13_rec_key in (select z13_rec_key from xxx01.z13 where z13_title is null;

b. do a GUI search or run the ret-03 Service (Retrieve Records Using CCL) for suppressed records: use the Find command from col. 9 of the tab_base.lng entry {example: wst=(suppressed or deleted or analytic or circ-created)}.

c. find the number of z00p records for this particular publishing set:

SQL> select count(*) from z00p where z00p_set = 'xxxxxxx';
<where "xxxxxx" is the publishing set from col. 1 of tab_publish>

d. subtract b and c from a to get the number of bib records which should have a z00p but do not.


If the publishing set is *not* the full database, but some smaller logical base, then that base can be checked in util h/1/10 ("Z0102 Setup") to get a count of the number of "Docs" which are in the base. Then get the count of z00p records as described in "c" above, and subtract it from the count from util h/1/10 to get the count of records in the publishing set's base which lack a z00p.


Resolution:

Case #1. publish-04 failed for some reason. Correct the problem. If the problem is lack of space, add datafiles to the tablespace the z00p resides in.

Case #2. If the count of records missing a z00p is more than, say, 500, then do this:

1. run publish-05 for each affected dataset (such as "PRIMO-FULL") to delete its z00p records

2. run publish-04 for the each affected dataset to properly create its z00p records with:
Publishing Set PRIMO-FULL
From key 000000000
To key 999999999

3. run publish-06 with
"Create Tar File For" "Range of documents"
From key 000000000
To key 999999999
Update Date Flag Y

4. On the Primo side, drop the entire existing STA01 PRIMO-FULL publishing set, run the pipe to harvest the tar file created by publish-06 and index it.

Symptoms

qui

Additional Information


A.  Articles about z00p records which exist but do not appear in Primo: 
p_publish_06: records not processed into PRIMO; re-running **MASTER RECORD**
Analyzing failure to publish Aleph BIB records into Primo

B. Complete analysis of publish-04 failure problem:
The $alephe_scratch/xxx01_p_publish_04.00394 has: "Exiting due to job suspension."

And the ./xxx01/scratch/p_publish_04.cycles file has this:
...
...
0011 + 000500001 000550000 $
0012 ! 000550001 000600000 $
0013 ! 000600001 000650000 $
0014 + 000650001 000700000 $
0015 + 000700001 000750000 $
0016 + 000750001 000800000 $
0017 ! 000800001 000850000 $
0018 ! 000850001 000900000 $
0019 * 000900001 000902924

that is, cycles 12, 13, 17, 18, and 19 did not complete.


And ./xxx01/scratch/p_publish_04.err has this:

b_publish_04_b: failure

Job Suspended !!!

To identify this error in your log files use the command:
grep "FAILURE Tue Mar 11 11:33:14 EDT 2014" $data_scratch/*.log

Which shows:

aleph@test-aleph01(a20_1) GPO01> grep "FAILURE Tue Mar 11 11:33:14 EDT 2014" $data_scratch/*.log
/exlibris/aleph/u20_1/gpo01/scratch/p_publish_04_b_1.log:FAILURE Tue Mar 11 11:33:14 EDT 2014 ==================
/exlibris/aleph/u20_1/gpo01/scratch/p_publish_04_b_2.log:FAILURE Tue Mar 11 11:33:14 EDT 2014 ==================
/exlibris/aleph/u20_1/gpo01/scratch/p_publish_04_b_3.log:FAILURE Tue Mar 11 11:33:14 EDT 2014 ==================
/exlibris/aleph/u20_1/gpo01/scratch/p_publish_04_b_4.log:FAILURE Tue Mar 11 11:33:14 EDT 2014 ==================

And in the ./xxx01/scratch/p_publish_04_b_1.log:

Error: write_xml_to_clob: ORA-01691: unable to extend lob segment GPO00.Z00P_SEGNAME by 512 in t


The $usr_library util a/17/11/1 shows that the Z00P_SEGNAME is in tablespace TS2D:

Z00P TABLE TS2D 53248 6656 13
Z00P_ID INDEX TS2X 12288 1536 3
Z00P_ID1 INDEX TS2X 12288 1536 3
Z00P_ID2 INDEX TS2X 12288 1536 3
Z00P_ID3 INDEX TS2X 12288 1536 3
Z00P_ID4 INDEX TS2X 12288 1536 3
Z00P_ID5 INDEX TS2X 12288 1536 3
Z00P_SEGNAME LOBSEGMENT TS2D 1748992 218624 427

Thus, more datafiles need to be added to the TS2D tablespace.

Category: Publishing


  • Article last edited: 5/1/2014